CORNELL UNIVERSITY. THE Sftosmed ^, 3^Iotoer Itibrarg THE GIFT OF ROSWELL P. FLOWER FOR THE USE OF THE N. Y. STATE VETERINARY COLLEGE 1897 8349-1 RB 149.F52T915'''"'''''-''''"^ "I'iiimiiiIm'im*' "^P''"*'s; a critical, experim 3 1924 000 232 243 Cornell University Library The original of tiiis bool< is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924000232243 OTHER WORKS BY Dr. FISCHER PUBLISHED BY JOHN WILEY & SONS, Inc. The Physiology of Alimentation. viii +348 pages. Sii by 8. 30 figures. Cloth, S2.00 net. TRANSLATION Physical Chemistry in the, Service of Medicine. Seven Addresses by Dr. WoI/Fgang Pauli, Privatdooent in Internal Medicine at the University of Vienna. Authorized Transla- tion by Dr. Martin H. Fischer, ix+156 pages. 5 by 7J4. Cloth, ^1.25 net. (EDEMA AND NEPHRITIS A CRITICAL, EXPERIMENTAL AND CLINICAL STUDY OF THE PHYSIOLOGY AND PATHOLOGY OF WATER ABSORPTION IN THE LIVING ORGANISM BY MARTIN H. FISCHER Doctor of Medicine Eichberg Professor of Physiology in the Unimrsity of Cincinnati SECOND AND ENLARGED EDITION NEW YORK JOHN WILEY & SONS, Inc. London: CHAPMAN & HALL, Limited 1915 E.V. No-v^-ia 1 Copyright, 1910, 1911, 1915 BT MARTIN H. FISCHER THE SCIENTIFIC PRESS ROBERT DRUMMOND AND COMPANY BROOKLYN. N. Y. TO C. R. F. Science commences when for a great num- ber of experiences one general conception is found which will embrace all cases. Thus, if you know that a certain remedy has cured Callias of a certain disease and the same rem- edy has produced the same effect on Socrates and on several other persons, that is Expe- rience, but to know that a certain remedy will cure all persons attacked with the same dis- ease is Science. Experience is the knowl- edge of individual things. Science is that of Universals.— Aristotle. PEEFACE TO THE SECOND EDITION These pages give in combined form the contents of the 1909 Nathan Lewis Hatfield Prize Essay of the College of Physicians of Philadelphia, and of the 1911 Caetwright Prize Essay of the Alumni of the College of Physicians and Surgeons of Columbia University, New York, previously pub- lished as separate volumes bearing the titles " (Edema " and " Nephritis." The close association between the two made their appearance in combined form seem advisable. The chief changes which time has rendered necessary consist of additions to the general text embodying the results of later experimental and clinica,l observations in good part not readily accessible to English readers — the main argument remains as before. Throughout the text are mentioned those who with me did the day's work. These references to James J. Hogan, Edmund M. Baehr, Gehteude Mogee, Maeian 0. Hookee, Thomas H. Kelly, Haywaed G. Thomas, William H. Steietmann, Anne Sykes and Carl Hillee but ill express my indebtedness to them. Nor would I fail to acknowledge the intelligent and faithful help of my technical assistant, Josef Kupka. In these paragraphs I can only record my keen appreciation of what they did. To Lauder W. Jones I have often turned for help in matters chemical, to Louis Teenchaed More in matters physical. Alfred Beodbeck made possible the observations on athletes. Chaeles Goosman gave of his time and skill to pre- pare the ' photomicrographs. Chaeles Heckee and Petee Scheerer made many of the other photographs. To all these I would express my sincere thanks. The now old views restated and elaborated in this volume have not gone unchallenged. Where these challenges have VI PREFACE TO THE SECOND EDITION sprung from scientific doubt it has not been difficult to reach common ground by discussion. Where the personal has entered into the spirit of the attack I have succeeded, in part at least, in keeping silent. As this is my wish for the future, will the interested reader of disputed points examine the evidence away from the noise of the pleading attorneys? Much of the older work and most of the newer in this volume was shaped in the Joseph Eichbeeg Laboratory of Physiology in the University of Cincinnati. The atmosphere prevailing there, which recognizes that the new is born as a minority point of view and hence is unpopular, and that the function of a university is to give it sanctuary, has been made possible by the generosity of Harey M. Levy. Martin H. Fischer. University of Cincinnati, 1914. TABLE OF CONTENTS PART ONE THE ARGUMENT (Pages 1-34) PART TWO ABSORPTION AND SECRETION IN INDIVIDUAL CELLS AND TISSUES PAGE I. The Probj^em 37 II. The Absorption of Water by Colloids 40 1. Remarks on Colloids. Nomenclature 40 2. Observations on the Swelling of Fibrin^ 43 3. Observations on the Swelling of Gelatin 57 4. Observations on the Swelling of Gluten 102 5. Hydration and Dehydration in Liquid Colloids 106 III. The Analogy between the Swelling op Certain Protein Col- loids AND the Swelling op Protoplasm 109 1. The Analogy between the Absorption of Water by Certain Pro- tein Colloids and by Muscle 109 2. The Analogy between the Absorption of Water by Certain Pro- tein Colloids and by the Eye 127 3. The Analogy between the Absorption of Water by Certain Pro- tein Colloids and by Nervous Tissue 136 IV. The Biological Signipicance op the Analogy between the Absorption op Water by Certain Protein Colloids and THE Absorption op Water by Dipferbnt Tissues 151 1. Introductory Remarks 151 2. Criticism of the Osmotic Theory of Water Absorption by Pro- toplasm 163 3. Criticism of the Lipoid Membrane Theory 159 4. Adequacy of the Colloid-chemical Theory of Absorption and Secretion 162 5. Absorption and Secretion of Dissolved Substances by Proto- plasm 164 vii Viii TABLE OF CONTENTS PART THREE (EDEMA PAGE I. Introduction 175 II. The Cause of (Edema Resides in the Tissues 178 III. On the Nature and Cause op (Edema 190 1. An Abnormal Production or Accumulation of Acids or Condi- tions Predisposing Thereto Exist in All States in which We Encounter (Edema 191 2. Any Means by which an Abnormal Production or Accumulation of Acid in a Tissue May Be Brought about Is a Means of Producing an (Edema 201 3. Those Conditions which Are Capable of Decreasing the Hydra- tion of (Protein) Colloids Decrease (Edema while Those Un- able to Do So Do Not Affect It 207 4. On (Edema Due to Other than Acid Causes 220 IV. On the Passive Congestion (Edemas op the Kidney and the Liver 225 V. On the Nature and Cause op Pulmonary (Edema 233 VI. Syneresis and the Accumulation of Fluid in the Body Cavi- ties IN (Edema 240 VII. Concluding Remarks 242 PART FOUR ABSORPTION AND SECRETION IN THE COMPLEX ORGANISM I. The General Problem .'.... 248 II. On Absorption 252 1. General Remarks on the Physico-chemical Structure of an Ab- sorbing Sy.'stem in the Complex Organism 252 2. Absorption from the Peritoneal Cavity 255 3. Absorption from the Gastro-intestinal Tract 267 4. Historical and Critical Remarks on the Theory of Absorption. Peritoneal and Alimentary Absorption of Dissolved Sub- stances 271 III. On Secretion 281 1. Introduction 281 2. General Remarks on a Secreting System in a Complex Organism. 282 3. A Model Illustrating Some Phases of Urinary Secretion 283 4. The Output of Water by the Kidney 286 5. On the CoUoid-chemical Action of the Diuretic Salts. How the Saline Diuretics Produce Diuresis 295 6. On the Colloid-chemistry of Sugar Diuresis 305 7. The Kidney in Secreting Water Does Work. Discussion of Some General Conditions Influencing Water Output by the Kidney. Diuretics of the Second Order 314 8. Historical and Critical Remarks on Urinary Secretion 319 9. Transition from the Physiological to the Pathological in Kidney Function 322 10. The Secretion of Dissolved Substances 324 11. Concluding Remarks on Absorption and Secretion. Lsrmph Formation. Vasomotor and Secretory Nerves 327 TABLE OF CONTENTS ix PAGE IV. Maintenance op the Circulating Fluids in the Body 333 1. Why the Blood Remains in the Blood Vessels 333 2. On the Treatment of Shook 346 PART FIVE THE COLLOID-CHEMICAL THEORY OF WATER ABSORPTION AND SOME PROBLEMS IN BIOLOGY, PHYSIOLOGY AND PATHOLOGY. I. TuBGOR, Plasmoltsis and Plasmoptysis 355 II. On the Absorption or Water by Spermatozoa, Epithelial Cells and White Blood Corpuscles 356 III. On the Interpretation op Some Experiments on Water Ab- sorption BY Muscle 359 IV. On the Nature op Hemolysis 364 V. On Growth and Some Growth Phenomena 372 VI. On the Contraction of Catgut and the Nature op Muscle Contraction 377 1. Observations on the Contraction of Catgut 378 2. Interpretation of Experimental Findings 385 3. On the Analogy Between the Described Contractions of Catgut and the Contraction of Striated Muscle 387 4. Historical and Critical Remarks 389 PART SIX NEPHRITIS I. The Thesis 399 II. An Abnormal Production or Accumulation of Acid in the Kidney Occurs in Every Case of Nephritis 401 III. Any Means which Leads to an Increased Production or Accumulation op Acid in the Kidney is a Means op Pro- ducing Nephritis 414 IV. Nephritis Due to Other than Acid Causes 428 V. The Albuminuria 430 1. Introductory Remarks 430 2. Observations on the "Solution" of Colloid (Protein) Gels. 434 3. Critical Remarks 445 VI. The Morphological Changes in the Kidney 447 1. Introduction 447 2. Classification of the Nephritides. Correlation of the Mor- phological Changes in the Kidneys with Some Clinical Manifestations 448 3. The Changes in the Size and in the Color of the Kidney in Nephritis (Cloudy Swelling) 455 4. The Bleeding into and from the Kidney in Nephritis (Hem- orrhage by Diapedesis) 470 5. On the Origin and the Different Types of Tube Casts 475 X TABLE OF CONTENTS PAGE VII. On the Alleged Consequences of Kidney Disease 482 1. On the Relation of Vascular Disease to Nephritis 482 2. On the CEdema of Nephritis 491 3. On the " Uremia " of Nephritis 493 4. Reinterpretation of the Relation of Nephritis to the Clinical Manifestations Associated Therewith 494 5. Remarks on the Etiology of Vascular Disease 499 VIII. The Disturbances in Secretion in Nephritis 501 1. General Considerations 601 2. The Secretion of Water by the Nephritic Kidney 503 3. The Secretion of Dissolved Substances by the Nephritic Kidney 510 IX. Some Experimental Foundations for the Treatment of Nephritis. Fallacy op Salt Restriction in Nephritis AND CEdema 518 1. Introduction 518 2. Asphyxial Nephritis 519 3. Nephritis Produced by Injecting Acid 525 4. Nephritis Due to Temporary Closure of the Renal Vessels. . 529 5. Interpretation of Experimental Findings 533 6. Inhibitive Effects of Alkaline Salts on the Albuminuria of Hard Work 535 X. On the Treatment op Nephritis 537 1. Introductory Remarks 537 2. Diet in Nephritis 539 3. Water Consumption in Nephritis 544 4. The Role of Salts in Nephritis 547 5. More Aggressive Methods of Alkali and Salt Administration . 549 6. The Treatment of Severe Cases of Nephritis 556 7. Clinical Abstracts and Comment 566 8. On CEdema as an Alleged Consequence of Sodium Chlorid Retention 602 9. On the Treatment of CEdema. Comment on the Sodium Chlorid Restriction Therapy 607 10. Judging the Nephritic. Prognosis 614 11. Underlying Principles and Clinical Value of Kidney Efficiency Tests 621 12. On Acidity Measurements of the Urine 629 13. Ammonia Determinations 638 14. Prophylactic Measures against Nephritis 644 PART SEVEN GLAUCOMA I. On the Nature and Cause op Glaucoma 653 II. On the Relief of Glaucoma 655 1. Local Measures 655 2. Systemic Measures 657 III. Some Comments 660 IV. On the Nature of Corneal Opacities 664 V. Closing Remarks 668 PART ONE THE ARGUMENT GEDEMA AND NEPHRITIS PART ONE THE ARGUMENT In this first part is given in running form a r^sum6 of the entire volume. The busy reader may accept as much of the argument as here outUned upon the abbreviated evidence accom- panying it as he chooses. By turning to the page references given in the text the detailed observations will be found on which are based the more dogmatically stated conclusions con- tained in these first paragraphs. PART TWO So practical a question as the treatment of glaucoma, uremia or anuria is intimately associated with the problem of oedema. This problem of oedema, the question of how a cell, an organism or the body as a whole comes to hold an abnormally large amount of water is in its turn but a subheading of that still greater prob- lem, why living matter holds any water at all and why under nor- mal circumstances it holds so constant an amount. Various hypotheses have been proposed by animal and plant physiologists, by pathologists and clinicians to explain such normal and ab- normal water absorption, but all are inadequate to account for more than the smallest fraction of the phenomena observed. The absorption of water by living matter under physiological and pathological circumstances is determined by the colloids contained in it and their state. The major portion of the chemical substances which make up protoplasm exists there in 3 4 CEDEMA AND NEPHRITIS colloid form. A discussion of the properties of colloids with particular reference to the matter of water absorption is there- fore in order. (See pages 37-40.) About the middle of the last century it was noted! that chem- ical substances differ in the rate at which they diffuse through solvents. The group which diffuse but slowly and which are for the most part amorphous, of high molecular weight, without osmotic pressure, and pseudo-soluble, are known as colloids, while those which diffuse rapidly, are of low molecular weight, have osmotic pressure, and form true solutions, are known as crystalloids. Among the former are found glue, gelatin, albumin and starch; among the latter, cane sugar, common salt and lu-ea. While we are seemingly classifying substances we should really speak of the colloid and crystalloid states, for later observations have shown that it is not its chemical character which identifies the colloid, but rather the physical state of the substance, prac- tically all substances being capable of existing in either of these two forms. (See pages 40-42.) The colloids themselves do not all show the same properties. On the one hand there are those which enter into close association with their solvent (lyophilic or emulsion colloids); on the other, those which do not (lyophobic or suspension colloids). Gelatin, dextrin, starch, glue, vegetable fibers, albmnin and gums are examples of the former class; the colloid metals, metallic hy- droxids and sulphids are examples of the latter. (See pages 42- 43.) The former of these two groups is of greatest biological im- portance, for the bulk of protoplasm is made of it. To discover some of the properties of the lyophilic colloids, particularly their relation to the absorption of solvent, the biologically important colloids, fibrin, gelatin, gluten and blood serum are studied. Fibrin absorbs more water (swells more) in any acid solution than it does in pure water. Within certain limits the amount of water thus absorbed increases with the concentration of the acid, but when a certain concentration is exceeded the fibrin swells less than in lower concentrations. The same is true of alkalies. The addition of any salt, even a neutral salt, to the solution of any acid or alkali reduces the amount that the fibrin swells and this the more the higher the concentration of the salt. When equivalent concentrations of different salts are THE ARGUMENT 5 compared some are found more active in this regard than others. Thus, the chlorid, bromid and iodid are less capable of dehy- drating fibrin than the acetate, while yet more powerful are the sulphate, phosphate and citrate. When the effects of basic radicals are compared potassium, sodium and ammonium are found less active than magnesium, calcium or strontium, while copper and iron are most effective of all. Non-electrolytes are comparatively ineffective in reducing the swelling in the presence of an acid or alkali. They are not, however, entirely without effect. Among the non-electrolytes the sugars deserve special mention because these produce a considerable dehydrating effect. When dextrose, levulose and saccharose are compared the last named is found particularly powerful. The taking up and giving off of water by fibrin represents in large measure a reversible process. In addition to the acids and alkalies there exist a number of other substances which are capable of increasing its hydration capacity. Urea and pyridin and possibly some of the amins are to be mentioned in this class. The hydrations produced by these substances are of a different type from those produced by acids and alkalies, for they are not readily reducible through salts, while they are through various non-electrolytes (sugars), a behavior the opposite of that observed in acid or alkali hydration. (See pages 43-57.) What has been said of fibrin is not characteristic of it alone, but is true of all proteins, as illustrated by the identical behavior of gelatin and gluten. (See pages 57-102 and 102-106.) What has been said of the solid colloids holds also for the liquid colloids such as blood serum, fluid gelatin or egg white. The hydration and dehydration in such liquid colloids can be followed by noting the changes in their viscosity, hydration being betrayed by an increase in viscosity, dehydration by a decrease. (See pages 106-109.) The behavior of the more solid colloids corresponds to the behavior of the more solid constituents of our body, as the tissues, that of the liquid with the fluids permeating these soUd structures, namely, the blood, the lymph, the cerebro-spinal fluid, etc. It can be shown that every tissue takes up water or gives it off under conditions identical with those which cause protein colloids to do this. Thus, muscle, eyes or nervous tissues swell more in acid or alkali than in water, and this the more, within 6 CEDEMA AND NEPHRITIS certain limits, the higher the concentration. Beyond a certain optimal concentration the tissues begin to lose water. The addition of any salt to the acid or alkali produces a dehydrating effect, and this is the greater the higher the concentration of the salt. At the same concentration different salts are unequally effective, and here again their order parallels that observed on simple proteins. The non-electrolytes as compared with the electrolytes are less effective. The processes of hydration and dehydration in all these tissues are largely reversible. Urea is observed in all these tissues to lead to an increased hydration. (See pages 109-150.) The acceptance of the view Jthat the colloids of the tissues and their state are primarily responsible for the amount of water held by them constitutes a tacit criticism of every theory of water absorption thus far advanced. These theories may be called for short, the osmotic theory conjointly with which we may consider that modification of it known as the lipoid membrane theory, and the pressure theory. The pressure theory is criticised later. (See pages 151-153.) The osmotic theory assumes that cells are surrounded by semi- permeable membranes through which water but no dissolved substances can pass. The movement of water is occasioned by differences in the concentration of dissolved substances within and -without the cell, the water being carried in the direction of the higher concentration whether existing within or without the cell. In thus accounting for the migration of water into and out of cells it becomes impossible to get dissolved substances through. Such a conception of the living cell is an impossible one because the cell must be and is able to absorb all manner of foodstuffs and get rid of the products of its metabolism. To meet this situation the original osmotic theory has been modified by saying that the membrane is permeable to some or all substances at some or all times. But when this postulate is granted concentration differences between the inside and the outside of the cell can no longer come to pass, for the dissolved substances will simply move from regions of higher to regions of lower concentration. The forces active for the movement of water therefore disappear. Adherents of the osmotic theory can move either water or dis- solved substances, but they cannot move both, and yet in hving cells this must be possible. (See pages 153-159.) THE ARGUMENT 7 The lipoid membrane theory suffers from much the same defects as the original osmotic one. In assuming the outer layer of cells to be fatlike in character, it may become possible to explain more easily the entrance of fat-soluble substances, but since fat is no solvent for water, for salts and for many of the normal products of cell metabolism, this conception is also bio- logically impossible, for all these can and must be able to pass into and out of cells. (See pages 159-161.) The mosaic theory, which holds part of the cell membrane to be " protoplasmic " in character, another part lipoid, suffers from the combined defects of the osmotic and lipoid theory. (See page 161.) In the colloid-chemical theory there is no need for membranes about the cells. The absorption of water is governed by the laws which govern the hydration and dehydration of lyophilic colloids. The differences between the concentration of dissolved substances found inside and outside of the cell are accounted for through differences in relative solubility, differences in ad- sorption and differences in chemical constitution. The cell conceived of as a colloid matrix cannot only absorb and secrete water, but it can also absorb and secrete any dissolved substance at the same time, and the two processes may run in the same direction or in opposite directions as physiological and biological observation demands. (See pages 162-164.) . The conception that cells are surrounded by osmotic or lipoid membranes has been used to explain not only the absorption of water, but also the biological characteristics of the absorption of dissolved substances. In essence this problem asks why dif- ferent parts of the same cell or different cells bathed by the same blood and lymph do not all hold the same amount of dissolved substance. The membrane theory attempts to explain such concentration differences by saying that the membranes are partially permeable or impermeable to these dissolved sub- stances. In discarding the osmotic conception of water absorp- tion we discarded also this mechanism for the maintenance of concentration differences. What have we to put in its place? (See pages 164-165.) We are familiar in physical chemistry with a number of illustrations of inequalities in the distribution of dissolved sub- stances between different phases even though no membranes exist 8 (EDEMA AND JSTEPHEITIS between them. Thus, when a substance is more soluble in one solvent than in another, it will collect in greater concentration in that in which it is the more soluble; or the adsorption pro- perties of the colloids may be different as in different cells; or the presence of certain chemical compounds in one phase permits it to combine chemically and so hold a greater amount of a given dissolved substance than another phase in which such are not present or present in less amount. Finally, all three may be active at one and the same time. (See pages 166-171.) On this basis our conception of the cell becomes that of a mass of protein intimately mixed with more or less fat and fat- like material, the whole immersed in a liquid from which the protein-fat mixture soaks up a certain amount of water and of various substances dissolved in the water, the whole being gov- erned by the laws of equilibrium. (See page 171.) PART THREE The problem of oedema is also a problem in colloid-chemis- try — that of the ways and means by which the normal hydra- tion capacity of the body colloids is heightened. The school of pathologists quite generally upholds the teaching that oedema is produced by changes in the pressure of the circulating fluids of the body (increased blood and lymph pressure) together with an increase in the permeability of the vessel walls. This pressure theory is completely unsatisfactory, for not only may we have extreme degrees of oedema without changes in blood or lymph pressure, but measures which increase blood pressure and should therefore increase oedema are known to produce just the oppo- site effect. In place of the pressure idea, changes in the tissues and cells have therefore been made to play a r61e in the develop- ment of oedema, but the attempt to define clearly the real nature of these was not made until quite recently when it was taught that increases in the osmotic pressure of the cell contents might lead to an increased absorption of water and thus to cedema. (See pages 175-178.) The cause of cedema resides in the tissues, which become cedematous not because water is forced into them, but because they suffer changes which make them suck up water. This is THE ARGUMENT 9 proved by the fact that the severest grades of oedema may be produced in the entire absence of any circulation, and conse- quently in the entire absence of any blood pressure. (See pages 178-190.) A state of oedema is induced whenever, in the presence of an adequate supply of water, the capacity of the tissue colloids for holding it is increased above that which we are pleased to call normal. Any agency capable, under the conditions existing in the body, of thus increasing the hydration capacity of the tissue colloids constitutes a cause of oedema. The accumulation of acids within the tissues, brought about either through their abnormal production or through the inadequate removal of such as some consider normally produced in the tissues, is chiefly responsible for this increase in the hydration capacity of the colloids, though the possibility of explaining at least some of it through the production or accumulation of substances (of the type of urea, pyridin, certain amins, etc.) which hydrate col- loids as do acids, or through the conversion of colloids having but little capacity for water into such as have a greater capacity must also be borne in mind. (See pages 190-191.) As it has already been proved that the amount of water absorbed by a tissue is dependent upon its content of lyophilic colloids it must next be shown that conditions capable of in- creasing their normal hydration capacity are produced in all states of oedema. Of those that might be discussed, an ab- normal production and accumulation of acids is considered the most potent, wherefore it receives detailed consideration. (See page 191.) The proof that such occurs in all states of oedema consists of the following: An abnormal production or accumulation of acids or conditions predisposing thereto exist in all states in which we encounter oedema. Circulatory disturbances, whether gen- eralized or local, conditions which decrease the oxygen-carrying powers of the blood, as the anemias, various states of inanition, the fevers, the chemical changes following death, as well as poisons of various kinds which are followed by oedema, all repre- sent methods by which the chemistry of the tissues is so altered as to lead to an abnormal production or accumulation of acid in them. (See pages 191-201.) Conversely, any means by which an abnormal production 10 CEDEMA AND NEPHRITIS and accumulation of acid may be brought about in a tissue is followed by an oedema. Thus, the acid production occurring in tissues after death or in poisoning by arsenic, morphin, uran- ium, chloroform, ether, cocain, etc., is always followed by a reten- tion of water in the body and an oedema. (See pages 201-207.) Finally, those conditions which are capable of decreasing the hydration of protein colloids decrease oedema, while those which are ineffective in this regard do not do so. Thus, the cedemas developed by amputated frogs' legs laid in water are reduced by all salts, and this the more the higher their concen- tration. At the same concentration different salts are unequally effective in this regard, and this in the order in which they de- hydrate simple proteins swelling in the presence of acid. To meet the criticism that such reduction of oedema is possible in " dead " frogs' legs but not in " living " animals, a parallel series of experiments is introduced on " living " frogs made oedematous by injections of uranium. Sodium chlorid is no exception to the rule; it reduces oedema as does any other salt. (See pages 207-220.) The possible role of other agencies besides an increased acid content in the development of oedema is discussed, and the importance of such known hydrating substances of protein as the alkalies, urea, pyridin and the amins is emphasized. (See pages 220-225.) A number of the specific problems presented by cedema as it affects special organs is next considered. Thus, another explana- tion must be found for the generally accepted belief that the oedema observed in a passively congested kidney, produced, for example, by ligation of the renal vein, is due to an increased capillary pressure and a forcing of fluid into the kidney tissues. Ligation of the renal artery with its consequent decrease in blood pressure is followed by exactly the same kind of change. (See pages 225-228.) Such a result, which cannot be interpreted by any of the older work, is readily understood on the colloid-chemical basis. Whether we deprive an organ of its oxygen supply by preventing the normal efflux of blood from it, or by preventing the normal influx, the resulting accumulation and production of acid is, of course, the same, and so it was to be anticipated that the organ would swell equally in both cases. (See pages 228-233.) THE ARGUMENT 11 This idea can be further tested on the liver, which besides having a (venous) blood supply through the portal vein, receives a second (arterial) blood supply through the hepatic artery, the two blood streams leaving the liver through the hepatic vein. Ligation of the portal vein does not lead to an oedema of the liver, but an extreme grade of it is produced when the hepatic artery is tied, even though there results herefrom a fall in blood pressure. Passive congestion oedema of the liver sec- ondary to heart disease is really produced through interference with the normal oxygen supply to the parenchyma of the liver through the hepatic artery due either to a deficient blood flow because of a defectively acting heart, or because the overfilled veins dam back the arterial blood so that it cannot get into the hver. (See pages 229-233.) The problem of pulmonary oedema is essentially the same as that of the passive congestion oedema of the liver, for the lung also has two blood supplies, the pulmonary circuit , and a ^ supply of arterial blood through the bronchial arteries. While interferences with the pulmonary circuit scarcely lead to an oedema of the lungs, such is promptly produced by interference with the systemic circulation. The most potent method of producing a pulmonary oedema consists in interfering with the oxygen supply through the bronchial arteries. (See pages 233-239.) It remains to account for the accumulations of fluid, so fre- quently encountered in states of oedema, in the serous cavities and " tissue spaces." These accumulations represent dilute solutions of protein. The squeezing off of such dilute colloid mixtures (the " transudates ") by more concentrated and solid ones (the oedematous tissues) is analogous to the syneresis ex- hibited by colloids. When heavily hydrated sohd colloids such as silicic acid or gelatin are permitted to stand, a thin colloid solution separates from them after a time. Such separation is not noted in slightly hydrated colloids, but is marked in heavily hydrated ones, and increases with time. In the same way the severer and more chronic types of oedema are the most likely to be accompanied by accumulations of fluid in the serous cavities and " spaces." (See pages 240-242.) Some concluding sentences indicate how the experimental findings on oedema of older observers are to be interpreted in the terms of the colloid-chemical theory. (See pages 242-246.) 12 CEDEMA AND NEPHRITIS PART FOUR The problem of absorption and secretion in the higher animals seems at first sight very different from this same problem in so simple a structure as the individual cell or tissue. The ameba, for instance, takes up or gives off water according to changes in its surroundings, whereas in a complex organism we find whole organs seemingly set apart for absorption or secretion alone. But we need to appreciate that the mucosal cell, for example, is an absorbing cell only so long as we look at it from the side of the lumen of the gut. From the blood vessel side it is a secreting cell, for what it absorbs from the gut it gives up to the blood. What characterizes absorption and secretion in the higher animals is that under normal circumstances and from the point of view of the organism as a whole, absorption and secretion occuj pre- dominantly in one direction. The reason for this resides in the fact that unlike the ameba which is surrounded on all sides by the same medium, the cells of an absorbing or secreting organ (in a mammal, for example) are through different portions of their cell protoplasm in contact with entirely different media. The effort to get into equilibrium with these produce all the phenomena of absorption and secretion characteristic of the higher animals. (See pages 248-252.) Every absorbing system consists of three phases: 1. The material to be absorbed — essentially an aqueous solu- tion of dissolved substances. 2. An absorbing membrane — physico-chemically, a solid col- loid behaving not unlike a leaf of gelatin. 3. The blood and lymph — a Hquid colloid acting, as a whole, like a (liquid) solution of gelatin. With it are admixed various solid colloids known as blood cells. (See pages 252-255.) In discussing absorption we must at all times distinguish between the absorption of water and the absorption of the dis- solved substances in the water. Absorption from the peritoneal cavity is first taken up. When water is introduced into the peritoneal cavity it is quickly absorbed. This is because the colloids of the peritoneum in consequence of their constant car- bonic acid production are not saturated with water. When the arterial blood enters the peritoneum the carbonic acid produced THE ARGUMENT 13 in the cells is given off to it. This increases the hydration capacity of the blood colloids, which therefore take water out of the peri- toneum. As long as the circulation is maintained and the cells continue to produce carbonic acid, absorption of water must therefore occur. (See pages 255 262.) All salt solutions are absorbed more slowly than pure water. This is because the salts diffhse into the colloid absorbing mem- brane and tend to dehydrate it, thus starting a counterstream of water secretion to meet the normal stream of water absorption. The end result so far as the absorption of water is concerned will represent the algebraic sum of the two. The higher the concentration of the salt solution the greater the counterstream and consequently the slower the absorption of water from the peritoneal cavity. (See pages 262-263.) At the same concentration the different salts in solution delay in different degrees the absorption of water, and this in the order in which they tend to dehydrate protein colloids. The non- electrolytes as compared with the electrolytes are rather in- effective. Only glycerin and the sugars delay definitely the absorption of water by the peritoneum. Both acids and alkalies delay it also. When water is offered the peritoneum in the form of a colloid solution, in a form, therefore, in which it is bound to the colloid, it cannot be absorbed. This is why blood, lymph and ascitic accumulations remain for long periods in the peri- toneum, while pure water, dilute salt solutions, etc., are readily absorbed. (See pages 263-267.) The available facts on the absorption of water by the intestine are explained in identical fashion. Water is absorbed best, and colloid solutions, in which all the water is bound to the colloid, not at all. All salts delay the absorption of water, and this the more the higher the concentration of the salt. Different salts at the same concentration are unequally effective in this regard. A saline cathartic is merely a salt which is very powerful in its dehydrating effect without possessing marked poisonous action. (See pages 267-271.) Critical remarks on previous theories of absorption are entered into and the laws governing the absorjjtion of dissolved sub- stances discussed. (See pages 271-281.) Secretion represents the mirror image of absorption. In this problem, as in absorption, we need again to distinguish 14 CEDEMA AND NEPHRITIS between the secretion of water and the secretion of any dissolved substance in the water. As the kidney represents from both a qualitative and a quantitative point of view the great secreting organ of our bodies, it receives chief consideration, though what is said of it may with little modification be applied to any of the other secretory organs, as the skin, saUvary glands or stomach. (See pages 281-282.) A secreting system in a complex organism is made up of three phases: 1. A secretion which for the most part represents a watery solution of various crystalloids. 2. A secreting membrane which may be likened to a solid colloid like a leaf of gelatin. 3. A source of some kind, the liquid colloid known as blood or lymph. (See pages 282-283.) A model is described which illustrates some phases of urinary secretion. It consists of a layer of fibrin through which solutions of various kinds may pass at constant pressure. Water or a " physiological " salt solution passes through at a certain rate. An acid solution makes the fibrin swell and lowers the amount of " secretion " to the point of stoppage. When to this acid is added any salt the fibrin shrinks and the secretion recommences, the saline diuretics acting more powerfully in this regard than other salts. When acid passes through, the fibrin goes into solution so that an albumin ring is obtainable; this also disap- pears on the addition of salts. (See pages 283-286.) The kidney can secrete water only as it is furnished this organ in " free " form. In absolute starvation no free water comes to the kidney and so secretion ceases. On the other hand, the giving of water by mouth, intraperitoneally, subcutaneously or intravenously, increases the secretion according to the amount given (loss through lungs, skin, etc., ignored). (See pages 286- 287.) When sodium chlorid solutions of different concentrations are injected intravenously, the amount of urine given off increases with the concentration of the salt, a point being finally reached where the water output is greater than the amount injected. Such a result is usually interpreted by saying that the salt " stimu- lates " the kidney and thus, as it were, pulls water out of the blood. What really happens is that the salt content of all THE ARGUMENT 15 the tissues of the body is increased, in consequence of which they give up water, and this " free " water is then added to the amount that is being injected. The " diuretic " action, there- fore, really depends primarily upon an effect on the body as a whole and only incidentally on the kidney. (See pages 287-292.) The conclusion that only free water can be given off by the kidney can be tested by injecting instead of a salt solution a colloid solution in which all the water is bound to the colloid. Injection of no amount of blood or blood serum increases the output of water from the kidney. (See pages 292-295.) The saline diuretics produce their diuretic action as do the stronger sodium chlorid solutions by their effect upon the body as a whole. They act primarily not upon the kidney, but after injection intravenously diffuse into the tissues, and, dehydrating jbhem, aid in bringing free water to the kidney. The diuretic action of any salt is predictable by adding together the dehy- drating effect of its constituent radicals upon a simple protein. (See pages 295-305.) Those non-electrolytes which are cajJable of producing some dehydration of simple protein colloids are also able to dehydrate the body colloids and so to produce diuresis. This diuretic action is again to be attributed, in the main, not to an effect upon the kidney, but to an effect upon the body as a whole. .Just as dextrose and levulose produce equal degrees of dehydration in simple protein colloids, they have when injected intravenously an equal diuretic action. Saccharose, on the other hand, which acts more powerfully on pure pToteins, also acts more powerfully as a diuretic. As the sugars are relatively more effective in producing protein dehydration in high concentrations than in lower ones (the opposite of what is true for salts), they also produce a relatively greater diuresis in the higher concentrations than in the lower. The dehydrating effect of sugar helps to ex- plain not only the dryness of the diabetic's tissues, but his thirst and his increased urinary output. (See pages 305-314.) The kidney in secreting water does work; in other words, in order to get the water out of the blood into the uriniferous tubules energy is required. To obtain it oxidation of various materials in the kidney must occur, and for this oxygen is needed. This explains why every scheme which interferes with the normal oxygen supply to the kidney or with the proper utilization of 16 (EDEMA AND NEPHRITIS oxygen in the kidney is followed by a diminution in the urinary output even though free water is available. Agencies which are capable of aiding in the restoration of a normal oxygen sup- ply to a kidney suffering from its lack constitute a second method of producing diuresis. This is why caffein and its deriva- tives, digitalis, etc., which favor respiration and the circulation of arterial blood through the kidney, act as diuretics, while, on the other hand, chloroform, ether and alcohol which in large doses lead, in toto, to a state of lack of oxygen in the tissues, all decrease the urinary output. Drugs like caffein, digitalis, etc., I shall call for short diuretics of the second order. (See pages 314-319.) The soundness of these ideas on urinary secretion is tested by applying them to the interpretation of the experimental results of other authors. (See pages 319-322.) It is pointed out that the transition from the physiological to the pathological in kidney function is not abrupt, but that we pass gradually from such a state 'as is considered normal to that in which we recognize the pathological extreme of a parenchyma- tous nephritis. (See pages 322-324.) The secretion of dissolved substances by the kidney (or any other organ) is a problem that must be considered independently of the secretion of water. A secretion of water is necessary before we can get the secretion of any dissolved substance. Urine is secreted primarily as water, and only secondarily as it washes down the tubules do substances come to be dissolved in it, giving it the qualities of urine. The feature which it has been difficult to explain in secretion is the quantitative differences between the concentration of any dissolved substance in the blood and the concentration of this same substance in the urine. Upon the existence of such differences has been based a faith in a " vital " element in secretion. Actually, such a conception is premature if not absurd. The distribution of a dissolved substance be- tween any three such phases as blood, kidney substance and water (the original urine) is a matter of equilibrium, and this is by no means sflways attained when the concentration is the same in all. Tne equilibrium point may be shifted in either direction, depending upon the solution properties, the adsorption char- acteristics and the chemical nature of the three phases. (See pages 324-327.) ^ THE ARGUMENT 17 The formation of lymph is in many respects analogous to the secretion of urine and is governed by similar laws. Anything that makes the cells or tissues of an organ give up water increases lymph flow, a lymphagogue being any substance which will aid in dehydrating the tissues of the body. (See pages 327-328.) Some remarks on the vasomotor mechanism follow. Changes in the amount of blood going to a gland are controlled by this. As we would expect, vasodilatation with its better supply of oxygen and free water makes for secretion, while vasoconstriction does the reverse. Vasodilatation with an adequate arterial blood flow may take place and yet there be no secretion from a gland, but only when poisons have been introduced which keep the gland from utilizing the oxygen. During their so-called periods of rest the gland cells swell and develop granules, while during activity they shrink and the granules disappear. The compli- cated interpretation given these facts is replaced by the simple statement that in the absence of a circulation the gland cells develop an oedema which disappears when the blood vessels dilate and more oxygen becomes available. The granules repre- sent precipitates of a second colloid as the acid content runs up, which disappear when with a better oxygen supply the acid is removed. A so-called resting gland represents the parallel of what in pathology we call cloudy swelling. The existence of secretory nerves is questioned. They are vasomotor nerves. The distribu'tion of secretory nerves and vasomotor nerves is identical. Secretion does not occur without coincident vaso- dilatation. Vasodilatation may occur without secretion, as when defectively oxygenated blood is furnished, but the reverse, never. A large arterial blood supply is furnished to some glands con- stantly, to others temporarily through vasodilatation governed by nferves. (See pages 328-330.) Whatever favors secretion interferes with absorption, and vice versa. (See pages 330-333.) Looking at the problem of secretion from another viewpoint it is now asked why all the blood does not pass out of the body as urine or some other secretion. Why does the blood remain in the blood vessels? The blood remains 'in them because all its water is bound to colloids. It is for this reason that the in- travenous injection of no amount of blood or of any other colloid solution in which all the water is combined with colloid is followed 18 (EDEMA AND NEPHRITIS by an increased urinary output. On the other hand, a salt solu- tion at once leaves the body because its water is " free." The lymph remains in the lymph vessels for the same reason. (See pages 333-345.) These facts are of importance in establishing the principles that must guide us in any attempt to increase the blood pressure in patients suffering from an abnormally low one. The salt solutions injected into the blood vessels by way of increasing the pressure produce their good effects but temporarily because they consist of free water which leaves the body in the secretions or is sucked out of the blood vessels into the tissues. Only transfusion mixtures in which all the water is held in colloid combination can remain in the blood vessels. It is for this reason that transfusion with blood, blood serum, hydrocele fluid, ascitic fluid or a properly prepared gelatin solution yields good and lasting results. (See pages 345-351.) PART FIVE The normal, abnormally low and abnormally high water content of cells discussed by the biologists under the terms turgor, plasmolysis and plasmoptysis are next considered, and it is indicated how these phenomena as observed in spermatozoa, epithelial cells, white blood corpuscles, muscle and other cells and tissues when subjected to the action of acids, alkalies, salts and various non-electrolytes, are more easily explained through colloid-chemistry than through the older, more' popular "osmotic " notions of water absorption. (See pages 355-364.) The problem of hemolysis receives special consideration. It is shown that changes in the size of the red blood corpuscles and the escape of hemoglobin from them are frequently asso- ciated, but not identical, processes. The changes in size follow the laws of water absorption by simple colloids, the loss of color entirely different ones. The hemoglobin and the stroma of the corpuscle are united as an adsorption compound similar in nature to the combination existing between certain colloids and dyes. The parallelism between the laws governing the absorption of water and the loss of color by carmine-stained fibrin and the ab- sorption of water and loss of color by red corpuscles is described in detail. (See pages 364-372.) THE ARGUMENT 19 The source of the energy for growth, which is defined as in- crease in volume, is found in the sweUing of the colloids produced in the process of growth. The mechanism by which curvatures are produced in consequence of the directive action of various external stimuli (tropisms due to light, heat, chemicals, elec- tricity, water) is found in the unequal swelUng of hydrophilic colloids, and it is indicated how many of these phenomena can be mimicked in the laboratory by the use of cylinders, strips and leaves of gelatin irregularly painted with acids, alkalies and other substances. (See pages 372-377.) Catgut strings when subjected to the action of acids shorten. This shortening is due to an absorption of water. They lose their water and lengthen again when the acid is washed out or neutralized. Various salts affect this contraction and relaxation entirely similarly to the way in which they affect the absorption and giving off of water by various protein colloids. When such contractions are registered upon a drum, a series of curves is obtained identical with those produced by the contraction and relaxation of muscle. By varying the conditions surrounding the catgut the phenomena of rigor, fatigue, staircase, residual contraction, increased tone and tetanus can be mimicked. The analogy between the contraction and relaxation of catgut and the contraction and relaxation of muscle under similar circum- stances leads to the conception that the muscle contraction represents a problem in colloid-chemistry, the contraction repre- senting a swelUng and shortening of the muscle fibrils under the influence of a temporary acid production; the relaxation, the neutralization of this acid with loss of water. The contributions of various authors toward the establishment of such- a colloid- chemical theory are reviewed. (See pages 377-396.) PART SIX The term nephritis is used to cover that symptom complex which clinically is characterized by the appearance of casts and albumin in the urine, by certain morphological changes in the kidney, by changes in the amount of water given off, by changes in the amount of dissolved substances secreted in the urine and by the associated oedema, increased blood pressure and cardiac hypertrophy. The changes that characterize nephri;^ 20 CEDEMA AND NEPHRITIS tis are colloid-chemical in nature and due to a common cause — the abnormal production or accumulation of acid and of sub- stances which in their action upon colloids behave like acid in the cells of the kidney. To \he action of these upon the colloid structures that make up the kidney are due the albuminuria, the specific morphological changes noted in the kidneys, the associated production of casts, the quantitative variations in the amount of urine secreted, the quantitative variations in the amounts of dissolved substances secreted, as well as the other signs of nephritis which appear in direct connection with the kidney. The alleged consequences of kidney disease such as oedema, high blood pressure, uremia, etc., are not consequences, but accompanying signs and symptoms which demand separate discussion and analysis. Proofs of the correctness of these contentions follow two lines: 1. A consideration of the chemical factors which bring about the colloid changes. 2. A consideration of the colloid changes themselves. (See pages 399-401.) As under the first heading, an abnormal production and accumulation of acid is considered of chief importance, it re- ceives main emphasis. The proof that an abnormal production or accumulation of acid occurs in the kidney in every case of_ nephritis comes from three directions: 1. The acidity of the urine determined either by titration or by measurement of its hydrogen ion content is constantly high. 2. The blood shows a so-called decreased alkalinity. 3. Dyes of various kinds which show a characteristic color only when the acid content of a tissue is sufficiently high and which show no color when injected into normal tissues, stain the kidneys when the signs and symptoms of a nephritis are present. (See pages 401-414.) On the other hand, it can be shown that any method which leads to an increased production or accumulation of acid in the kidney is a means of producing nephritis. Direct introduction of sufficient acid into the organism is followed by the appear- ance of casts and albumin in the urine and a diminished output of water. When the acid is produced by the organism itself, as in hard muscular work (long marches, athletic games, running matches, etc.) these signs again follow. The real reason why THE ARGUMENT 21 the signs of a nephritis appear after hard muscular work is be- cause the lactic acid produced in the muscles and necessary for their contraction is not oxidized as fast as formed. In other words under such " normal " circumstances there occurs a sufficient accumulation of acid in the body and in the kidney specifically to give rise to albumin and casts in the urine if the normal supply of oxygen to the tissues is reduced. Under path- ological circumstances a reduction in oxygen supply comes to pass in heart disease, respiratory disease, the anemias, carbon monoxid poisoning, etc., and this explains why albumin, casts, etc., appear in such cases. Under all these circumstances the original conditions leading to the nephritis lie entirely outside the kidney. Direct inter- ference with the oxygen supply to the kidney itself, as through pressure upon its blood vessels, arterio-sclerosis, etc., is followed by a local production or accumulation of acid in this organ and by the signs of nephritis. A final way of causing an abnormal production or accumulation of acid is by interfering with the power of the kidney to utilize oxygen even though this is supplied in sufficient amount. The. metallic salts, the alkaloids, the anesthetics and other poisons belong to this group. The albu- minuria of the newborn and that occasioned by salt starvation and excessive consumption of water receive consideration. (See pages 414-428.) Incidentally, it will be observed that the argument regarding the nature and cause of nephritis is much the same as that pre- viously given for the nature and cause of oedema. Actually, nephritis is in good part an oedema of the kidney. As noted in discussing oedema, other substances besides acids, notably the alkalies, urea, pyridin and certain amins are capable of increasing the hydration capacity as well as the solution of proteins. Since swelling and solution of its proteins characterize the nephritic kidney we are not surprised to find these same substances capable of inducing a nephritis. In cases of alkali poisoning, or when alkali is introduced intravenously into animals, all the signs and symptoms of a nephritis develop. (See pages 428-430.) How now does such a single factor as an abnormal produc- tion and accumulation of acid in the kidney lead to the signs and 22 (EDEMA AND NEPHRITIS symptoms of a nephritis? Acid acting upon protein makes it not only swell, but go into solution. The albuminuria of nephritis not traceable to gross lesions, as to bleeding, to diapedesis of red or to migration of white blood corpuscles, etc., is due to such a solution of the kidney proteins in the urine. (See pages 430-447.) Before discussing the mechanism by which each of the dif- ferent morphological changes characteristic of nephritis is brought about, it is necessary to come to an understanding of what is the relation to each other of the different types of nephritis which we recognize clinically, and how these are correlated with anatomical findings. There is only one kind of nephritis, parenchymatous nephritis, but this may affect all the kidney or only spots in it. When the former occurs, as after an intoxi- cation which involves the whole kidney, we speak of a general- ized parenchymatous nephritis. Anatomically this is a large swollen kidney, while cHnically it is characterized by much albumin, many casts and little or no urinary secretion. If the patient does not die, one of two things may happen so far as the kidneys are concerned. They may recover entirely or pieces in them may die, be absorbed, and naught but a scar remain to mark the place of death. If this happens a secondarily con- tracted kidney (" small red kidney ") results. As long as one- quarter of the total kidney substance is saved it suffices for the patient's needs, and he may live and die with this without ever being aware of his state. Nor can his state be recognized clinically. (See pages 447-449.) Depending upon the factors responsible for it, a spotty paren- chymatous nephritis may also recover or the involved areas die. If the latter occurs, replacement by connective tissue again follows, and the " small red kidney," which, as morpholo- gists, we call chronic interstitial nephritis, again results. (See pages 449-450.) The mechanism, however, by which a spotty nephritis is brought about is usually totally different from that which brings about a generalized nephritis. A poison circulating in the blood affects for the most part the whole kidney at once and uniformly. For the focal lesions there must be focally acting causes. Most commonly these are found in the changes of blood vessel disease which lead to a defective blood supply with destruction and death of one piece after another of the kidney. As the con- THE ARGUMENT 23 nective tissue which replaces the lost areas has been very gen- erally regarded as the cause of the necrosis instead of consequent thereto, these Iddneys are called primarily contracted kidneys. As the small areas of kidney substance become involved there appear a few casts and a little albumin in the urine. But since the kidney between these spots is functioning normally and since one-quarter of the total kidney substance suffices for all ordinary needs, the urinary output in such patients remains normal. As will be shown later, the high blood pressure, cardiac hypertrophy, etc., so frequently observed with this type of chronic interstitial nephritis, are not the consequences of kidney disease, but ex- pressions of the vascular disease here held responsible for the kidney lesions. (See p'ages 450-453.) Infections of the kidney are not ordinarily considered with the nephritides, but they might as well be. In' the infections the poison is simply produced within the kidney itself instead of being carried into it from elsewhere in the body. Depending upon the amount of kidney involved, su.ch infections may give rise to either local or generalized nephritides, accompanied in the former case by few casts, little albumin and a normal water output, in the latter by the reverse. (See page 453.) The remaining portions of healthy kidney substance consti- tuting the small red kidney and anatomically diagnosed as chronic interstitial nephritis, whether secondary to a general- ized parenchymatous nephritis or consequent upon blood vessel disease, may at any time and for various reasons become the seat of a generalized parenchymatous nephritis. When this happens, the normal or increased urinary output with few casts and little albumin gives way to a diminished output with many casts and much albumin. (See pages 453-454.) The portions of a kidney involved in nephritis show: 1. An increase in size. 2. A loss in normal color due to the appearance of granules in the affected cells. 3. The appearance of blood corpuscles extravascularly. 4. Evidences of a falling apart of the kidney leading to the formation of casts. (See pages 454-455.) The increase in size is due to a swelling of the kidney colloids under the influence of the abnormal production and accumu- lation of acid characteristic of nephritis. The change in color 24 (EDEMA AND NEPHRITIS is due to the precipitation under the same circumstances of a second colloid of the nature of casein. The two together con- stitute " cloudy swelling.'' (See pages 455-470.) The bleeding from the kidney in nephritis is in part due to rupture of the blood vessels, in part to diapedesis. The mechanism of diapedesis is discussed. The tissues ordinarily are sufficiently stiff to prevent the red blood corpuscles from sinking into them. After absorbing water under the influence of an acid they become less rigid (their viscosity is decreased) and the corpuscles are now able to penetrate them. The process of diapedesis can be mimicked by allowing mercury drops to move in all directions through a solid gelatin. (See pages 470- 475.) When the surface of a fresh kidney is scraped one obtains only a granular detritus consisting of broken cells, but if the kidney is treated with a dilute acid or is simply subjected to the action of such as are produced post-mortem, it falls apart into its constituent elements, the cement substance dissolving first, while the epithelial cells stick together and slip out of the tubules as casts. The casts are originally epithelial, but become granular as the action of acid upon them is prolonged. When the concentration of the acid is increased the casts become hya- line. By proper regulation of acid and salt content the hyaline casts may be reconverted into granular casts. The danger is pointed out of drawing too large clinical conclusions from the character of the casts found in the urine, which within its ordinary limits of acidity, salt concentration, etc., can so markedly change them. (See pages 475-482.) Many of the clinical manifestations observed in patients with kidney disease are considered consequences of the impair- ment of kidney function. While there are Consequences to loss of kidney function, those clinically regarded as such almost without exception do not belong in this group. Blood vessel disease, high blood pressure and cardiac hypertrophy are not secondary to loss of kidney function. The primary disturbance in chronic interstitial nephritis associated with vascular disease and changes in the heart is the vascular disease, and the change's in the kidneys, heart and other organs are secondary to it. This can be proved both from clinical observation and experiment. The worst cases of nephritis in which there is greatest loss of THE ARGUMENT 25 kidney function, as in the toxic nephritides occurring in scarlet fever, pregnancy, etc., there is no high blood pressure; nor when the kidney substance of animals is experimentally reduced to the physiological minimum do vascular disease, high blood pressure or cardiac hypertrophy develop. The primary change being blood-vessel disease, it is easily understood why the other signs and symptoms must follow. In consequence of the de- struction of one piece of the kidney after another by the changes characteristic of vascular disease as displayed in the small blood vessels of this organ with subsequent replacement by scar tissue, the kidney undergoes gradual diminution in size. The pieces die with the signs characteristic of parenchymaitous nephritis, but because what remains of the kidney is healthy and but one- quarter is necessary to maintain life, patients with this type of chronic interstitial nephritis show no change in urinary output. (See pages 482-488.) The hypertrophy of the heart is not consequent upon loss of kidney function, but is the result of a demand for increased work and increased power. The source of these demands resides in the changes in the blood vessels. Because of their decreased caliber an increased force is necessary to drive the blood through them, and because they are inelastic the heart is required to push the blood through them in the time of a systole alone, instead of as ordinarily in the time of a systole, plus a diastole, plus a pause. More work is therefore done and in less time. In mechanics more work in less time requires a more powerful machine, and the hypertrophy of the heart is an expression of this law in the body. (See pages 488-490.) As it is only through the increased blood pressure that the different organs of the body are adequately supplied with blood when vascular disease is present, the high blood pressure cannot in itself be regarded as something evil, but must be looked upon as decidedly good. Except in hemorrhage, treatment which merely lowers blood pressure is of no value and often dangerous. Because of the weakened blood vessel walls, all those agencies which tend to increase blood pressure even in normal individuals should be controlled, but beyond this only that therapy has value which tries to combat the underlying cause of the increased pressure, namely, the vascular disease itself. " Hypertension " is not a clinical entity any more than is " fever " or " dropsy " 26 CEDEMA AND NEPHRITIS and schemes for reducing it which do not consider its underlying physiology and pathology are worthless. (See pages 490-491.) The oedema observed in some cases of nephritis, particularly in the parenchymatous types, is also not secondary to the kidney disease. Patients suddenly deprived of their renal function, as through accidental removal of an only kidney, or animals similarly treated by removal of both kidneys, develop no oedema. On the other hand, patients or animals poisoned with any of the popular " kidney poisons " like arsenic, salvarsan, uranium, etc., develop an oedema in a short time. But the generalized oedema is not secondary to the kidney disease, but represents in the in- volved tissues the same type of change as that which in the kidney we call nephritis. (See pages 491-493.) Similar clinical observations and experiments prove that the headache, vomiting, disturbances of vision, stupor, respira- tory changes, coma and death, clinically regarded as signs of a " uremia," are not secondary to loss of kidney function. Neph- rectomized patients or animals show no such symptoms. The alleged " uremic " symptoms are due to oedemas of various portions of the central nervous system and are caused by the same agents which are inducing the oedema elsewhere in the body, including the kidney. (See pages 493-494.) A reinterpretation of the clinical manifestations associated with nephritis is attempted on the basis of these facts. (See pages 494-499.) Since vascular disease is regarded as the cause rather than the consequence of kidney disease, its etiology is discussed. The pathological changes observed in the blood vessels in vas- cular disease are focal in type in even the most advanced cases. The whole of the media, or the whole of the intima of an aorta, for example, is never involved. A generalized intoxication cannot, therefore, underlie it. To get the focal lesions we must have focal causes, and infectious emboli are undoubtedly to be regarded as the underlying cause. The emboli lodge in the smaller blood vessels and thus give rise to the local necroses observed in the kidney, retina, brain, etc. When they lodge in the small blood vessels supplying the coats of the larger ones (as in the aorta) the patchy spots of softening, absorption, connective tissue overgrowth, and calcification characteristic of " atheroma " and arteriosclerosis follow. On this basis the importance of looking THE ARGUMENT 27 for sources of infection in patients with vascular disease is em- phasized. (See pages 499-501.) Disturbances in secretion in nephritis are of two types, those affecting the output of water and those affecting the output of dissolved substances. (See pages 501-503.) The introduction of acid into the kidney by any means what- soever is followed by a prompt decrease in urinary secretion even to the point of absolute stoppage. The same factor so largely responsible for the other signs of nephritis is therefore responsible for this also. How this acid works is discussed. (See pages 503-510.) The changes in the excretion of dissolved substances by the kidney must be considered under two headings. The decrease in absolute amount put out is secondary to the decrease in total amount of water secreted. The change from the normal in the relative proportions of the dissolved substances given off is secondary to changes in the adsorption properties, etc., of the kidney colloids as afftected by the presence in them of acid and similarly acting substances. The action of acid^ in thus altering the adsorption properties of protein colloids is illustrated by experiments on the taking up and giving off of dyes by fibrin. (See pages 510-518.) Experiments are now introduced to establish further some principles of treatment. As the various salts, including sodium chlorid, were found to decrease the swelling and solution of proteins, it was to be expected, since similar changes characterize nephritis, that they would also be able to inhibit or make these subside when experimentally induced. Asphyxial nephritis, that secondary to intravenous injection of acid, and that con- sequent upon temporary clamping of the renal artery are studied in this regard. The decrease in urinary output, with the appear- ance of albumin, blood and casts in such urine as is secreted, can be overcome almost entirely if animals rendered nephritic by such experimental means are treated with various salts. Sodium chlorid is no exception to this rule. (See pages 518- 535.) To meet the objection that such suppression of the signs of nephritis can be obtained only in animals, a series of observations on the albuminuria consequent upon hard muscular work is introduced. When athletes are fed liberally on citrus fruits 28 OEDEMA AND NEPHRITIS the albuminuria developed is decreased in severity. (See pages 535-537.) Were we to formulate a general rule for the prophylaxis and the treatment of nephritis we should evid'ently have to say that this lies in an avoidance and removal as far as possible of every condition that favors the abnormal production or accumulation of acid in the kidney, or of such other substances which in their effects on tissue colloids behave like acid. After this is done, attention must be directed to combating the effects of such conditions as cannot be removed. The rule to be followed may be summarized in the words: Give alkali, salts and water. To this may be added a fourth rule: Give dextrose. The alkali is needed to neutralize the acid present in abnormal amount. The salts are indicated, and sodium chlorid is no exception, because the changes induced in the body colloids by the action of acids upon them are counteracted by adding salt. Water is necessary in order to have this present in the body over and above the amount necessary to saturate its colloids so that free water may be left over out of which to make urine. Dextrose or other carbohydrates are given not alone from a chemical point of view, in that an abnormal production and accumulation of acid is frequently the consequence of carbohydrate starvation, but because the sugars are peculiarly powerful in reducing certain types of increased hydration in protein not produced by acid. (See pages 537-539.) The importance of the diet in treatment is considered. Foods high in the mineral acids or in those organic acids (benzoic, oxalic, tartaric, etc.) which cannot be readily oxidized to carbonic acid in the body should be avoided unless special pains are -taken to give with such foods an adequate supply of alkali to neutralize these acids. A protein diet yields, after oxidation in the body an excess, roughly, of 25 per cent of acid, while a vegetajjle and fruit diet yields under the same circumstances a 25 per cent excess of alkali. Hence the advantage of the vegetable diet over the meat diet. Practically, however, drastic restrictions in the dietary are not to be recommended. It is easier and better for the patient to be liberal with the diet, but to protect him against the effects of an excess of acid by a continuous feeding of alkali to the point where his urine is kept persistently neutral to litmus. (See pages 539-544.) THE ARGUMENT 29 The question of water consumption in nephritis resolves itself into two parts, into the use of water in cases where nephritis is likely to arise and into its use in the established case. The patient neetls water in order to have free water available out of which to make urine, and since most of the nephritides depend in the end upon an intoxication, water is needed to reduce its concentration, for intoxication depends upon concentration. For both purposes water must be given both day and night. The giving of water does not materially increase the work thrown on the heart as generally taught, wherefore heart disease, blood- vessel disease, etc., do not contraindicate its use. Neither is there any scientific reason against giving water in chronic interstitial nephritis. The objections to the use of water are two: first, when the hydration capacity of the body colloids is increased, the giving of water makes possible their swelling; second!, pure water in washing through the kidney washes out not only poison- ous substances of which we would be rid, but also salts of various kinds which we would keep. To give the organism the benefits of water without its accompanying bad effects, we need to give along with the water properly chosen salts in sufficient amount. (See pages 544-547.) The natural and artificial means of establishing and main- taining a proper salt concentration in the body are discussed. (See pages 547-549.) When the gastric route alone does not suffice to get adequate amounts of alkali and salt into the organism the rectal or the intravenous route may be employed. The formulae of proper solutions to use under such circumstances are given. Never must alkaline mixtures be used subcutaneously or intramus- cularly. (See pages 549-566.) The case histories of a series of patients are given and com- mented upon to illustrate the practical use of the principles of treatment outlined in this volume. (See pages 566-601.) How alkali, salt and dextrose may be used to relieve the alleged consequences of kidney disease (" uremia," vomiting of central origin, papillo-oedema, etc.) as well as other conditions in which an oedema of the affected organs constitutes a charac- teristic feature is emphasized. (See pages 590-602.) (Edema as an alleged consequence of sodium chlorid reten- tion is next discussed. Retention is not due to an inability of 30 (EDEMA AND NEPHRITIS the kidney to eliminate chlorid, but to a change in the (protein) colloids of the body, which, under the influence of an abnormal production and accumulation of acid, not only swell (become oedematous) but at the same time retain more chlorid. Experi- ments on gelatin and fibrin are introduced to support this contention. (See pages 602-607.) The generalized csdema so frequently observed as an accom- paniment of certain types of kidney disease needs to be treated on the same principles as the cedfema of the kidney itself. All salts are indicated because they decrease this generalized cedema as they do the swelling of the kidney itself, and sodium chlorid is no exception to the rule. (See pages 607-608.) The accumulations of fluid in the body cavities in states of cedema represent colloid solutions in which the water is largely bound to the colloid as hydration water. They are therefore absorbed only with difficulty, and this is why when suflSciently large in amount they need to be and can be gotten rid of only by tapping. How the administration of salt and alkali while reducing the cedema of the body tissues generally may increase the accumulations of fluid in the cavities is commented upon. (See pages 608-612.) An effort is made to explain how the salt restriction scheme of therapy practiced by many, leads to the good results reported. It is not the salt restriction, but the accompanying water re- striction that does the work. By sufficiently reducing the intake of water we succeed in losing more water (by evaporation, etc.) in the unit of time than is taken in, and so all the organs of the body dry out. This at times succeeds in breaking into the vicious circles established in many organs when once they begin to swell. The swelling compresses their blood supply and thus aggravates their already precarious state. Dehydration of the organ through water starvation may suffice to save it. It is pointed out, how- ever, that by obtaining such dehydration through administration of properly chosen salts with water instead of through water starvation, we gain the advantage for our patient of having water available to float off his poisonous products. (See pages 612- 614.) The meaning of the signs and symptoms displayed by patients, the victims of nephritis, and their prognostic value are taken up, and the correlation between urinary findings and the clinical THE ARGUMENT 31 manifestations originating in organs other than the kidney is made. (See pages 614-621.) The physiological principles to be borne in mind in any at- tempt to develop an efficiency test for an organ are considered. Because of the great reserve available in nearly all of them they continue to show a normal function as long as more than a physio- logically necessary minimum remains preserved. Three-quarters to seven-eighths of the normal functional capacity of an organ may be lost before the organism as a whole shows the effects of it or before it can be discovered by functional tests. Even then the test must be heightened to the point of straining what re- mains before the loss becomes apparent. These considerations hold for the kidney. The functional capacity of the kidney is best tested by its power to eliminate water. I have never seen a kidney that would secrete water which would not also secrete all dissolved substances. Tests dependent upon an elimination of dissolved substances are fraught with greater possibilities for error and are therefore less satisfactory. As long as one- quarter (more probably one-eighth) of the total kidney function is preserved all such tests yield normal figures. Animals in which three-quarters of the total kidney substance has been removed excrete water and all dissolved substances as do normal animals. When less than one-quarter (or more probably one-eighth) of the total kidney substance is functioning a diminished output of water and of certain dissolved substances becomes evident, but when such extreme states of renal insufficiency are experimentally produced or encountered clinically elaborate tests are not neces- sary to bring them to light. Efficiency tests are of greatest service for the discovery of unilateral kidney lesions. The teach- ing that successful prognostications regarding the onset of " ure- mia," etc., can be made on the basis of kidney efficiency tests, needs to be examined critically, for most of such alleged conse- quences of kidney disease are not consequences. (See pages 621- 629.) Since 'an abnormal production or accumulation of acid in the kidney is so largely responsible for the development of the signs and symptoms of nephritis, the importance of following the acidity of the urine is emphasized The meanings of titration acidity and of hydrogen ion acidity are defined. The deter- mination of either or of both furnishes valuable data in the 32 (EDEMA AND NEPHRITIS clinical control of every case of nephritis, but neither can alone serve as an index to the severity of the intoxication occurring in the kidney. How the physician may use simple indicators in the urine by way of determining roughly its hydrogen ion acidity and the meaning of such determinations is explained. (See pages 629-638.) It is characteristic of animals when subjected to intoxication with acid to draw first upon their supply of fixed bases to neutralize the acid. When these have been largely exhausted the carnivora have a second method of meeting the acid intoxication, namely, by the production of ammonia. An absolute or relative increase in the ammonia output in the urine (or other secretion from the body) therefore becomes evidence for and a quantitative guide to the degree of acid intoxication. The usefulness of ammonia determinations in nephritis is therefore emphasized. (See pages 638-644.) The section closes by pointing out how a diagnosis of nephri- tis has in itself but little meaning, in that after this has been made it is ever necessary to say why the nephritis has come to pass. As the conditions leading to or likely to lead to the signs and symptoms of nephritis are largely known, the great impor- tance of prophylactic measures is emphasized. (See pages 644-650.) PART SEVEN Glaucoma from a pathological standpoint represents one of the local oedemas, and from a clinical point of view all its signs and symptoms are referable to the increased intraocular pressure resulting from the csdema. An eye becomes glaucomatous not because water is forced into it, but because it suffers changes which make it suck up an increased amount. The eye is built up of a series of colloids which normally have a certain hydration capacity. Anything which in the body is capable of increasing this hydration capacity leads to a swelling of the eye and con- stitutes therefore a cause of glaucoma. As in other forms of oedema, evidences may be found in cases of glaucoma for an abnormal production or accumulation of acids in the eye and of substances which in their action upon colloid's behave like acids. (See pages 653-655.) THE ARGUMENT 33 As the abnormally high hydration of the ocular colloids which characterizes glaucoma may be reduced through various salts, so can clinical cases of glaucoma be given relief from symptoms by the subconjunctival injection of properly selected salts such as sodium citrate. (See pages 655-657.) The problem of glaucoma and its treatment is identical with the problem of nephritis, and exactly as we err in the treatment of nephritis when we consider only the kidney, so do we go wrong when in glaucoma we consider onily the eye. Starvation, an excessive protein diet, hard muscular and mental work, excessive consumption of sour wines, various intoxications (anesthetics, alcohol and arsenic), the infections, the severe anemias, arterio- sclerosis, uncompensated heart lesions, exposure to cold, etc., are all associated with an abnormal production of acid in the body, and constitute in consequence potent factors in the precipitation of the glaucomatous attack. These must be recognized and removed if we expect the attack to subside. If they cannot be removed, then we need to meet their consequences, and since we deal here with factors affecting the whole patient we must treat him. For this reason alkali, salts, dextrose and water are indicated in the same way and for the same reasons as in nephritis. (See pages 657-660.) The prognosis in glaucoma depends entirely upon the nature of the factors appearing in its etiology. A diagnosis of " glau- coma " is as complete as one of " dropsy." When the swelling is due to a transient intoxication, or to temporarily acting infections, mere reduction of tension by any means whatsoever, whether surgical or medical, can easily lead to brilliant results and permanent relief, but when irremovable causes such as blood vessel disease are responsible for the glaucoma — by far its commonest cause in older individuals — no scheme of treatment which ignores the blood vessel disease and which merely centers attention upon the eye can yield anything but temporary and ultimately disappointing results. (See pages 660-664.) The nature of the corneal opacities and the cloudiness of the clear media so often observed in glaucoma and certain other pathological states are discussed. They are not due to oedema, but represent precipitations of protein of the nature of casein. The " clouding " thus caused by the precipitation of one type of colloid while another is swelling, together make glaucoma 34 (EDEMA AND NEPHRITIS but another example of the widely observed " cloudy swelling " of the pathologists. (See pages 664-668.) A paragraph asking that those who feel tempted to make clinical use of any of the therapeutic methods discussed in the volume feel tempted also to study the scientific principles upon which they are based in order that misunderstanding and dis- appointment through improper application of the suggested remedial measures may be prevented, closes the volume. (See pages 668-669.) PART TWO ABSORPTION AND SECRETION IN INDIVIDUAL CELLS AND TISSUES PART TWO ABSORPTION AND SECRETION IN INDIVIDUAL CELLS AND TISSUES THE PROBLEM When, in the routine of our day's work, we are brought face to face with so practical a matter as the treatment of a patient suffering with glaucoma, or from the convulsions of uremia, or from an anuria, our behavior, if it is not a blind following of empirically transmitted teachings, is determined by what we think of the natm-e and the cause of these deviations from the normal. In good part we deal in these illustrations with an oedema of the affected organs — of the eye, of the brain, of the kidney — and so our treatment becomes but a specialized expression of what we think regarding the nature and cause of oedema in general. But this problem of oedema — the problem of the presence of abnormally large amounts of water in tissues and tissue spaces ■ — ^is in itself only a phase of a still greater problem: Why does protoplasm hold any water at all, and why does it hold under normal circtmistances so nearly constant an amount? It is easily seen why an interest in oedema or in such special expressions thereof as glaucoma, uremia or nephritis should have overshadowed the greater and really simpler problem, for all these have a human interest that is entirely lacking to the ques- tion of why protoplasm generally holds water. That attempts should in consequence have been made to answer the question of oedema first is not surprising. The ways and means adopted 37 38 (EDEMA AND NEPHRITIS may, however, well serve as an example of the short-cut methods which clinicians and pathologists have only too often adopted in order to obtain hght, and with disastrous results. Since oedema constitutes a pathological state of interest chiefly in man, various hypotheses were formulated to account for the condition on the basis of his complex anatomy — such, for instance, as his blood and lymph circulatory systems; and when experiments on the higher animals failed to bring the corroborating evidence which should convert the shadowy hypothesis into the healthy theory, recourse was had to the still more shadowy properties of " living " cells. To this day the accepted explanation of oedema remains an ill-defined mixture of the physical concepts of pressure and filtration with the mysterious forces of "living" matter. A little thought will show that variations in the amount of water held by cells and tissues occur in a great variety of animals and plants. To cite but a single example, and one common to both plants and animals, mention may be made of the long studied phenomena of normal, increased and decreased water content in cells which are discussed under the terms turgor plasmoptysis and plasmolysis. Isolated animal and plant cells have normally a certain water content. Under a variety of circumstances they may be made to absorb more. These are " cedemas " as true as any ever observed in man, or produced experimentally in dog or rabbit. Such reflection should by itself have created suspicion against any conception of oedema which demands for its production circulatory systems or any structures not common to all protoplasm, vegetable as well as animal. It will not seem strange therefore that the best contributions toward the solution of the problem of the ways and means by which cells and tissues absorb water have in recent years really come through the plant physiologists. Not led into erroneous paths through the presence of circulatory systems at all similar to those found in the higher animals, the plant physiologists early sought the explanation of the variations in the amount of water held by the plant tissues in the cells themselves. As we shall see shortly, this is where the problem belongs, and the attempts of later years to make differences in osmotic pressure responsible for the movement and storage of water in animal cells as well as in plant cells under normal and pathological con- ABSORPTION, SECRETION— CELLS AND TISSUES 39 ditions cannot be too highly commended. While the theory of osmotic pressure is incapable of accounting for more than a very small portion of the phenomena observed — even in plants — the great value of an attempt to explain variations in the water content of animal and plant tissues on a healthy physico-chemical basis cannot be questioned. To a consideration of the various hypotheses and theories that have been proposed to account for the normal water con- tent of cells and for the abnormally high water content which the plant' physiologists discuss under plasmoptysis, the path- ologists under cedema, the clinicians under a variety of terms, we shall have occasion to return later. Here we would only emphasize the fact that all these diversified phenomena bear a close relation to each other and need to be discussed together. The clinician meeting his medical problem has common interest with the pathologist; the pathologist in turn can discuss his more abstract notions of oedema only as he considers the physi- ologist and his ideas of normal water absorption. This will explain why we need in this volume to begin with a discussion of fundamental physiological principles. And at all times sub- sequently must we find it easy to pass from the extremes of pathology back to physiology, or vice versa. The unit for consideration in physiology, and so in pathology and applied medicine, remains the cell. As practical men we are likely to lose sight of this fact or to take exception to it, yet the most practical will heed the physiology of the individual cell most. Even the highly specialized functions of groups of cells or of organs in a complex individual rarely represent more than exaggerations of functions common to all cells. This is why this volume begins with and constantly reverts to the be- havior of the individual cell. To be familiar with the effect of various external conditions upon the general behavior of the individual cell is to be familiar with the behavior of these same conditions upon groups of cells. When such groups of cells (an organ) are part of and determine the behavior of yet other groups of cells in a complex organism, famiharity with the action of these external conditions upon the isolated groups of cells becomes synonymous with a knowledge of the action of these external conditions upon the organism as a whole. An understanding of the most specialized therapeutic procedure is almost invariably 40 (EDEMA AND NEPHEITIS dependent upon such a knowledge of the cell. We shall find many an illustration of this as we proceed. With this let us turn to our specific problem of water absorption by the living cell under physiological and pathological circum- stances. Since I consider the colloids of the tissues and their state as of chief importance in this matter, a review of the prop- erties of colloids with particular reference to their behavior toward water is first in order. II THE ABSORPTION OF WATER BY COLLOIDS 1. Remarks on Colloids: Nomenclature It is now more than fifty years since Thomas Graham recog- nized that chemical substances differ greatly in the rate with which they diffuse through solvents of various kinds. On the basis of this observation he made a distinction between those which diffuse but slowly and those which diffuse very rapidly. As the former are for the most part amorphous, and since ordi- nary glue (/coXXa) is an example of this class, he called them colloids. The group that diffuse readily he called crystalloids, for such beautifully crystalline substances as cane sugar, ordinary salt and urea are found in it. Since Ghaham's studies we have become familiar with further characteristics of colloids and crystalloids. Crystalloids are ordinarily stated to form true solutions. This, colloids do not — they form pseudo-solutions, that is to say, they simply remain suspended in the solvent. Colloid solutions are, therefore, not homogeneous, but heterogeneous in their make-up. Solutions of crystalloids show an osmotic pressure, which is proportional to the number of particles of dissolved substance in the unit volume of the solvent. The most typical colloids, on the other hand, show practically no osmotic pressure. The enormous differences in osmotic pressure between crystal- loids and colloids correspond to similar differences in the molec- ular weight of the substances composing the two groups. The molecular weight of the most pronouncedly colloid bodies may ABSORPTION, SECRETION— CELLS AND TISSUES 41 be measured in thousands, while two or three hundred covers the weight of even very complex organic compounds. It must be stated at once, however, that between the two ex- tremes of the typical colloids and the typical crystalloids, there is found an infinite number of substances which lean more or less strongly toward one side or the other. It is possible, for example, to obtain in crystalline form, certain albumins which may ordinarily be taken to represent our most typical colloids. Egg albumin may be obtained in such a state, and the physio- logical chemist is rarely satisfied with a hemoglobin that is not beautifully crystalline. On the other hand, comparatively simple bodies, such as silicic or tungstic acids, are found in the group of our most representative colloids. These few facts will suffice to show that no hard and fast line can be drawn between the colloids on the one hand and the crystalloids on the other. It should also be clearly understood that while we speak of colloids and crystalloids, and, therefore, are seemingly classifying substances, we ought really to speak only of the colloid and crystal- loid state. Our familiar use of the nouns colloid and crystalloid has grown out of the fact that certain chemical compounds are best known to us in the colloid state, while others we see almost always in a crystalloid state. As a matter of fact, it is probably safe to assume that any substance may be obtained in a colloid form, even those simplest and most typical crystalloids, the chlorids of the various metals. That many typical colloids may, on the other hand, be obtained in crystalline form is evidenced daily by the ever-growing list of biological products long known to us only in the form of amor- phous powders, mucilages and syrups which chemists are obtain- ing in crystalloid form. These considerations are not without biological significance, for a chemical substance in a colloid form may and usually does possess entirely different properties from the same chemical substance in a crystalloid form. This fact has been abundantly proved within the last decade in many striking ways. The question of how a crystalloid passes over into the colloid form, or vice versa — a question which has been but little investigated physico-chemically — is, therefore, of the greatest importance to biological chemistry, for these very con- versions of colloids into crystalloids, and of crystalloids into colloids, are among the commonest observed in the living organ- 42 CEDEMA AND NEPHRITIS ism. The question of glycogen formation from dextrose, for example, represents such a change. Coupled with the chemical change of a dehydration, there is in this case a physical change which converts a freely soluble, osmotically active, diffusible crystalloid, into an insoluble (pseudo-soluble), osmotically in- active, non-diffusible colloid. Marked as are these general differences between colloids and crystalloids (the colloid and the crystalloid states), the colloids themselves do not all possess the same properties. Various attempts have been made to classify them. As this book deals particularly with the relationship existing between colloids and the water they hold, the distinction of A. A. Notes ' between those colloids which are viscous, gelatinizing and not readily coagulated by salts (colloid solutions) and those which are non- viscous, non-gelatinizing, and readily coagulated by salts (colloid suspensions) is of great interest. To the former belongs, for example, a solution of gelatin, glue or dextrin, while in the latter might be mentioned the colloid solutions of ferric hydroxid, of aluminium hydroxid and of various metallic sulphids. The essential difference between the two groups resides in the relation which the colloid bears to its solvent. A colloid of the first-named type is united to much of the solvent, one of the latter holds but little or none. For this reason the former are also known as lyophilic, or, if water is the solvent, hydrophilic colloids; the latter as lyophobic or hydrophobic colloids. (J. Pek- KiN^ and H. Freundlich ^.) Wolfgang Ostwald,* who has done so much for the proper definition and classification of the colloids, distinguishes between the emulsion colloids (emulsoids) and the suspension colloids {sus- pensoids). The first are formed through mixture of two liquid phases, the second through mixture of a soUd with a liquid phase. Since it is harder to separate one liquid from another than a solid from a liquid, a separation of the two phases is obtained in emulsion colloids (which correspond, it will be seen, with ' A. A. NoYEs: Jour. Am. Chem. Soc, 27, 85 (1905). 2 J. Perkin: Journal de Chimie Physique, 3, 84 (1905). ' H. Fbeundlich: KoUoid Zeitsohr., 3, 80 (1908). Kapillarchemie, 309, Leipzig (1909). * Wolfgang Ostwald: Kolloid Zeitschr., 1, 291 and 331 (1907). Grund- riss der KoUoidchemie, Dritte Auflage, Dresden (1913). ABSORPTION, SECRETION— CELLS AND TISSUES 43 NoYEs' first group and Perrin's hydrophilic colloids) with greater difficulty than a similar separation in the suspension colloids of OsTWALD (which bring to mind Notes' second group and the hydrophobic colloids). When we recall that the hydrophilic colloids which have thus far been accorded most study — gelatin, dextrin, starch, glue, vegetable fibers, albumin, gums — are for the most part derived from biological sources, their probable importance in the living animal or plant must at once be suspected. Not only is the chief mass of the living organism built up of colloid material, but most of it belongs in the hydrophilic group. We will not be surprised, in consequence, to find that the same physico-chemical characteristic which makes for the division of all colloids into two great classes will show itself of importance in determining the behavior of the tissues toward water. It will give us a better conception of just what this absorption of water by colloids represents, and how it is influenced through various external conditions, if we study the swelling of some proteins. 2. Observations on the Swelling of Fibrin In these experiments ordinary blood fibrin was used, which after having been thoroughly washed to free it from adhering salts was dried at a low temperature and pulverized in a mortar. When weighed amounts of such powdered fibrin (0.25 gram) are introduced into definite volumes (25 cc.) of various solutions contained in test-tubes of the same diameter (1.7 cm.) the fibrin swells to very different heights. From the results of many series of experiments, the following facts which are of importance in our discussion have been determined.^ (a) Fibrin swells more in the solution of any acid than it does in distilled water. Table I illustrates this fact. While the exact order changes somewhat at different concentrations the table also serves to indicate that when equinormal acids 1 See MartiW H. Fischer and Gertrude Moore: Am. Jour. Physiology, 20, 313 (1907); Kolloid Zeitschr., 5, 197 (1909); Martin H. Fischer, Pfluger's Archiv., 125, 99 (1908). 44 (EDEMA AND NEPHRITIS are compared, they are found to be very unequally effective in producing the swelling. A hasty glance suffices to show that we are not dealing with the simple effects of hydrogen ions deter- mined by the relative degrees of dissociation of the various acids, for while a " strong " acid (hydrochloric) stands at the top of the list, another (sulphuric) stands at the very bottom, while a series of " weak " organic acids are found between. TABLE I Fibrin — Acid All acids n/10 Height of fibrin column in mm. after 24 hours. All acida n/10 Height of fibrin column in mm. after 24 hours. Water 6 28 27 27 24 Oxalic Nitric 24 Hydroohlorio Phosphoric Lactic 20 Acetic 10 Citric Sulphuric 9 Pormic 8 The amount that fibrin swells in any acid solution is dependent upon the concentration of the acid. Within certain limits fibrin swells the more the higher the concentration of the acid. In the case of the " strong " acids, however, a maximum is attained, above which a further increase in the concentration of the acid does not lead to a greater, but to a diminished absorption of water. The swelling of fibrin in acid solutions of progressively higher concentrations may therefore be represented graphically by a curve which rises at first, attains a maximum and then falls again. These facts are brought out in Table II. In the case of acetic acid it will be noted that the highest concentration of acid used in the table induces the greatest amount of swelling. At concentrations above n/10, 1 obtained a height up to 41 mm. with this acid. As yet I have not, however, been able to deter- mine if with such a " weak " acid a point is finally reached be- yond which, as with the " strong " acids, a further increase in concentration brings about a diminished absorption. No greater swelling of fibrin than that noted in the table can be obtained by using concentrations of sulphuric acid above those given in Table II. ABSORPTION, SECRETION— CELLS AND TISSUES 45 TABLE II Fibrin — Add Concentration of acid. Height of fibrin column in mm. after 24 hours in Hydro- chloric acid. Nitric acid. Acetic acid. Sul- phuric acid. 1 ■ 2 3 4 5 6 7 8 9 10 12H 15 17K2 20 25 25 n/10 acid +24 c -f23 -1-22 +21 +20 +19 +18 +17 +16 +15 +12H +10 + 7Ji + 5 . H2O. 12 23 37 47 4S 41 '31 21 13. S 26. 29. 37.5 35. 30. 30. 25. 23. 21.5 18.5 17. 14.5 14. 11.5 8.5 10. 10.5 11. 12. 12. 13. 13. 14. 14.5 15. 16. 17. 18. 18.5 9. 9.5 10. 10. 11. 11. 10. 10. 10. 10. 9. 9. 8.5 water (control) . (6) Fibrin swells more in the solution of any alkali than in pure water, but the amount of this swelling is greater in some alkalies than in others. This statement is the analogue of the corresponding one for acids. When equinormal solutions are compared fibrin swells more in potassium hydroxid than in sodium hydroxid, and more in either of these than in calcium hydroxid or ammonium hydroxid in the order named. The first three of these have in such dilute solutions about the same degree of dissociation. Clearly the amount of swelling is not simply a function of the hydroxyl ions. As in the case of acids, the amount of swelling is here also dependent upon the concentration of the alkali. For the " strong " alkalies there is within certain limits an increase in the amount of swelling with every increase in the concentration of the alkali, but, after a certain point is exceeded a further increase in concentration is followed by a diminution in the height of the fibrin column. Tables III and IV illustrate these facts. If the amounts that fibrin will swell in acid and alkali solu- tions having the same H or OH concentration are compared, it is found that fibrin swells less in the solution of an acid than in an equally concentrated solution of an alkali. While, for 46 CEDEMA AND NEPHRITIS example, in n./50 KOH or NaOH, the fibrin column may be found to measure 83 and 77 mm. respectively, in n/50 HCl or HNO3 it measures only 48 and 35 mm. TABLE III Fibrin — Alkali Concentration of alkali. Height of fibrin column in mm. after 24 hours. KOH NaOH NHiOH 2 4 6 10 15 25 n/10 alkali +23cc. H2O. +21 ■■ . + 19 •■ + 15 ■• . +10 " . 25 cc. water (control) . 23 64 83 80 72 58 22 68 77 75 62 57 10 10.5 11 11 12 13 8 TABLE IV Fibrin — Alkali Concentration of alkali. Height of fibrin column in mm. after 24 hours. 10 12J^ 15 17H 20 25 25 cc cc. n/IO alkali +24 +23 +22 +21 +20 + 19 +18 +17 + 16 +15 +12H + 10 + TA + 5 , HjO. water (control) . » Actually these solutions were prepared by diluting n/30 Ca(0H)2 . (c) We come now to the interesting fact that the addition of any salt to the solution of an acid or an alkali decreases the amount that fibrin will swell in that solution. The only excep- tions to this rule are formed by the salts which react with the acids. If barium chlorid, for example, is added to a sulphuric acid solution, the amount of swelling is not decreased, but in- ABSORPTION, SECRETION— CELLS AND TISSUES 47 creased. This is because insoluble barium sulphate is produced and thrown down, while hydrochloric acid is formed in which fibrin swells more than in an equally concentrated sulphuric acid solution. The higher the concentration of the added salt, the less does the fibrin swell, and if enough is added the effect of the acid or alkali may be suppressed almost entirely. These facts are illustrated in Tables V and VI and in Fig. 1. The tube on the Figure 1. extreme right contains the unit weight of powdered fibrin in water. The tube marked HCl contains the same weight of fibrin in n/40 acid. The remaining tubes from right to left contain the same amounts of acid and of fibrin, but progressively greater concentrations of sodium nitrate (from m/40 to m/5 in the finished solution). (d) If the effect of equimolar ^ salt solutions is compared, ' To make proper comparisons between the physiological or pharmaco- logical actions of different chemical compounds, ordinary equivalents by weight (as in percentage solutions) cannot be used. We must compare 48 CEDEMA AND NEPHRITIS TABLE V Fibrin — Acid+Salt Concentration of solution. Height of fibrin column in mm. after 24 hours. KCl MgClj (NH4)2SO< KI 15 cc. n/5 HCl +15 cc. m/2 salt solution 15 " " +15 " m/4 " 9 10 13 14 15 6 8 9 10 10 5 6 7 10(?) 9 5 5 15 " " +15 " m/S " 6 15 " " +15 " m/16 " 10 (?) 9 15 " " +15 cc. water (control) 17 TABLE VI Fibrin — Alkali+Salt Concentration of solution. Height of fibrin column in mm. after 24 hours. 5 cc. n/lONaOH+12 cc. m/1 NaCl+ 8 cc. H2O. +10 + 8 " + 6 " + 4 " + 3 " + 2 •' + 1 " +20 cc. water. +10 +12 +14 + 16 +17 +18 +19 19 21 21 24 28 30 32 43 74 25 cc. water (control) . amounts that are equivalent from certain chemical or physico-chemical points of view. For many purposes molar {pram-molecular or molecular) solutions serve very well. A molar solution (m/1) is made by dissolving the molecular weight of the substance (including its water of crystallization, if it has any) expressed in grams in enough water to make a liter. If only one-half the gram-molecular weight is dissolved in enough water to make a liter, we have a one-half molar solution (m/2), etc. Solutions which con- tain the same number or fractions of a gram-molecule in the unit volume are equimolar. In the case of acids and alkalies it is usually best to employ normal solutions. A normal solution (n/1) is a molar one provided the dissolved substance is monobasic. In other words, its power to displace hydrogen is taken into consideration. A normal solution of a dibasic compound has half the concentration of a molar solution of the same compound; a normal solution of a tribasic compound but one-third, etc. Equinormal solutions of different acids or alkalies therefore all contain the same amount of replace- able hydrogen or hydroxyl. The " physiological " or " normal " salt solutions of our laboratories and hospitals have absolutely nothing to do with the normal solutions of the chemists which we are discussing. The terms are meaningless, and should disappear. We should speak of 0.85 per cent or 0.9 per cent sodium chlorid solutions if that is what we mean by these terms. ABSORPTION, SECRETION— CELLS AND TISSUES 49 they are found to affect the sweUing of fibrin in solutions of acids or alkalies to very unequal degree. This is readily apparent from Fig. 2, where the effect of adding molecularly equivalent amounts of various sodium salts to a hydrochloric acid solution is portrayed. The tube on the extreme left contains pure water only. The next contains pure hydrochloric acid (n/40). From left to right the succeeding tubes contain the same amount of hydrochloric acid plus various sodium salts (n/40 HCl in m/40 Figure 2. salt solution). The salts added from left to right are respectively the chlorid, bromid, nitrate, iodid, acetate, tartrate (of sodium and potassium), sulphate, phosphate and citrate of sodium. The tube on the extreme left in Fig. 3 contains a pure n/40 solution of sodium hydroxid. The remaining tubes show the effect of adding molecularly equivalent (m/40) amounts of various sodium salts to the pure sodium hydroxid solution. From left to right the salts added are the bromid, nitrate, acetate, tartrate, sulphate, citrate and phosphate of sodium. 50 (EDEMA AND NEPHRITIS ABSOKPTION, SECRETION— CELLS AND TISSUES 51 Prom the study of many series of salts it has been found that the effect of any salt is made up of the sum of the effects of its constituent radicals. In any series of salts having a common base the order in which the acid radicals are effective is always found to be the same, and when series having a common acid are compared, the order in which the basic radicals are effective is always the same. From such experiments the two following lists have been constructed. The radical least effective in bring- ing about a diminution in the amount that fibrin will swell in the solution of any acid or alkah is in each case placed first: Acid radicals. Basic radicals. Chlorid Potassium Bromid Sodium Nitrate Ammonium Sulphocyaijate Magnesium • lodid Calcium Acetate Barium Sulphate Strontium Phosphate Copper (ic) Tartrate Citrate Iron (ic) The table for the acid radicals is more accurate than the table for the basic radicals. This is because the amount of difference in swelhng produced by the end members of each of the two series is decidedly greater in the case of the acid radicals than in the case of the basic radicals. The general grouping of the basic radicals is, however, entirely trustworthy. While the differ- ence between the amount of swelhng in an acid solution con- taining a magnesium salt may not differ decidedly from a similar solution made up with a calcium or barium salt, there is never any question about the difference between the action of any of these three and that of a radical found in the list either above or below them. (e) Non-electrolytes do not share with electrolytes their marked power of reducing through their presence the amount that fibrin will swell in the solution of any alkali or acid. In concentrations that are from an osmotic standpoint comparable to those used above in the case of salts, the non-electrolytes are almost without effect, as shown in Tables VII and VIII : 52 (EDEMA AND NEPHRITIS TABLE VII Fibrin — Acid+Non-eledrolytes Concentration of the solution. Height of fibrin column in mm. after 24 hours. 10 CO. n/5 HCl+lO cc. H2O 10 " " +10 ' ' m/1 glycerin. . 28 30 10 " " +10 " " saccharose. ... 28 10 " " +10 " " urea 27.5 10 " " +10 " " dextrose 27 8 TABLE VIII FiBRiK — A lkali+ Non-electrolytes Concentration of the solution. Height of fibrin column in mm. after 24 hours 10 cc 10 10 10 10 10 10 10 10 10 10 10 2S cc. n/10 KOH+ 4 cc. m/1 + 4 + 4 + 4 + 4 +15 +10 +10 +10 +10 +15 +15 r (control) . H2O. 2/m m/1 HjO. ethyl alcohol +11 cc. methyl alcohol +11 glycerin +11 urea +11 saccharose +11 HjO . ethyl alcohol + 5 methyl alcohol + 5 glycerin + 5 urea + 5 saccharose 76 77 72 83 75 77 73 73 64 80 75 78 Not until employed in rather concentrated solutions do glycerin, saccharose, dextrose, ethyl alcohol and methyl alcohol change in any marked way the height to which a fibrin column will swell in various concentrations of acid or alkali, h (/) For purposes of biological apphcation a series of tables are inserted here which show the effect of different non-electrolytes upon the absorption of water in a neutral medium. Saccharose, dextrose, levulose, methyl alcohol, propyl alcohol and acetone will inhibit more or less markedly the amount of water that fibrin will absorb. 1 Of these, the behavior of saccharose, dextrose and levulose deserve special mention. While all three dehydrate fibrin in increasing amount with increase in concentration, at the same concentration saccharose is the most powerful in ' Martin H. Fischer and Anne Sykes: Kolloid Zeitschr., 14, 215 (1914). ABSORPTION, SECRETION— CELLS AND TISSUES 53 this regard. Tables IX, X and XI suffice to illustrate this fact, which will be more emphatically brought out in discussing the absorption of water by gelatin. TABLE IX Fibrin — Saccharose Concentration of solution. Height of fibrin column in mm. 40 1 2 3 S 7^ 10 20 cc. water (control) cc. 2/m saccharose +39 cc. H2O. +38 +37 +35 +321^ +30 +20 23 22 22 21 21 21 20 18 TABLE X FiBEiN — Dextrose Concentration of solution. Height of fibrin column in mm. 40 cc. water (control) 1 cc. 2/m dextrose +39 cc. H2O. 2 3 5 10 20 40 +38 +37 +35 +32J^ +30 +20 21 21 20 20 19 19 18 17 15 TABLE XI Fibrin — Levulose Concentration of solution. Height of fibrin column in mm. 40 cc. water. . 1 cc. 2/m levulose +39 00. H2O 2 " +38 3 " +37 5 " +35 yVi ■■ +323^ ■■ 10 " +30 20 " +20 21 20.5 20.5 20.5 20 20 20 19 {g) The taking up and giving off (absorption and secretion) of water by fibrin represents in high degree a reversible process. 54 OEDEMA AND NEPHRITIS If for a hydrochloric acid solution in which fibrin has attained its maximal swelling, an equally concentrated sulphuric acid solution is substituted, the fibrin column shrinks. The same occurs if a potassium hydroxid solution is replaced by an equally concentrated calcium or ammonium hydroxid solution. When equilibrium is finally established the height of the fibrin column in each of these solutions is approximately equal to that which would have been attained had the fibrin been placed directly in these solutions. In the same way fibrin which has attained its maximal swelling in an acid solution will shrink rapidly if for the pure acid there is substituted one of equal concentration containing a salt. Similarly, if water replaces the solution of an acid or an alkali, the fibrin will either shrink or swell more, depending upon whether the addition of the water makes the concentration of the alkali move toward or away from that which is optimal for the swelling of fibrin. (See paragraph b of this section.) The reverse of all these experiments can also be accomplished, although not with the same ease. If, for example, hydrochloric acid is substituted for sulphuric, or potassium hydroxid for the calcium compound, an increase in the amount of swelling is noted, but the column does not rise as high as it would have done if placed directly in these solutions. Similarly, fibrin which has once been in an acid or an alkali solution containing a salt, when placed in pure solutions of acid or alkali does not swell to the amount which it would have done if it had been put in these solutions from the first. All this would seem to indicate that fibrin suffers more or less permanently from every external con- dition to which it has been subjected. To explain this phe- nomenon, which is of great importance from both the theoretical and the practical aspects of biology and medicine, we can advan- tageously call to mind the well-known property of colloids of attaching to themselves, and holding fast the various substances with which they come in contact.^ (h) For reasons associated with our analysis of the problem of oedema we are particularly interested in substances which are capable of increasing the amount of water held by such a colloid as fibrin. Among other substances besides acids and alkalies 1 See page 168, where is discussed the taking up of dissolved substances and the phenomena of adsorption. ABSORPTION, SECRETION— CELLS AND TISSUES 55 capable of thus increasing the hydration capacity may be men- tioned urea and pyridin.i The hydrating effect of urea is already indicated in Tables VII and VIII, but is more clearly evidenced in Table XII. The hydrating effect of pyridin is illustrated in Table XIII. The calibrated test-tubes used in these particular experiments were 22 mm. in diameter, 40 cc. of solution were prepared and a gram of dry fibrin was employed. TABLE XII Fibrin — Urea Concentration of solution. Height of inbrin column in mm. after 24 iiours. 40 1 2 3 5 10 20 40 cc. water (control) cc. 5/m urea +39 cc. H2O. +38 " +37 •■ . +35 " +32^ ■• +30 " +20 " 17 18 19 19 19 20 22 25 39 TABLE XIII Fibrin — Pyridin Concentration of solution. Height of fibrin column in mm. after 24 hours. 40 1 2 3 5 71^ 10 cc. water (control) cc. 10/m pyridin +30 cc. HzO +38 " ., +37 " .. +35 " .. +32M " . 20 22 23 24 25 26 28 (x) An interesting and biologically important difference exists between the increased hydration brought about by substances of the type of urea or pyridin and that brought about through acids. That produced through acids is readily reducible through all salts. Salts do not reduce the increased hydration brought about either through lirea or pyridin, as shown by Tables XIV 'See Martin H. Fischer and Anne Sykes: Science, 38, 486 (1913). Some of the amins seem also to belong in this group, but they are so pro- nouncedly alkaline in watery solution that a large part of their hydrating effect is dependent upon this alone. 56 (EDEMA AND NEPHRITIS and XV. The slight power of some salts to increase the hydration of fibrin is merely found added to that produced by urea or pyridin alone. On the other hand, various non- electrolytes, such as the sugars, which affect the swelling of fibrin in acid solutions but little, produce a marked shrinkage when the increased hydration has been produced by urea or pyridin. This is shown in Tables XVI and XVII. TABLE XIV Fibrin— Urea + NaCl Concentration of solution. Height of fibrin column in mm. after 24 Ijours. 40 20 20 20 20 20 20 20 cc. wat r (control) cc. 5/m urea +20 +20 cc. + 15 " +10 ' + 5 ■■ + 2Vi" + 1 " HzD. m/1 NaCl, + 5 cc. H2O. + 10 " +15 " .. + 17^2 ■■ .. +19 " .. 18 30 36 35 as 33 33 32 TABLE XV Fibrin — Pyridin-\-'ii&C\ Concentration of solution. Height of fibrin column in mm. after 24 hours. 40 cc. water (control) 10 cc. 10/m pyridin +30 cc. HzO 10 " " +20 cc. m/1 NaCl +10 cc. H20 10 " " +10 " ■' +20 ■■ 10 " " +5 " " +25 10 •• " + 2H " " +27H '■ 20 29 32 32 31 30 TABLE XVI Fibrin — Urea +Saccharose Concentration of solution. 40 cc. water (control) . 20 cc. 6/m urea +20 20 20 20 20 20 20 20 +20 cc + 15 ■ + 10 • + 5 ' + 3 ' + 2 ' + 1 ' H20 2/m saccharose . + 6 cc. H2O. + 10 " +15 ■• . + 17 • + 18 " . +19 " . Height of fibrin column in mm. after 24 hours. 19 29 20 20 20 20 20 20 21 ABSORPTION, SECRETION— CELLS AND TISSUES 57 TABLE XVII Fibrin — Pyridin +Saccharose Concenta*ation of solution. Height of fibrin column in mm. after 24 hours. 40 cc. water (control) 10 cc. IC/m pyridin +20 cc. HaO 10 " " +30 cc. 2/m saccharose +10 cc. H-iO. " " +10 " " +20 " 10 " " +5 " " +25 " 10 •■ ■• + 2ii " " +27H '• 20 29 25 26 27 28 3. Observations on the Swelling of Gelatin We have now to consicler whether the behavior of fibrin in various solutions is characteristic of this substance alone, or whether we have simply discussed as applicable to one colloid, properties that are really common to many. A partial answer to this question can be found in the careful studies available on the swelling of gelatin and other proteins. The observations of Franz Hofmeister,^ Wolfgang Padli,^ K. Spiro ^ and Wolf- gang OsTWALD * show gelatin to behave in many ways simi- larly to fibrin. We will review some of these in so far as they are of interest to us in the study of our problem. At the same time experiments of our own will be introduced which not only serve to corroborate the various findings already made on the swelling of gelatin but augment these, particularly in the following directions. They show (1) the unequal effect of different equi- normal and equally dissociated acids and alkalies upon the swell- ing; (2) the antagonism between neutral salts and acids or alkalies upon it; (3) the comparative lack of antagonism between non- electrolytes and acids or alkalies upon the absorption of water by this substance; (4) the reversibihty of the absorption of water by this substance. They discuss also (5) other substances besides acids which are capable of increasing the hydration capacity of gelatin and show (6) how such hydration is not reduced through salts, but readily through various non-electrolytes which are ' Feanz Hofmbister: Archiv. f. exp. Path. u. Pharm., 27, 395 (1890). ' Wolfgang Pauli: Pfluger's Archiv, 67, 219 (1897); 71, 1 (1898). ' K. Spiro: Hofmeister's Beitrage zur chem. Physiologie, 5, 276 (1904). •1 Wolfgang Ostwald: Pfluger's Archiv, 108, 563 (1905). 58 CEDEMA AND NEPHRITIS comparatively ineffective in reducing the swelling induced through acid. Our experimental methods differed in no material way from those usually followed by workers in this field. Ostwald's scheme was adopted. One part of the best commercial gelatin was dissolved at a low temperature (45° C.) in four parts of water and poured into shallow pans. After having hardened in an ice- chest the gelatin was cut with the aid of a sharp knife and a ruler into squares of uniform size. These squares were allowed to dry upon glass plates at room temperature. The drying process took from six to ten days, and was not sufficiently rapid to distort the squares. When completely dry the squares measured about 18X18X2.5 mm. and weighed approximately 0.8 gram. As a uniform material is necessary to obtain comparable results, it is well to mention that all the gelatin discs used in any extended series of experiments were always prepared at the same time. The course of the absorption of water by the discs was followed by immersing the weighed gelatin discs in solutions of various kinds and weighing them at intervals. In order to facihtate comparison with the results obtained on fibrin the paragraphs on gelatin are lettered in the same way as the paragraphs on fibrin. It will be seen that gelatin is a colloid which behaves in many ways like .fibrin. Important differences, however, exist between the two, which we shall later find to be not without biological interest. (a) Gelatin swells more in the solution of any acid than it does in water. This fact is readily apparent even to the naked eye. If two gelatin discs are dropped at the same time, the one into water, the ether into n/20 hydrochloric acid, the in- equality in the amount of swelling is plainly to be seen at the end of six hours, and at the end of twenty-four or forty-eight it is very marked. While at this time the gelatin disc in the water still has a shghtly brownish-yellow and opaque appearance, that in the acid is hyahn and perfectly clear, so clear, in fact, that it can scarcely be seen at the bottom of the dish. Spiro, who first discovered this difference in the amount that gelatin will swell in water and in acids, found that while a gelatin plate gained 1.97 times its weight in water, it gained 3.49 times its weight in n/500 hydrochloric acid, and 5.45 times its weight in n/200 acid. OsTWALD came to the same conclusion from comparison ABSORPTION, SECRETION— CELLS AND TISSUES 59 of his results on the swelling of gelatin plates in acids of various kinds with the absorption curves of gelatin in water, as given by HOFMEISTER. While the gelatin swells more in the solution of any acid than in water, the acids are by no means equally potent in this regard when equinormal solutions are compared. Most authors are inclined to the belief that the swelUng induced in gelatin discs is exclusively a function of the hydrogen ion concentration. It seems to me that this is only in part responsible for the observed effects. I have taken the liberty of constructing from Ostwald's ^ tables the curves contained in Figs. 4 and 5. 72 Figure 4. The hours that the gelatin discs were in the acid solutions are plotted on the horizontal, the amount of water absorbed, expressed in units of the original weight of the disc, is shown on the vertical. We have no difficulty in recognizing in Fig. 4 the order: Nitric, Acetic, Sulphuric, Boric. The position of the " weak " acetic acid between the " strong " nitric and sulphuric acids (which two are about equally dis- sociated, and yield a higher concentration of hydrogen ions than the equinormal acetic acid) is by itself an argument against the explanation which considers only the concentration of the hydrogen 'Wolfgang Ostwald: Pfluger's Arcliiv, 108, 577 and 578 (1905). 60 (EDEMA AND NEPHRITIS ions. A look at Fig. 5 brings with it similar conclusions. Except in the first hours of the experiment, we again find the order: Hydrochloric, Nitric, Acetic, Sulphuric, Boric. This order in which the different acids make gelatin swell is identical with that in which they make fibrin swell. The amount that gelatin swells in any acid solution is de- pendent in a complex way upon the concentration of the acid. This is shown in Fig. 6, which has been copied from Ostwald's article. The curve marked HCl indicates the amount of water absorbed by gelatin plates after twenty-four hours residence in hydrochloric acid solutions of various concentrations. With the exception of the initial fall in the curve (which simply indicates that in hydrochloric acid solutions of certain concentrations a gelatin disc may absorb even less than in pure water) we notice a rapid rise in the curve indicative of an increase in the amount of swelling with every increase in the concentration of the acid. An optimal point is reached when the concentration of (approxi- mately) n/38 hydrochloric acid is attained, beyond which a ABSORPTION, SECRETION— CELLS AND TISSUES 61 further increase in the concentration of the acid is not followed by a greater absorption of water, but by a less. An analogous relationship between concentration of acid and amount of swelling exists in the case of fibrin. (6) Gelatin swells more in the solution of any alkali than in water. Macroscopic examination alone evidences this fact. Spiro,! who first noted it, found that while a gelatin disc kept in pure water gained only 3.02 times its weight of water, one kept •K. Spiro: Hofmeister's Beitrage zur ehem. Physiologie, 5, 277 (1904). 62 CEDEMA AND NEPHRITIS in n/100 sodium hydroxid solution gained 5.08 times its weight, one in n/50 solution, 11.82 times its weight, and one in n/10 solution, 12.61 times its weight of water. When the effect of equinormal solutions of different alkalies is compared, it is found that a gelatin disc swells more in some alkalies than in others. This statement, which has its analogue in the acids, is illustrated in Fig. 7. The hydroxids show the following grouping, in which that which allows of the greatest swelling is placed first: Potassium, Sodium, Calcium, Ammonium. At the concentrations employed, the electrolytic dissociation of the first three is about the same. The conclusion, therefore, seems justified that the swelling of gelatin in various alkalies FiGTJRE 7. is not solely determined by the concentration of the hydroxyl ions, but perhaps by these minus the effect of the kation, calcium being more active in bringing about a reduction in swelling than sodium, and this more than potassium. Fig. 7 has been con- structed from the data contained in Table XVIII. As the increase in weight in these experiments on gelatin is very consid- erable, uselessly large figures have been avoided by expressing changes in weight in -parts of the original weight of the (dry) gelatin. One part, therefore, corresponds to an increase in weight of 100 per cent. ABSORPTION, SECRETION— CELLS AND TISSUES 63 TABLE XVIII Gelatin — Alkali Dry weight of gela- tin disc. 0.830 0.830 0.830 0.822 Solution. 150 cc. n/30 KOH. 150 cc. n/30 NaOH. 150 cc. n/30 Ca(0H)2. 150 cc. n/30 NH.OH. Hours in the solution. Gain in parts of one part of gelatin. 10.05 21.25 34.25 58.20 5.5 8.9 12.8 18.6 5.3 8.7 12.5 17.7 3.4 6.2 7.4 11.9 3.4 5.0 6,8 8.8 As in the case of hydrochloric acid, we find with potassium hydroxid also that the amount of swelling is dependent in a com- plex way upon the concentration of the alkali. This is well shown in the curve marked KOH in Fig. 6, copied from Ostwald. It indicates the amount of water absorbed by gelatin discs after twenty-four hours residence in various concentrations of potas- sium hydroxid. The initial rise in the curve indicates how with an increase in the concentration of the alkali there is an increase in the amount of swelling; but, as with the acid, an optimal point is soon reached beyond which a further increase in the concen- tration of the potassium hydroxid leads to a diminished absorption of water. If the amounts that gelatin will swell in equinormal solu- tions of acids and alkahes are compared, it is found that gelatin swells somewhat less in the solution of an alkaU than in an equally concentrated acid. This fact, which is the reverse of that found for fibrin, is well illustrated in Fig. 6 and in the upper two curves of Fig. 5, copied from Ostwald's studies. It finds a ready ex- planation, it seems to me, in the experiments which follow. Commercial gelatin is distinctly acid. When placed in the solu- tion of an alkaU, a salt is therefore formed, the presence of which, as the next paragraph shows, markedly decreases the amount that gelatin will swell in any acid or alkaline hquid. (c) The addition of any salt to the solution of an acid or an alkah decreases the amount that a gelatin disc will swell in that solution. As the number of insoluble hydroxids is large, studies on the antagonism between acids or alkahes and salts were carried out chiefly with acid solutions. Fig. 8, as well as Figs. 9, 10, 11, 64 CEDEMA AND NEPHRITIS 12, 13, 14 and 15, illustrates this point. In Fig. 8 is compared the swelling of a gelatin disc in a pure hydrochloric acid solution, with the sweUing of gelatin discs placed in equally concentrated hydrochloric acid solutions to which have been added equimolar amounts of various ammonium salts. As clearly evident, the amount of swelling is in every instance much less in these solutions Figure 8. than in the pure hydrochloric acid. Fig. 8 has been constructed from the data contained in Table XIX. TABLE XIX Gelatin — Add+Salt Dry weight of gelatin disc. 0.802 0.806 0.813 _ 0.814 0.817 0.81 Solution. 50 cc. n/10 HCl +50 cc. H2O. 50 cc. n/10 HCl +50 cc. m/2 ammonium acetate. 50 cc. n/10 HCl +50 cc. m/2 ammonium bromid. 50 cc. n/10 HCl +50 cc. m/2 ammonium chlorid. 50 cc. n/10 HCl +50 cc. m/2 ammonium nitrate. 50 cc. n/10 HCl +50 cc. m/2 ammpnium sulphate. Hours in the solution. Gain in parts, of one part of gelatin. 10.25 21 25 34.25 58.20 7.6 11.4 15.3 21.0 2.8 3.9 4.8 5.8 3.3 5.0 6.7 9.5 3.4 5.2 6.9 9.3 3. ' 4.9 6.6 9.5 2.9 4.2 5.3 6.9 ABSORPTION, SECRETION— CELLS AND TISSUES 66 The higher the concentration of the added salt, the less does the gelatin swell, and if enough is added the effect of the acid or alkali may be almost entirely suppressed. This fact is brought ■ BS P 2 out in Figs. 9, 10, 11, 12, 13 and 14. In each of these figures the curve for the swelling of the gelatin disc is found to he nearer the base line with every increase in the concentration of salt employed. 66 (EDEMA AND NEPHRITIS (d) When the action of equimolar salt solutions on the swell- ing of gelatin discs in acid or alkaline solutions is compared, it is found that some salts depress the amount of swelling more P o M than others. This is already apparent in Fig. 8, in which the sulphate and acetate of ammonium have brought about a dis- tinctly greater inhibition in swelling than the chlorid, bromid ABSORPTION, SECRETION— CELLS AND TISSUES 67 and nitrate. The point is further illustrated by comparing with each other Figs. 9, 10 and 11; also Figs. 12, 13 and 14. The hydrochloric acid curves of Figs. 9, 10 and 11 are practically identical. In all the figures a diminution in the amount of swell- ing is apparent through the addition of the salts, and the more salt added, the greater is this diminution. When Fig. 9 is com- 68 OEDEMA AND NEPHRITIS pared with Fig. 10, it is readily apparent that at the same con- centration potassium citrate brings about a greater depression of swelling in an acid solution than potassium chlorid. When, now, we compare Fig. 11 with Fig. 9 we note that potassium sul- phocyanate acts more powerfully than potassium chlorid. When we compare Fig. 10 with Fig. 11 we find that the extremes of the potassium citrate series lie be- tween the extremes of the potassium sulphocyanate se- ries. We cannot, in consequence, give an exact table indi- cative of the order in which the vari- ous acid radicals of salts with a com- mon base are active in depressing the amount that gela- tin will swell in an acid solution with- out stating the ex- act concentrations used. Figs. 12, 13 and 14 permit a compaiison of various basic radicals. When Figs. 12 and 13 are compared, it is readily apparent that calcium chlorid is more effective in inhibiting the swelling ABSORPTION, SECKETION— CELLS AND TISSUES 69 solution than in potassium chlorid. All (with the exception of the pure hydro- of gelatin in an acid the curves of Fig. 13 chloric acid curve) lie distinctly below the corresponding curves in Fig. 12.i If we make a Uttle allowance for ex- perimental errors we are probably safe in saying that the curves for sodi- um chlorid in Fig. 14 occupy a posi- tion between those given for potassium chlorid and calcium chlorid. As the acid radical is the same in these salts, the differences may be attributed to the effect of the basic radicals which as- sume the following familiar order in which that least effective in reduc- ing the swelling of gelatin is placed first. Potassium, Sodium, Calcium As the concen- trations of acids and salts employed 1 That curve V in Fig. 13 lies above IV represents an experimental error. The dry gelatin disc used for curve V was not as heavy as that used for curve IV. Thin discs swell not only faster, but somewhat more than thicker ones. 70 (EDEMA AND NEPHRITIS are the same in the experiments from which Figs. 9 and 12 (the two potassium chlorid series) have been constructed, the ques- tion arises why the curves in the latter lie lower than those in the former. The gelatin and all ex- ternal conditions were the same in these two sets of experiments except the temperature, and it is to the higher temperature prevailing when the experiments of Figs. 9, 10 and 11 were carried out (Sep- tember 4 to 7, 1908) than when those of Figs. 12, 13 and 14 were made (November 11 to 20, 1908), that I at- tribute the marked absolute differences in the amount of the swelling. A point that we will find of biologi- cal interest later is well brought out in Figs. 9 to 14. This is the amount of inhibition in the swelling with any unit increase in the concentration of the added salt. It is clearly evident that to double the concentration of the salt is not to double the diminution in swell- ing — in every case the diminution is less than might be expected. ABSORPTION, SECRETION— CELLS AND TISSUES 71 In Fig. 15 is illustrated the effect of adding equimolar solutions of different sodium salts to a solution of sodium hydroxid. It is easily seen how much more powerfully the citrate, phosphate, tartrate and sulphate interfere with the swelling of the gelatin discs in this alkaline solution, than the various univalent acid radicals. The general grouping of the salts as to the way in which 13 12 11 10 9 8 7 6 5 4 3 S 1 ; . /? GELATIN X y NaOH-Na=Salt Series y/ xx^ - y* yy^ NaOH/ ^oyyy' y^. ^^y\^ ^^^ulphale / yy^^ ^^^^ Tartrate, .^ :S=^^ospbate Citrate //^^(^ 12 Hours FiGUEE 15. 16 20 they affect the swelling of gelatin in solutions of adds and alkalies is therefore the same as that discovered in our study of the swelling of fibrin. Tables XX, XXI, XXII, XXIII, XXIV, XXV and XXVI contain the experimental data from which have been constructed, respectively, Figs, 9, 10, 11, 12, 13, 14 and 15. 72 (EDEMA AND NEPHRITIS TABLE XX Gelatin — Acid+Salt Dry wt. of gelatindiso. 0.800 0.802 0.803 0.809 0.810 0.813 50 cc. 50 cc. 50 cc. 60 cc. n/10 n/10 n/10 n/10 n/10 HCl+50 CO. m/1 HCl+40 cc. HCl+30 cc. HCl+20 CO. HCl+10 cc. n/10 HCI+SO cc. Solution. H2O+IO cc. H2O+2O cc. H2O+3O CO. H2O+4O cc. m/1 m/1 m/l m/1 potassium H2O. potassium potassium potassium potassium chlorid. chlorid. chlorid. chlorid. Hrs. in the solution. Gain in parts, of Dne part of gelatin. 1.40 1.36 1.21 1.21 1.03 1.00 1.81 6.05 3.02 2.49 2.39 2.05 1.87 4.26 8.05 4.11 3.42 3.17 2.75 2.53 6.86 12.40 5.41 4.48 4.06 3.57 3.32 6.63 23.35 7.27 6.01 5.40 4.66 4.48 10.82 35.25 8.56 7.06 6.34 5.67 5.29 12.79 47.05 9.58 8.01 7.21 6.51 6.11 14.16 I II III IV V TABLE XXI Gelatin — Acid+Salt Dry wt. of gelatindiso. 0.763 0.765 0.766 0.772 0.775 0.778 Solution. 50 cc. n/10 HCl+40 cc. H2O+IO CO. m/1 potassium citrate. 50 cc. n/10 HCl+30 cc. H2O+20 cc. m/1 potassium citrate. 60 cc. n/10 HCl+20 CO. H2O+3O oc. m/1 potassium citrate. 50 cc. n/10 HCl+10 cc. H2O+40 CO. m/1 potassium citrate. 50 oc. n/10 HCl+50 CO. m/1 potassium citrate. 50 cc. n/10 HCl+60 CO. H2O. Hrs. in the solution. Gain in parts of one part of gelatin. 1.40 6.05 8.05 12.40 23.35 35.25 47.05 1.11 1.93 2.64 3.39 4.42 5.12 5.57 I 1.01 1.86 2.46 3.13 4.04 4.67 5.09 II 0.95 1.68 2.61 (?) 2.81 3.63 4.15 4.52 III 0.73 1.25 1.64 2.08 2.80 3.26 3.59 IV 0.61 1.02 1.33 1.65 2.21 2.56 2.84 V 2.20 4.55 6.22 6.93 11.13 13.13 14.61 ABSORPTION, SECRETION— CELLS AND TISSUES 73 TABLE XXII Gelatin — A cid -j-Salt Dry wt. of gelatindisc. 0.778 0.782 0.783 0.788 0.790 0.794 Solution. 50 cc. n/10 HCl+40 cc. HjO+lO cc. m/1 potassium aulpho- cyanate. 50 CO. n/10 HCl+30 CO. H2O+20 CO. m/1 potassium sulpho- cyanate. 50 cc. n/10 HCl+20 00. H2O+3O 00. m/1 potassium sulpho- cyanate. 50 CO. n/10 HCl+10 CO. H2O+40 cc. m/1 potassium sulpho- cyanate. 50 CO. n/10 HCl+50 CC. m/1 potassium sulpho- cyanate. 50 00. n/10 HCl+50 00. H2O. Hrs. in the solution. Gain in parts, of one part of gelatin. 1.40 6.05 8.05 12.40 23.35 35.25 47.05 1.33 2.68 3.46 4.47 6.28 7.50 8.74 I 0.86 1.54 1.96 2.54 3.36 3.93 4.56 II 0.82 1.40 1.65 1.93 2.42 2.56 2.66 III 0.73 1.25 1.32 1.47 1.85 1.56 1.56 IV 0.68 1.07 1.16 1.18 1.43 1.11 Too sticky to weigh V 2.06 4.61 6.27 8.43 11.49 13.58 15.07 TABLE XXIII Gelatin — Acid+Salt Dry wt. of gelatindisc. 0.755 0.792 0.755 0.771 0.722 0.750 Solution. 50 cc. n/10 HCl+40 cc. H2O+IO CO. m/1 potassium chlorid. 50 CO. n/10 HCl+30 ce. H2O+20 ce. m/1 potassium chlorid. 50 CO. n/10 HCl+20 CO. HiO+30 CC. m/1 potassium chlorid. 50 cc. n/10 HCI+10 cc. H2O+4O cc. m/1 potassium chlorid. 50 cc. n/10 HCl+50 cc. m/1 potassium chlorid. 50 CO. n/10 HCl+50 CC. H2O. Hrs. in the solution. Gain in parts, of one part of gelatin. 14.00 27.35 36.15 48.45 75.35 144.15 213.15 4.5 5.9 6.4 6.8 7.5 9.2 10.3 I 3.7 4.9 5.3 5.8 6.3 7.9 8.9 II 3.2 4.2 4.6 5.0 5.8 7.0 7.8 III 3.2 4.0 4.5 4.8 5.6 6.9 7.7 IV 3.0 3.8 4.2 4.6 5.3 6.4 7.2 V 6.5 8.6 9.2 9.8 11.0 12.9 14.4 74 (EDEMA AND NEPHRITIS TABLE XXIV Gelatin — Acid+Salt Dry wt. of gelatindisc. 0.724 0.743 0.723 0.788 0.738 0.740 Solution. 50 cc. n/10 HCl+40 cc. HjO+lO cc. m/1 calcium chlorid. 50 cc. n/10 HCl+30 cc. H2O+2O cc. m/1 calcium chlorid. 60 cc. n/10 HCl+20 cc. H2O+30 cc. m/1 calcium chlorid. SO cc. n/10 HCl+10 cc. H2O+40 cc. m/1 calcium chlorid. 50 CO. n/10 HCl+50 00. m/1 calcium chlorid. 50 cc. n/10 HCl+50 cc. H2O. Hrs. in the solution. Gain in parts, of one part of gelatin. 14.00 27.35 36.15 48.45 75.35 144.15 213.15 3.8 3.5 5.3 5.6 6.4 7.6 8.5 I 2.9 3.9 4.4 4.7 5.5 6.8 7.6 II 2.9 3.7 4.2 4.6 6.4 6.8 7.8 III 2.3 3.2 3.7 4.1 5.0 6.8 7.9 IV 2.4 3.4 3.9 4.4 5.4 • 7.4 8.7 V 6.5 8.4 9.2 9.7 10.9 12.9 14.6 TABLE XXV Gelatin — A aid +SaU t Dry wt. of gelatindisc. 0.632 0.690 0.686 0.597 0.700 0.676 SO CO. 50 cc. 50 cc. SO cc. 50 00. n/10 n/10 n/10 n/10 n/10 50 cc. HCl +40 cc. HCl +30 cc. HCl +20 CO. HCl+lOoc. HCl +50 CO. n/10 HCl +60 CO. Solution. HiO+lOcc. H2O+20ec. H2O+3O CO. H2O +40 cc. m/1 m/1 m/1 m/1 m/1 sodium H2O. sodium sodium sodium sodium chlorid. chlorid. chlorid. chlorid. chlorid. Hra. in the solution. Gain i n parts, of ne part of g elatin. 14.00 6.0 3.9 3.5 3.6 2.8 7.1 27.35 6.2 4.9 4.4 4.3 3.6 8.7 36.16 6.6 5.3 3.5(?) 4.6 3.9 9.3 48.46 7.0 5.7 5.1 4.9 4.2 9.8 75.35 7.8 6.3 5.6 5.4 4.8 10.8 144.15 9.1 7.6 6.9 6.5 5.9 12.7 213.15 10.0 8.4 7.5 6.9 6.6 14.2 I II III IV V ABSORPTION, SECRETION— CELLS AND TISSUES 75 TABLE XXVI Gelatin — Alkali +SaZf Dry wt. of gelatin disc. 0,799 0.798 0.797 0.787 0.782 Solution. 50 cc. n/10 NaOH +50 cc. H2O. 50 cc. n/10 NaOH +50 cc. m/5 sodium acetate. 50 cc. n/10 NaOH +60 0. m/5 sodium bromid. 50 cc. n/10 NaOH +50 cc. m/5 sodium chlorid. 50 cc. n/10 NaOH+50cc. m/5 sodium citrate. Hrs. in the solution. Gain in parts, of one part of gelatin. 4.45 17.00 24.30 47.30 3.94 8.83 12.38 melted 2.71 6.00 8.12 too soft to weigh 3.37 7.42 9.83 too soft to weigh 3.14 6.91 9.29 too soft to weigh 2.22 4.73 6.23 8.65 much broken Dry wt. of gelatin disc. 0.776 0.770 0.756 0.755 0.754 Solution. 50 cc. n/10 NaOH +50 cc. m/5 sodium iodid. 50 cc. n/10 NaOH +50 cc. m/5 sodium nitrate. 60 cc. n/10 NaOH +50 CO. m/5 disodium phosphate. 60 CO. n/10 NaOH +50 cc. m/5 sodium sulphate. 50 cc. n/10 NaOH+50co. m/6 NaK tartrate. Hrs. in the solution. Gain in parts, of one part of gelatin. 4.45 17.00 24.30 47.30 3.34 7.40 11.02 melted 3.46 7.58 10.40 almost melted 2.41 4.78 6.10 9.11 good body 2,95 6,04 8,08 breaks on handling 2,88 5.95 7,82 firmer than preceding (e) Non-electrolytes do not share with electrolytes their marked power of reducing through their presence the amount that gelatin will swell in the solution of any acid or alkali. Figs. 16, 17 and 18 illustrate this better than words. The upper curve of Fig. 16 indicates the amount and rate of swelling of a gelatin disc in a pure hydrochloric acid solution. The black circles, crosses, squares and triangles just below this curve give the gains in weight of gelatin discs kept in equally concentrated hydro- chloric acid solutions to which various amounts of ethyl alcohol have been added. While present in amounts osmotically more than equivalent to the salts added in the previously described experiments, there is practically no reduction in the amount of swelling of the gelatin discs. The same is true when methyl 76 (EDEMA AND NEPHRITIS alcohol is added to a hydrochloric acid solution. To avoid confusion only one of these curves has been filled in, and the whole series has been placed somewhat to the right in the drawing. The methyl alcohol series is indicated in white crosses, squares, circles and triangles to distinguish it from the ethyl alcohol ABSORPTION, SECEETION— CELLS AND TISSUES 77 series. As readily apparent, the characters practically coincide with each other. CO Ci tH o In Fig. 17 is shown the effect of adding various amounts of glycerin and urea to a hydrochloric acid solution. The curve 78 (EDEMA AND NEPHRITIS Gelatin HC)-Sacoharose 23- HCl/ 20 'III TVy -L. Hours IB Figure 18. 24 32 ABSORPTION, SECRETION— CELLS AND TISSUES 79 for the pure hydrochloric acid solution occupies a position at about the middle of the series. The curves marked I, II, III and IV show the effect on swelUng of adding progressively larger amounts of glycerin to the hydrochloric acid solution. Glycerin produces a definite decrease in the amount of swelling, though as compared with the effect of any electrolyte, it is slight. Urea, on the other i6r hand, distinctly favors the swelling of gelatin in a hydrochloric acid solution, and this the more the higher the con- centration of the urea. The curves marked I', II', III' and IV' demon- strate this fact. Because the n o n - electrolytes are so comparative- ly ineffective in re- ducing the swell- ing of protein colloids in the presence of an acid many have made this state- ment read, entire- ly without effect. This is by no means the case, a fact which must be remembered for future discussion. The various sugars, for example, have, like glycerin, a decided dehydrating effect, especially in the higher concentrations. Fig. 18 illustrates this in the case of saccharose, which represents the most active of this class of compounds. The effect of various non-electrolytes on the swelling of gelatin in an alkaline solution is shown in Fig. 19. Only the curve for the pure sodium hydroxid has been filled in. As with acids, urea again favors the swelling. The addition of ethyl and methyl FlGUBE 19. 80 (EDEMA AND NEPHRITIS alcohols and glycerin is without effect, for these curves practically coincide with that for the pure alkali. In contrast hereto the addition of an electrolyte, hthium chlorid, produces a distinct diminution in the amount of the swelhng. The curves of Figs. 16, 17, 18 and 19 have been constructed from the experimental data contained in Tables XXVII, XXVIII XXIX and XXX respectively. TABLE XXVII Gelatin — A cid + Non-elecirolytes Dry wt. of gelatin disc. 0.773 0.765 0.756 0.748 0.723 Solution. 50 00. n/10 HCl +40 CO. H2O+IO CO. 2/m ethyl alcohol. 50 cc. n/10 HCl +30 CO. H2O+2O CO. 2/m ethyl alcohol. 50 cc. n/lO HCl+20 CO. H2O +30 CO. 2/m ethyl alcohol. 50 cc. n/10 HCl +50 CO. 2/m ethyl alcohol. 50 cc. n/10 HCl +50 CO. H2O. Hrs. in the solution. Gain in parts, of one part of gelatin. 13.20 24.45 67.30 97.05 113.35 162.45" 212.05 13 days 6.9 9.2 12.5 13.5 13.8 14.4 14.7 15.4 Black Circle 6.7 9.1 12.7 13.8 14.1 14.7 15.0 15.6 Black Square 6.8 6.2 8.8 8.7 12.4 12.3 13.5 13.4 13.9 13.9 14.5 14.4 14.8 14.7 15.3 15.0 Black Black Cross Triangl 7. 10. 13. 14. 14. 15. 15. 16. Dry wt. of gelatin disc. 0.733 0.729 0.727 0,724 Solution. 50 ec. n/10 HCl +40 CO. H2O+IO CO. 2/m methyl alcohol. 50 CO. n/10 HCl +30 00. H2O+2O00. 2/m methyl alcohol. 50 00. n/10 HCl+20 00. HjG +30 00. 2/m methyl alcohol. 50 CO. n/10 HCl +50 CO. 2/m methyl alcohol. Hrs. in the solution. Gain in parts, of one part of gelatin. 13.20 24.45 67.30 97.05 113.35 162.45 212.05 ' 13 days 7.6 10.1 13.1 14.0 14.4 14.9 15.4 16.0 White Star 6.9 9.4 12.7 13.7 13.9 14.6 14.9 15.5 White Circle 7.1 9.4 12.7 13.7 14.1 14.6 14.9 15.4 White Square 6.7 9.2 12.5 13.5 13.9 14.3 14.1 15.2 White Triangle ABSORPTION, SECRETION— CELLS AND TISSUES 81 TABLE XXVIII Gelatin — A cid + Non-electrolytes Dry wt. of gelatin disc. 0.829 0.872 0.842 0.816 0.810 50 cc. n/10 50 cc. n/10 50 CC. n/10 HCl +40 cc. HCl +50 cc. HCl +20 00. 50 00. n/10 HCl +50 CO. 50 CO. n/10 Solution. H2O+IOCC. H2O +20 cc. HiO+SO cc. HCl +50 CO. 2/m 2/m 2/m H2O. glycerin. glycerin. glycerin. glycerin. Hrs. in the solution. Gain in parts of one part of gelatin. 13.20 6.6 6.1 5.9 5.6 6.9 24.45 9.0 8.2 8.1 7.8 9.4 67.30 12.5 12.2 11.7 11.4 13.1 97.05 13.6 13.3 12.8 12.4 14.2 113.35 13.9 13.7 13.1 12.7 14.6 162.45 14.5 14.3 13.7 13.2 15.2 212.05 14.8 14.6 14.0 13.5 15.6 13 days 15.3 15.1 14.5 14.0 17.0 I II III IV Dry wt. of 0.851 0.848 0.837 0.832 gelatin disc. 50 cc. n/10 60 cc. n/10 50 cc. n/10 HCl +40 cc. HCl +30 cc. HC1+20CC. 50 CO. n/10 H2O+IO0C. H2O +20 CO. H2O+3O CC. HCl +50 CO. 2/ni urea. 2/m urea. 2/m urea. 2/m urea. Hrs. in the solution. Gain in parts of one part of gelatin. 13.20 6.4 7.2 7.2 7.3 24.45 9.1 10.1 10.2 10.8 67.30 13.3 14.7 15.0 18.1 97.05 14.5 16.1 16.7 20.7 113. 35 15.1 16.6 17.3 21.5 162.45 15.7 17.4 18.4 23.1 212.05 16.0 17.7 18.8 23.7 13 days 16.8 18.6 19.5 24.5 I' II' III' V (/) For purposes of biological application there are introduced here a series of experiments showing the effect of various non- electrolytes upon ordinary commercial gelatin when no acid or alkali has been added from without.^ No special comments are necessary upon the results reproduced in this paper, obtained by comparing the swelling of gelatin in water with the swelling of this same gelatin in. differently con- 1 Martin H. Fischer and Anne Sykes: Science, 38, 486 (1913); Kolloid- Zeitschr., 14, 215 (1914). 82 CEDEMA AND NEPHRITIS centrated solutions of various non-electrolytes. We used sac- charose, levulose, dextrose, methyl alcohol, propyl alcohol, TABLE XXIX Gelatin — A cid -{-Saccharose Dry wt. of gelatin disc. .518 .515 .515 .507 .502 5 CO. 5 CO. 5 CO. n/10 n/10 n/10 5 cc. n/10 HCl HCl HCl HCl + 5 CO. +30 CO. +50 00. +95 CO. 2/m 2/m 2/m H2O saccharose saccharose saccharose +90 00. +65 CO. +45 cc. H2O H2O H2O .501 .500 Solution. 100 cc. H2O 5 CO. n/10 HCl +75 cc. 2/m saccharose +20 00. H2O 5 00. n/10 HCl +95 00. 2/m saccharose Hrs. in the Solution. Gain in parts of one part of gelatin. 18.45 26.30 42.00 50.45 5.28 6.27 9.88 12.26 28.36 47.54 In sol 21.38 41.71 ution 15.90 31.90 58.17 65.07 9.72 17.27 36.25 51.75 3.62 7.18 13.50 17.50 1.89 3.98 6.08 7.78 TABLE XXX Gelatin — Alkali -\- Non-electrolytes Dry wt. of gelatindisc. 0.710 0.712 0.714 0.715 0.716 0.705 Solution. 50 cc. n/10 NaOH +50 00. m/5 LiCl. 50 00. n/10 NaOH +50 CO. 2/5 m urea. 50 cc. n/10 NaOH +50 cc. 2/5 m glycerin. 60 cc. n/10 NaOH +50 cc. 2/5 m ethyl alcohol. 50 cc. n/10 NaOH +50 00. 2/5 m methyl alcohol. 50 CO. n/10 NaOH +50 cc. H2O. Hra. inthe solution. Gain in parts, of one part of gelatin. 3.15 16.15 24.00 3.46 9.07 9.98 3.70 12.45 13.57 Melting Black circle 3.71 11.53 15.79 White circle 3.70 11.79 15.29 Triangle 3.64 11.51 15.60 Square 3.80 11.70 15.82 propylene glycol and acetone. The presence of all these non- electrolytes reduces the amount of the swelling, and this the more the higher the concentration of the added silbstance. ABSOEPTION, SECRETION— CELLS AND TISSUES 83 When these curves are paralleled with those available on the effect of different electrolytes (salts) on the swelhng of gelatin K a in the presence of acid one is impressed with the fact, when equi- molar or osmotically equivalent solutions are compared, that the non-electrolytes are relatively most powerful in their de- 84 (EDEMA AND NEPHRITIS hydrating effects in the higher concentrations, while of the elec- trolytes the reverse is true. Thus, in even low concentrations fa « P C5 the salts produce a great dehydrating effect, but with every unit increase in concentration the degree of shrinkage becomes pro- gressively less. Just the opposite holds for the non-electrolytes. ABSORPTION, SECRETION— CELLS AND TISSUES 85 where low concentrations are comparatively ineffective, but where an unexpectedly great dehydrating effect is observed as the concentration rises. The reasons for this behavior are not as yet clear, but it opens an interesting theoretical field wherein the problems of chemical combination and of adsorption 86 OEDEMA AND NEPHRITIS TABLE XXXI Gelatin — Saccharose 1 Dry wt. of .726 .721 .722 .723 .724 .725 .726 gelatin disc. 10 cc. 20 oc. 30 CO. 50 cc. 75 CO. 100 oc. 2/m 2/m 2/m 2/m 2/m 100 CO. saccharose saccharose saccharose saccharose saccharose 2/m +90 cc. +80 cc. +70 cc. +50 cc. +25 CO. saccharose HjO H2O HjO H2O HiO Hrs. in the Solution. Gain in parts of one part of gelatin 2.45 1.51 0.81 1.13 1.03 0.78 0.52 0.26 17.15 5.61 5.03 4.66 4.01 2.69 1.43 0.65 25.00 6.65 6.11 5.59 4.97 3.41 1.76 0.90 42.15 9.51 8.29 7.83 7.21 5.18 2.65 1.15 65.15 11.69 9.36 9.34 8.74 6.51 3.53 1.54 TABLE XXXII Gelatin — Leuulose Dry wt. of .732 .744 .750 .759 .797 .797 .797 gelatin disc. 10 00. 20 cc. 30 oc. 50 00. 100 00. 2/m 2/m 2/m 2/m 100 00. 100 00. Solution. H2O levulose levulose levulose levulose 2/m 4/m +90 00. +80 CO. +70 oc. +50 CO. levulose levulose ao H2O H2O H2O Hrs. in the G ain in parts of one part of gela :in. solution. Hours. 7.00 2.90 2.90 2.78 2.71 2.55 3.50 1.16 0.46 22.30 6.01 5.80 5.67 5.38 5.29 19.00 3.74 1.13 31.00 6.77 ■ 6.47 6.25 6.23 6.03 28.20 4.89 1.43 46.30 8.44 7.97 7.87 7.71 6.82 43.60 7.24 2.11 55.00 8.77 8.07 8.17 8.11 8.27 TABLE XXXIII Gelatin — Dextrose Dry weight of gelatin disc. .678 .699 ;699 .713 Solution. 100 CO. H2O 10 CO. 2/m dextrose +90 00. H2O 20 CO. 2/m dextrose +80 CC. H2O 30 00. 2/m dextrose +70 cc. H2O Hours in the solution. Gain in parts of one part of gelatin. 7.30 23.15 31.15 47.15 55.45 2.44 5.85 6.51 8.26 8.72 2,76 5.82 6.46 8.09 8.19 2.75 5.61 6.31 7.92 8.17 2.27 4.97 5.71 7.29 7.62 ABSORPTION, SECEETION— CELLS AND TISSUES 87 between protein and dissolved substances is further complicated by the possibility of having no water present for the swelling of BS P the protein because it is all combined with the various non- electrolytes (especially in the case of the viscid sugar solutions) . But while all these non-electrolytes reduce the swelling of 88 (EDEMA AND NEPHRITIS gelatin there exist some interesting quantitative differences between them. As shown in Figs. 20, 21 and 22 and the corre- sponding Tables XXXI, XXXII and XXXIII, the various sugars all reduce the sweUing of gelatin, but at the same con- centration saccharose is far more powerful in this regard than ABSORPTION, SECRETION— CELLS AND TISSUES 89 either levulose or dextrose, which produce approximately equal degrees of dehydration. This is readily apparent on comparing curves I, II, III, IV and VI of Fig. 20 with the corresponding curves I, II, III, IV and V of Fig. 21, or the first three of these with curves I, II and III of Fig. 22. 90 (EDEMA AND NEPHRITIS Methyl and propyl alcohols, propylene glycol and acetone all approximate the monosaccharids in the degree of dehydration a which they bring about. Only in very high concentrations are they able to bring about a dehydration which saccharose brings about in much lower ones, as readily apparent when Figs. 23, 24, ABSORPTION, SECRETION— CELLS AND TISSUES 91 25 and 26 are compared with Fig. 20. Tables XXXIV, XXXV, XXXVI and XXXVII contain the experimental data from which Figs. 23 to 26 have been constructed. TABLE XXXIV Gelatin — Methyl Alcohol Dry weight of gelatin disc. .747 .729 .729 .739 .744 .746 2 DC. 5 cc. 10 cc. 20 cc. 50 cc. 10/m 10/m 10/m 10/m 10/m methyl methyl methyl methyl methyl alcohol alcohol alcohol alcohol alcohol +98 cc. +95 CO. +90 cc. +80 CO. +50 cc. HjO H2O HzO H2O H2O .746 Solution. 100 cc. H2O 100 cc. 10/m methyl alcohol Hours in the solution. Gain in parts of one part of gelatin. 5.30 21.30 29.30 48.00 53.30 2.95 6.32 7.12 8.54 9.52 2. 85 2. 57 2.55 2.36 1.60 6.27 5.87 5.50 3.15 3.12 7.01 6.63 6.27 5.50 3.58 8.30 7.83 7.43 6.45 4.09 8.21 8.39 7.87 6.76 4.29 0.93 1.69 1.97 2.14 2.19 TABLE XXXV Gelatin — Propyl Alcohol Dry weight of gelatin disc. Solution. .708 100 cc. H2O 2 cc. 10/m propyl alcohol +98 cc. H2O .683 5 cc. 10/m propyl alcohol +95 cc. H2O .683 10 cc. 10/m propyl alcohol +90 cc. ao .689 20 00. 10/m propyl alcohol +80 cc. H2O .690 50 cc. 10/m propyl alcohol +50 cc. H2O .707 100 cc. 10/m propyl alcohol Hours in the solution. Gain in parts of one part of gelatin. 5.00 20.00 29.00 46.00 2.85 7.28 8.71 11.16 2.72 2.59 2.46 2.08 1.19 7.51 7.21 6.74 4.25 2.15 8.85 8.83 8.07 5.70 2.75 10.39 10.37 8.54 7.61 3.43 0.28 0.51 0.55 0.58 (g) The absorption and secretion of water by gelatin represent in large part reversible processes. This fact is brought out in Fig. 27 and Table XXXVIII, from which it is constructed. When a gelatin disc is transferred from a pure hydrochloric acid 92 (EDEMA AND NEPHRITIS or sodium hydroxid solution into an equally concentrated one containing a salt, a prompt fall in the absorption curve is noted A rise in the curve follows the reverse process. A further fact of interest in Fig. 27 is that at the same concentration potassium citrate inhibits the swelling of gelatin in a hydrochloric acid solution more than in an equinormal sodium hydroxid solution. TABLE XXXVI Gelatin — Propylene Glycol Dry wt. of .718 .755 .752 .748 .747 .743 .733 .709 gelatin disc. 2 CO. 5 cc. 10 cc. 20 cc. 30 cc. 50 cc. 80 cc. 10/m 10/m 10 m 10/m 10/m 10/m 10/m 100 CO. propy- propy- propy- propy- propy- propy- propy- Solution. lene lene lene lene lene lene lene glycol glycol glycol glycol glycol glycol glycol +98 cc. +95 CO. +90 cc. +80 00. +70 cc. +50 cc. +20 CO. H2O H2O H2O H2O H2O H2O H2O Hrs. in the solution. Gain in E arts of on 2 part of gelatin. 6.15 2.49 2.19 2.07 1.59 1.36 1.09 0.89 0.40 20.45 4.60 4.49 4.36 3.43 3.06 2.32 1.61 0.74 29.45 5.30 5.98 4.85 3.91 3.53 2.85 1.79 0.92 45.00 6.82 6.26 6.11 4.97 4.69 3.72 2.86 1.43 64.30 7.33 7.14 6.69 5.47 6.00 4.04 3.17 1.46 TABLE XXXVII Gelatin — A cetone Dry weight of gelatin disc. .808 .808 .807 .810 .811 Solution. 100 cc. H2O. 2 cc. 10/m acetone +98 CO. H2O. 5 CO. 10/m acetone +95 CO. H2O. 10 cc. 10/m acetone +90 cc. H2O. 20 cc. 10/m acetone +80 CO. H2O. Hours in the solution. Gain in parts of one part of gelatin. 16.00 26.30 39.30 49.15 66.00 4.06 5.59 7.49 8.39 10.39 3.72 6.05 6.55 7.43 9.27 3.56 4.99 6.48 7.20 8.44 3.37 4.78 6.15 6.69 7.57 3,18 4.28 6.78 6.10 6.60 (h) There are other substances besides acids which will in- crease the amount of water that gelatin can hold. In paragraph (e) above, we found urea to do this when added to an acid solu- ABSORPTION, SECRETION— CELLS AND TISSUES 93 16 / \ HCl+Salt 15 / V^_ 14 / / 13 / / 12 11 / NaOH J 10 HCl / / NaOn// 9 /NaOH / /hCI 8 / J I ■? NaOH +„--^galt / 6 y-^ / 5 HCl+Salt___-J- 4 ^^'^^ 3 GELATIN 2 1 0- 1 1 12 Hours Figure 27. 16 20 TABLE XXXVIII Gelatin 24 Dry weight of gelatin disc. I 0.705 II 0.705 III 0.722 IV 0.721 Solution. 60 CO. n/10 NaOH +50 CO. H2O. 50 cc. n/10 NaOH +50 oc. m/5 K citrate. SO cc. n/10 HCl +50 CO. H2O. 50 cc. n/10 HCl +50 cc. m/5 K citrate. Houra in the solution. Gain in parts of one part of gelatin. 3.15 16.15 3.80 11.70 2.74 7.39 5.44 16.74 2.26 5.02 Disc I is put into Solution II; Disc II into Solution I. Disc III is put into Solution IV; Disc IV into Solution III. 18.25 20.35 24.00 10.49 10.63 11.21 9.85 11.19 11.34 15.20 14.70 14.73 7.64 12.26 14.77 94 (EDEMA AND NEPHRITIS tion. It will do this also in neutral solution as shown in Fig. 28 and Table XXXIX, from which this is constructed. When the urea is sufficiently concentrated the gelatin goes into solution. Pyridin represents another substance which has activities in this direction, as shown in Fig. 29 and Table XL. The hydrating ABSORPTION, SECRETION— CELLS AND TISSUES 95 effect of urea is not a simple alkali effect, for acids in no con- centration counteract it. This is also true of pyridin, which, however, exhibits a slight alkali effect in addition. Acids as well as alkalies simply increase the hydrating effects of these substances. Some of the amins seem also to belong in this group with urea and pyridin, but they are so intensely alkaline 96 (EDEMA AND NEPHRITIS in watery solution that special pains need to be taken to eliminate first this alkaline effect before their more specific hy- drating action becomes clearly evident. TABLE XXXIX Gelatin — Urea Dry weight of gelatin disc. Solution. .786 100 cc. H2O .782 2 cc. :o/m urea +98 CO. H2O .793 5 CC. 10/ m urea +95 CO. HzO .796 10 oc. 10/m urea +90 cc. H2O .759 20 cc. 10/m urea +80 ,< H2O .760 50 cc. 10/m urea +50 cc. H2O .788 100 00. 10/m urea Hours in the solution. 5.30 21.45 45.15 Gain in parts of one part of gelatin. 2.60 5.93 7.97 2.67 2.88 3.26 2.96 6.83 8.17 10.65 Complete 9.97 11.36 14.45 solution Complete solution. TABLE XL Gelatin — Pyridin Dry weight of gelatin disc. .780 .780 .781 .783 .783 .783 .784 .785 Solution 100 cc. H2O 0.1 cc. 6/m pyridin + 100 00. H2O 0.2 cc 5/m pyridin + 100 CO. H2O 0.5 oc. 5/m pyridin + 100 CO. H2O 1 CO. 5/m pyridin +99 cc. H2O 2 cc. 5/m pyridin +98 cc. H2O 3 00. 5/m pyridin +97 00. H2O 5 cc. 5/m pyridin +95 cc. H2O Hours in the solution. Gain in parts of one part of gelatin. 16.45 38.45 46.30 62.30 73.00 3.75 6.12 6.51 8.00 9.29 3.77 6.43 6.87 8.48 9.61 4.01 6.73 7.23 8.96 9.85 4.19 6.98 7.51 9.16 10.45 4.44 7.59 8.22 .10.12 13.23 4.63 8.56 9.28 11.91 13.48 4.84 9.85 10.77 14.19 16.60 5.00 10.40 11.52 15.63 18.38 (i) As previously emphasized for fibrin, the hydration induced through urea or pyridin is of a different type from that produced through acids or alkalies, for while the latter is reducible through the addition of different salts this is not the case with the former. As shown in Fig. 30 and Table XLI, the slight hydrating effect ABSORPTION, SECRETION— CELLS AND TISSUES 97 of pure sodium chlorid upon gelatin is merely added to that pro- duced by urea. On the other hand, various non-electrolytes, « P O such as the sugars, reduce a urea or pyridin hydration very markedly, while, as noted above, they are relatively ineffective in the case of an acid or alkaU hydration. The effect of dextrose 98 (EDEMA AND NEPHRITIS on urea and pyridin hydration is illustrated in Figs. 31 and 32 and Tables XLII and XLIII, from which they are drawn. It is well in concluding this section to say a word regarding the similarities and the differences to be noted between the swelling of fibrin and the swelling of gelatin. The two behave similarly ABSORPTION, SECRETION— CELLS AND TISSUES 99 in that both swell more in the solutions of acids and alkaUes than in water; both swell to different degrees in equinormal solutions of different acids or alkalies, and the order in which these acids 20 10 Gelatin Pyridin - Dextrose Pyridin , II H,0 Hours 16 Figure 32. 24 and alkahes are effective is much the same; the swelling of both in either acid or alkaUne solutions is markedly inhibited through the presence of electrolytes, and this the more the higher the 100 CEDEMA AND NEPHRITIS concentration of the electrolytes. In contrast to the action of the electrolytes, the non-electrolytes are comparatively ineffective in this regard. TABLE XLI Gelatin — Urea +NaCl Dry weight of gelatin disc. .792 .787 .786 .773 Solution. 20 cc. 5/m urea +80 cc. H2 20 cc. 5/m urea +10 cc. m/1 NaCl +70 cc. H2O 20 cc. 6/m urea +26 cc. m/1 NaCl +55 cc. H2O 20 cc. 5/m urea +50 CO. m/1 NaCl +30 cc. H2O Hours in the solution. Gain in parts of one part of gelatin. 5.15 20.15 29.15 43.15 52.00 2.53 6.54 7.73 11.03 12.85 3.05 7.65 9.03 12.64 13.87 3.50 9.02 10.80 14.54 17.41 3.32 8.93 10.52 14.91 17.75 TABLE XLII Gelatin — Urea -\- Dextrose Dry weight of gelatin disc. .812 .812 .812 .812 .814 Solution. 20 cc. 5/m urea +80 cc. HjO 20 cc 5/m urea + 5 cc. 2/m dextrose +75 cc. HiO 20 CO. S/m urea + 10 CC. 2/m dextrose +70 CC. H2O 20 CC. 5/m urea +25 cc. 2/m dextrose +55 cc. H2O 20 cc. 5/m urea +50 ec. 2/m dextrose +30 cc. H2O Hours in the solution. Gain in parts of one part of gelatin. 20.30 25.30 40.30 50.00 9.15 10.52 14.73 16.27 9.01 10.29 14.66 16.25 8.56 9.78 13.87 15.49 7.10 8.18 11.89 13.38 5.88 6.81 10.17 11.65 Urea and pyridin are representative of another class of sub- stances which increase the power of both fibrin and gelatin to swell, but this type of increased hydration, while not affected by salts is markedly reduced through various non-electrolytes such as the sugars, There exist, on the other hand, certain differences between the sweUing of fibrin and the swelling of gelatin. These are for the most part of a quantitative nature. Gelatin is able to absorb under optimal conditions about sixty-five times its weight ABSORPTION, SECRETION— CELLS AND TISSUES 101 in water. With fibrin I have obtained values up to forty times its weight in absorbed water. While fibrin swells more in alkaline solutions than in equally concentrated acid solutions, gelatin does the reverse. This may, however, be only a seeming differ- ence, because of the usual acid content of the commercial gelatins, and the consequent formation of salts when they are made to swell in alkaline solutions. Fibrin attains its maximal swelUng in concentrations of acid which are much below those necessary to produce the maximum amount of swelling in gelatin, and higher concentrations of electrolytes are necessary to reduce markedly the swelling of gelatin in acid solutions than are neces- sary in the case of fibrin. On the other hand, urea and pyridin seem able to induce a relatively higher hydration in gelatin than in fibrin. All these statements must, however, not be taken too strictly, for, depending upon the history of their preparation, etc., the gelatins differ widely from each other. TABLE XLIII Gelatin — Pyridin + Dextrose Dry weight of .793 .795 .792 .790 .767 .736 .736 gelatin disc. 5 cc. m/1 5 cc. m/1 5 cc. m/1 5 cc. m/1 5 cc. m/1 pyridin pyridin pyridin pyridin pyridin 5 cc. m/1 + 5cc. +10 cc. +20 cc. +30 CO. +50 cc. Solution. 100 cc. pyridin 2/m 2/m 2/m 2/m 2/m H2O +95 cc. dextrose dextrose dextrose dextrose dextrose HsO +90 cc. +85 cc. +75 cc. +65 cc. +45 cc. mo H2O H2O H2O HiO Hours in the solution. Gai a in parts of one part of gelatin. 18.00 6.92 12.35 11.60 10.32 9.24 8.79 6.67 26.30 8.33 15.84 14.46 13.12 11.77 10.92 8.29 42.00 10.01 20.36 18.46 16.94 15.09 13.69 10.53 51.00 9.72 21.28 18.85 17.65 15.78 14.32 11.18 66.00 10.06 23.25 17.72 19.28 17.13 15.63 12.62 Such similarities and differences in the behavior of different colloids toward the same external conditions demand detailed study, for they are of the utmost biological importance. Pro- toplasm consists of a mixture of many different colloids. Not only are different colloids found in the same cell, but essentially different colloids form the basis of different tissues (bone, car- tilage, muscle, connective tissue, parenchymatous organs, central nervous system). It is at once apparent, therefore, that not 102 CEDEMA AND NEPHKITIS only so far as water absorption and secretion is concerned, but so far as any physiological reaction dependent upon the colloid constitution of living matter is concerned, a single variation in internal or external conditions may be followed by quite a different response either qualitatively or quantitatively, not only by different tissues, but by different parts of the same tissue or even the same cell. In a study of the behavior of different colloids toward the same group of external conditions we may therefore hope to discover much to aid us in our attempt to analyze the apparently limitless variations in the reactions of protoplasm to various external " stimuli." 4. Observations on the Swelling of Gluten. The absorption of water by proteins has recently received interesting elaboration by the work of Fred W. Upson and J. W. Calvin ^ in their study of wheat gluten. The gluten was pre- pared by washing flour free of its starch with distilled water. It was rolled out between glass plates to uniform thickness, and small round pellets weighing approximately 1.25 gram were cut from this with a large cork borer. Gluten behaves very much like fibrin and gelatin, Thus, it swells more in any acid than in pure watei This is well shown in Fig. 33. The beaker on FlGUEE 33, the extreme left shows a pellet of gluten in distilled water. The six beakers to the right contain progressively stronger solutions of lactic acid ranging from n/500 to n/10. Entirely similar series may be arranged for other acids. The addition of any salt to an acid solution inhibits the swelling, and this the more the higher the concentration of the 1 Feed W. IJpsoisf and J. W. Calvin: Personal Communication (1914). ABSORPTION, SECRETION— CELLS AND TISSUES 103 added salt. This is well illustrated in Figs. 34, 35 and 36. Beaker 1 in each of the series contains pure n/100 lactic acid; the re- FlGUBE 34. maining beakers, increasingly greater amounts (from m/1000 to m/25) of different salts, potassium chlorid in Fig. 34, dipotassium phosphate in Fig. 35 and potassium tartrate in Fig. 36. , i 1 ■ tr>. i^< L«i. \9i J ^ n ^ V « ^ V as :^ Figure 35. Figs. 37 and 38 bring out these relationships yet more clearly. In Fig. 37 is shown the amount of water absorbed in different concentrations of three different acids. Concentration is plotted FiGUBE 36, on the horizontal, increase in weight in terms of the original weight of the (moist) pellet on the vertical. The optimal swelling point is exceeded earlier in the case of hydrochloric acid than 104 CEDEMA AND NEPHRITIS m a p o ABSORPTION, SECRETION— CELLS AND TISSUES 105 in the case of lactic or acetic. A highly interesting feature of these gluten experiments is the fact that even such " weak " acids as lactic and acetic show an optimal concentration for swell- ing beyond which the protein swells less than in lower concentra- tions of the acid. Fig. 38, in addition to showing that all salts reduce the swell- ing of gluten in an acid solution, also shows that at the same con- centration different salts are unequally effective in this regard. Figure 38. Thus, calcium chlorid produces a greater dehydration than potas- sium chlorid, and the tartrate is more powerful than the phosphate. Some earher experiments by T. B. Wood and W. B. Hardy ^ on the " cohesiveness " of gluten bring out from an experimental point of view what amount in essence to the same facts as those of Upson and Calvin. Wood and Hardy found gluten to "disintegrate" and "dissolve" in dilute acids. The loss of cohesion depended upon the nature of the acid and its concen- tration and in about the way in which the swelling of fibrin, 1 T. B. Wood and W. B. Hardy: Proc. Roy. Soc. London, Series B, 81, 38 (1908). 106 (EDEMA AND NEPHRITIS gelatin and gluten depends upon these factors. Salts inhibited the action of the acid, both their concentration and their nature being of great importance in the matter. These experiments show how plant protein behaves in a fash- ion identical with the previously studied animal proteins. Their importance in the general biological problem of water absorp- tion will become apparent as we proceed. The value of Wood, Hahdy, Upson and Calvin's work in many other directions, as for the theory and practice of flour manufacture, bread mak- ing, etc, needs no emphasis. 6. Hydration and Dehydration in Liquid Colloids. The three colloids discussed thus far are essentially solid in character, and their behavior corresponds, as we shall see imme- diately, with the more solid constituents of living matter such as the muscles, parenchymatous organs, nervous tissues or eyes of our own bodies. But permeating these more solid structures we find in the higher animals streams of liquid colloid material which we call blood, lymph or tissue juice. How do such liquid protein colloids behave when subjected to the action of acids, alkalies and salts? Do they " swell " and " shrink " as do the solid colloids already discussed? The answer to this question, which is of fundamental importance for the solution of a whole series of biological phenomena, has been given us through the work, more especially of Franz Hofmeister,i Wolfgang Pauli,^ W. B. Hardy,^ p. von Schroeder,* Hans Handovsky^ and K. Schorr.^ A liquid colloid such as a solution of gelatin, blood serum or egg albumin cannot, of course, be seen to swell or shrink ' Franz Hofmeisteb; Arch. f. exp. Path. u. Pharm., 27, 395 (1890 ibid., 28, 210 (1891). 2 Wolfgang Pauli: Pfliiger's Arch., 67, 219 (1897); ibid., 71, 1 (1898). Hofmeister's Beit. z. chem. Physiologie, numerous papers in the years 1902 to 1908; Biochem. Zeitschr., 17, 235 (1909); ibid., 18, 340 (1909); ibid., 24 239 (1910). A general statement of his views is found in Kolloid Zeitschr. 7, 241 (1910). 'W. B. Hardy: Jour. Physiol., 24, 288- (1899); ibid., 33, 251 (1905) Proc. Royal Soc, London, Series B, 79, 413 (1907); Zeitschr. f. physilc Chem'. 33, 385 (1900). < P. VON Schrobder: Zeitschr. f. physik. Chem., 45, 75 (1903). ' Hans Handovsky: Fortsohritte in der KoUoidchemie der Eiweiss- kOrper, Dresden (1911), where reference to his earUer napers will be found. °K. Schorr: Cited by Pauli and Handovsky. ABSORPTION, SECRETION— CELLS AND TISSUES 107 in a test-tube. We must therefore use some other method of discovering such changes and measuring them. This is accom- plished by determining the viscosity of the liquid colloid by permitting it to flow through a capillary tube. Evidently, as the separate colloid particles in a colloid solution swell, they take up the pure solvent about them, and as such sweUing pro- gresses it must become increasingly difiicult for the particles to move over each other. The viscosity of the solution must therefore rise, and this betrays itself by an increase in the time required for a certain volimie of the colloid solution to flow through a standard capillary tube. Conversely, as the particles shrink the pure solvent is squeezed off, and so the viscosity must tend to fall back toward that of the pure solvent. Wolfgang Patili ^ has in this way studied blood serum from which the various admixed crystalloids have been removed by long dialysis against pure water. Such a solution is perfectly clear and stable. If its viscosity is measured it is found to be considerably higher than that of pure water owing to the colloid material in it. If a trace of acid is added the viscosity is enor- mously increased. But with progressive additions an upper limit is reached in the case of such acids as hydrochloric, hydro- bromic, nitric or sulphuric, beyond which a further addition of acid does not further increase, but decreases viscosity. For the weaker organic acids, such as acetic, no such optimal point has yet been found. The addition of any salt to the acidified serum markedly reduces the viscosity. With the same salt the degree of reduction increases with the concentration of the salt. With a given concentration of any series of salts very different degrees of reduction in viscosity are obtained. Thus, when sodium salts are compared, the chlorid, nitrate and sulphocyanate are found to be less powerful than the acetate or sulphate, and in the order named. The addition of a non-electrolyte is con- spicuously less effective in this regard. A practically identical series of findings has been estabhshed for the effects of alkali or of alkali plus various salts or non-electrolytes. It is readily apparent that these statements are point for 'Wolfgang Pauli: Naturwissensch. Rundschau, 21, 3 (1906); Physical Chemistry in the Service of Medicine, 136, translated by M. H. Fischer, New York (1907). Pauli and H. Handovsky: Biochem. Zeitschr., 18, 340 (1909). 108 CEDEMA AND NEPHRITIS point analogous to those made previously regarding fibrin, gelatin and gluten, and hence justify the conclusion that liquid (protein) colloids behave toward various external conditions in the same way as do the more solid ones. These changes in the swelling of fibrin, gelatin or gluten, or the viscosity changes of a liquid colloid, may opportunely be correlated here with changes in certain other properties. When acids, bases or salts are added to a protein colloid we ob- serve variations not only in its swelling or viscosity, but in its precipitability or coagulability and in its optical behavior. What relation do these bear to each other? The fundamental change remains the same, namely, a change in the hydration capacity of the involved colloids. As already pointed out, whatever makes gelatin or fibrin swell increases viscosity, and vice versa. As the degree of hydration is increased, the intimacy of the colloid with its solvent is evidently increased, and so we should expect its stability to be increased. We are not surprised, therefore, to find that whatever increases hydration increases the stability of a colloid, while, conversely, whatever does the reverse favors instability, in other words, precipitation and coagulation. Thus, pure serum albumin is easily precipitated by heat or alcohol. When a little acid is added the hydration capacity of the colloid is increased and corresponding herewith, its precipitability through heat or alcohol is lost. But if yet more acid is added the hydration optimum is exceeded and now heat and alcohol regain their power of precipitating the protein. In a similar way the protein after being rendered non-precipitable througn acid can again be precipitated by heat if a salt is added to the acid pro- tein, for this again lowers the hydration capacity of the colloid.^ An analogous series of observations is available regarding changes in the optical behavior of protein colloids. We see from this that a series of reactions in certain protein colloids which at first seem to have nothing to do with each other are reducible in the end to a comparatively simple set of changes. And as we proceed we shall find that protoplasm, which is in essence but a colloid matrix of the type of fibrin, gelatin or blood serum, fol- lows similarly simple laws. In the normal water content of a ^ The ordinary heat coagulation test for albumin in the urine makes use of these principles. The albumin is coagulated best when acid and salt are first added to the urine. ABSORPTION, SECRETION-CELLS AND TISSUES 109 cell we shall see again a swollen colloid, and in oedema the same colloid swollen to a greater amount. Changes in the viscosity of the blood will come to mean changes in its degree of hydration, while corneal opacities in glaucoma and changes in the normal refraction and diffraction of the clear media will come to mean dehydration and precipitation of certain protein colloids present in the tissues of the eye. Ill THE ANALOGY BETWEEN THE SWELLING OF CERTAIN PROTEIN COLLOIDS AND THE SWELLING OF PROTO- PLASM Having become famiUar with the effect of various external conditions on the swelling of several simple so-called hydro- philic or emulsion colloids (fibrin, gelatin, gluten, blood serum), we have at our disposal some facts which we may utilize in an attempt to analyze the ways and means by which tissues hold their normal amount of water, and to discover how under altered external conditions they may come to hold more or less than is considered normal. It is evident that could we show that the same conditions which make fibrin, gelatin or gluten take up and give off water, affect protoplasm similarly, a real step forward in the solution of this problem of the absorption and secretion of water by the tissues would be made. This can be done and with great simplicity. As the following paragraphs show, the absorption of water by various tissues is entirely analogous to the absorption of water by fibrin, gelatin or gluten. 1. The Analogy between the Absorption of Water by Certain Protein Colloids and by Muscle Simple facts regarding the absorption of water by various cells and tissues are very numerous and date back to the earliest periods of modern physiology. We shall have occasion to review them later. So far as water absorption by muscle is concerned, 0. Nasse 1 studied this question from an osmotic standpoint as far back as 1869, and E. BHticKE ^ touched some aspects of 10. Nasse: Pfliiger's Archiv., 2, 97 (1869). ''E. Bbucke: Sitzungsber. d. math. Naturw. CI. d. kais. Akad. d. Wis- sensch., 55, 622 (1867). no (EDEMA AND NEPHRITIS the problem even earlier. Most of the investigations of this particular type of tissue made since then are useless for our purposes because they antedate the years in which adequate use of the principles of physical chemistry first began to be made in biological studies. The period of interest to us begins with 1898, when Jacques Loeb ^ published the results of some experiments on the influence of acids, alkalies and various salts on the absorption of water by the gastrocnemius muscle of the frog. He found that muscle absorbs much water if placed in distilled water or in solutions of various acids or alkalies. From his earlier experiments he concluded that a muscle does not change in weight if kept in a solution having an osmotic pressure equal to that of the blood, but that it gains or loses weight if placed in solutions having respectively a lower or a higher osmotic pressure. About the same conclusion had been previously reached by Nasse. But Nasse noted that certain salts, notably the sulphates, bromids and iodids, exhibited a greater than calculated " osmotic " effect. Loeb made a similar observation when he discovered that in spite of isos- moticity a frog's muscle will absorb more water from a potas- sium chlorid solution than from one of sodium chlorid, and more from this than from one of calcium chlorid. The analogy be- tween the latter fact and the absorption of water by potassium, sodium and calcium soaps was pointed out, but our conceptions of the colloids had not at that time advanced to the point of recognizing in the soaps examples of this class of bodies. As much controversy has hedged about the question of the historical development of the colloid-chemical theory of water absorption by protoplasm it is well to emphasize that Loeb not only never contributed anything to its establishment, but actually thrust such aside .2 The action of acids and alkalies on muscle Loeb Uacques Loeb: Pfliiger's Archiv, 69, 1 (1898); ibid., 71, 457 (1899); ibid., 75, 303 (1899). 2 Loeb : Pfliiger's Arch., 77, 305 (1899) says: " The analogy between the absorption of water by soaps and by muscle is of importance in explaining the mechanism of water-absorption. The majority of authors, for example, HoFMEiSTER, assume that in the absorption of fluids by tissues we deal with imbibition; that is to say, with capillary phenomena. But in the absorption of fluid by soaps we deal with solution phenomena. The forces active here are osmotic and not the surface tension forces active in capillary phenomena." (In Bezug auf die Mechanilc der Fliissigkeitsresorption ist die Analogie zwisohen dem Verhalten von Seifen und dem Muskel von Be- ABSORPTION, SECRETION— CELLS AND TISSUES HI brought into harmony with the then current osmotic conceptions of absorption by assuming that they induced changes within the muscle tissues whereby the osmotic pressure of the cell contents was raised, as previously emphasized for a series of other animal tissues by H. J. Hamburgek,i C. von Limbbck,^ GtJRBEK 3 and C. Eijkman> The experiments of Ralph W. Webster ^ and E. Overton ^ followed those of Loeb. Webster concluded that osmotic effects could only explain the absorption from water and solu- tions of cane sugar. His careful study of the effects of electrolytes showed unequivocally that simple osmotic effects are out of the question here. Overton came to essentially the same conclusion and attempted to help out the problem by his conception of lipoid membranes about living cells and their entire impermea- bility to salts. He showed conclusively that Loeb's explanation of the action of acids and alkalies cannot be correct, for were all the proteins, carbohydrates and fats contained in muscle, split into their simplest digestion products they would still not yield a sufficient number of molecules to account, through con- ceptions of osmotic pressure, for the amount of water absorbed by muscle in the solution of an acid or an alkali. To certain other of Overton's ideas we shall have occasion to return later. Both in individual experimental results and in the conclusions drawn from them there exist many contradictions between the findings of these various authors. It is needless to touch upon them in detail. For a majority of these differences an explanation can readily be foimd. None of the authors mentioned ever studied the curves of absorption of water by muscle under various condi- deutimg. Die Mehrzahl der Autoren z. B. Hofmeistbr, nehmen an dass es sich bei der Resorption von Fliissigkeiten in Geweben um Imbibition handle, d. h. um Capilljaritatsercheinungen. Bei der Fliissigskeitsaufnahme in Seifen handelt es sich aber um Losungsvorgange. Die dabei maasgebenden Krafte sind osmotische Drucke und nicht die bei capiEaren Vorgangen maasgebenden Oberiiachenspannungen . ) iH. J. Hambtogbr: Arch. f. (Anat. u.) PhysioL, 513 (1892); ibid., 153 (1893); Zeitschr. f. Biol., 35, 252 and 280 (1897), where references to his earlier papers are found. See also Arch. f. (Anat. u.) Physiol., 31 (1898). 2 0. VON Limbeck: Arch. f. exper. Path. u. Pharm., 35, 309 (1894). ' Gukber: Sitzberich. d. med. phys. Gesellsch., WUrzbxu-g, Feb. 25 (1895). ^C. Eijkman: Virohow's Arch. f. path. Anat., 143, 448 (1896), where references to his earlier papers wiU be found. * Ralph W. Webster: University of Chicago Decennial PubUcations, 10 (1900); cited from a reprint. 6E. Overton: Pfluger's Archiv, 92, 115, (1902). 112 CEDEMA AND NEPHRITIS tions. They weighed their muscles at arbitrary intervals of time, and drew their conclusions from these weighings — at times only one weighing. A moment's study of a few of the curves which accompany these paragraphs will show how wrong this is. (See Figs. 43 to 45.) To cite but one example, a muscle kept in any salt solution need not, and, in fact, usually does not, show a progressive increase or decrease in weight. It may at first show a very decided decrease and later an equally decided increase; or the reverse may be the case. If this fact is borne in mind, many of the statements made by these authors and not in har- mony with each other or with my own experimental results will find a ready explanation. We shall turn now to the conclusions to which I have been led from my own experiments, and see if in them we may not find an acceptable explanation of the apparently unattached and not easily accounted for facts observed by the previous workers in this field. My experiments were made with the hind legs of tree toads (Hyla) from which the skin had been removed, and with the gastrocnemius muscles of the frog (Rana) . The muscle preparations were carefully dried, weighed and placed in various solutions contained in lightly covered finger bowls. At various intervals they were removed from the solutions, carefully dried with filter paper, and weighed, and the amount of water they had lost or gained was calculated in per cent of the original weight of the muscle. From many such experiments the following conclusions of importance to the subject in hand were drawn. The conclusions are again lettered so as to permit ready com- parison with similarly lettered and corresponding conclusions reached in the study of the absorption of water by fibrin and by gelatin. (a) A muscle swells more in the solution of any acid than it does in pure water, but the amount of this swelling is greater in some acids than in others. Muscle swells most in a hydro- chloric acid solution, almost as much in a nitric acid solution of the same concentration, and less in acetic and sulphuric acids in the order named. Fig. 39 may serve as an illustration of this fact. The experiments upon which these curves are based were made with the hind legs of tree toads (Hyla) from which the skin had been removed.^ ' Maetin H. Fischer: Pfliiger's Aichiv, 124, 69 (1908). ABSORPTION, SECRETION— CELLS AND TISSUES 113 An important relationship exists between the concentration of the acid employed and the amount that the muscle swells. This is readily apparent in Fig. 40 and Table XLIV, which contains the experimental findings from which the curves were constructed. In this series of experiments the gastrocnemius muscles of frogs (Rana) were used. There is first to be noted an increase in the swelling with every increase in the concentra- 20 „ 25 Hours Figure 39 tion of the acid. But after a time a point is reached beyond which a further increase in concentration is followed by a dimin- ished absorption of water. This fact has its analogue in the absorption of water by fibrin or gelatin in acid solutions of various concentrations. Table XLIV is given in detail to show by what means were obtained all the data upon which conclusions in this section are based. The first figure in each of the columns indicates the original weight of the muscle. After each of the weighings 114 CEDEMA AND NEPHRITIS there is given, in parentheses, the gain in weight, expressed in per cent of the original weight of the muscle. Figure 40. (6) It is somewhat difScult to say what is the effect of alkalies on the absorption of water by muscle. The statement is un- ABSORPTION, SECRETION— CELLS AND TISSUES 115 ™3 O t> g+ffl aSq si* d m q " + M — W . oO ^ — o t)-iO>0OiC coinoiOO«>r-iTfHi-((N ■^COl>-O0Oi'-'C0Tt010OJOlOJ00000000 ooooooooo 10 cc. n/10 HCl +100 CO. H2O. % (0) C+ 48.8) (+ 87.5)' (+ 96.6) (+112.7) (+12r ') (+111.8) (+109.5) (+102.6) Mlv'l t-iO»OOOMOOOOO OiOosO'-i'N'-H'-'O 9 cc. n/10 HCl+101 cc. H2O. % 0.686(0) 1.072 (+ 56.2) 1.435 (+109.1) 1.505 (+119.4) 1.586 (+131.1) 1.582 (+129.9) 1.458 (+112.5) 1.455 '+112.1) 1.370(+ 99.7) (f) IX 8 cc. n/10 HCl +102 cc. H2O. r^t--o=ciFHiooicc ^ iOO'NCCCC!N>-i.H 3++++++++ COOOOiOOiOiMiCO 'NCCOOS'-tt--CCOO'^ d'H^r-ir-I.-^r-l.-ir-I -> S cc. n/10 HCl +105 cc. H2O. S" CC ^ 01 01 So "5 CCOCOt-^Ol^COl^ b? ■<*'iC00'-H'>»OI>CC S++++++++ £^ i-i01>(N0000"3 TjHO00CCC^M»O«« U3aiCCiOI>l>CD'^N od.-5^^rHTHF^^ 6 " >Sq % 491 (0) 815 (+ 65.9) 150 C +134. 2) 313 (+167.4) 473 (+200.0) 578 (+221.3) 420 (+189.6) 220 (+148.4) 095 (+123.0) S> OO^i-i,-(,H,-(^,-( .2 .2 gg W M CO -^ r>. 116 OEDEMA AND NEPHRITIS questionably true that muscle swells more in the solution of any alkali than in water. There seems to be a great difference, how- ever, both in the swelling of tree toad legs from which the skin has been removed and of the gastrocnemius muscles of frog with the season. In my original experiments with tree toads I got a decidedly greater swelling in dilute alkaline solutions than in water. In later experiments (December 11, 1908) with the gastrocnemius muscles of winter frogs (Rana) this difference was not so marked. I append Tables XLV and XL VI to illus- trate this point. In explanation of these results it should be noted that the amount of swelling in pure water runs unusually high. As, to my mind, this is brought about chiefly through the pro- duction of acid within the muscles, the high water absorption values indicate an unusually large production of acid (starvation acidosis in winter frogs?). When such muscles are placed in alkaline solutions, the alkali combines with the acid, and the salt formed by the union inhibits the swelling. (See paragraph c, below.) LoEB states that the gastrocnemius muscles of frogs swell more in the solution of an alkali than in acid solutions of the same normality. The few weighings that he gives are not suffi- cient to prove this, for only serial weighings can tell us whether the maximal swelling in a muscle has been attained, is being ap- proximated, or has been passed. The experiments just outlined indicate that, if the opposite is not true, the question is at least still an open one. TABLE XLV Gastrocnemius Muscles op the Frog Hours in the solution. 110 CO. H2O. 5 cc. n/10 NaOH+105cc. H2O. 10 cc. n/10 NaOH+100 cc. H2O. 1.05 3.25 4.45 8.45 20.50 34.10 45.60 70.20 % 0.984 (0) 1.397 (+41.9) 1.610 (+63.6') 1.659 (+68.5) 1.680 (+70.7) 1.626 (+65.2) 1.540 (+56.5) 1.630 (+55.4) 1.610 (+53.4) % 0..571 (0) 0.858 (+50.2) 0.995 (+56.7) 0.881 (+64.3) 0.854 (+49.5) 0.842 (+47.4) 0.838 (+45.9) 0.850 (+48.1) 0.885 (+64 9) 1 % 571 (0) 920 (+ 61.1) 104 (+ 93.3) 144 (+100.3) 141 (+ 99. S) 108 (+ 94.0) 125 (+ 97.0) 142 (+100.0) 135 (+ 98.7) ABSORPTION, SECRETION— CELLS AND TISSUES 117 (c) The addition of any salt to the solution of an acid decreases the amount that a muscle will swell in that solution, and the higher the concentration of the salt the greater is the amount of this inhibition. Fig. 41 illustrates this fact. The curve marked HCl was obtained by immersing a gastrocnemius muscle in a solution of hydrochloric acid, made by adding 10 cc. n/10 hydro- 20, HCl + }^mNaCl 10 15 20 25 Hours 30 35 40 ■45 Figure 41. chloric acid to 100 cc. of water. The three remaining curves show the changes in weight suffered by muscles immersed in solutions made by adding the same amount of acid to 100 cc, respectively, of m/8, m/4 or m/2 solutions of sodium chlorid. As plainly evident, the action of the hydrochloric acid is entirely inhibited, so far as the absorption of water is concerned when the last-named concentration of sodium chlorid is employed. 118 (EDEMA AND NEPHRITIS > 1-1 X! m i- o o « W O ;z: o o BS Eh CC ■C'-«'-«(N S+++++++++ 0i(NOO0S00TtC IC 10 10 CO -* Tji Tti S+++++++++ "*(NUDO(NOiCmOO or-0(NOt^i>osOTt< r-o.-(fH.-H0050ioo OOfMf-i.-H^OO^^ 1/2 cc. n/10 NaOH +109-1/2 cc. H2O. ■^"^ 00 00^io5J"f^t^ gs COOlOlOTjiCCCOCOCO S+++++++++ (OOOOtNOOOO'CiO Of-OiOt^-^iOr-cDCD t^0i0'~'000i0i0101 00r-..-<.-(^0000 1/4 00. n/10 NaOH+109-3/4cc. H2O. <0 i~iOOOOMf^O>^S" ^0 osr-osr-'<**ooooo ftv co»o»raioiCTjO CO(N»Ot^iO(rOCOCO(N'-i I>0^-ti-H-HTHOOOO d r-I ^' ^' ^' ^ ^ rt' ^' ^' OmCNcDro^OOOOO ^0 tOCOcOOiiOO'-'CTiOl e^ COlCt>lOlOTtiTtHCCCO S+++++++++ OOTfiOOOOiCOOO OOcDOCR'XiCO'iJ'cOCO 00i-i(NcOci SOOit-MCOiCOOCDiC gS; T(COCOiO+o gB« z COiOO>COi-i00cDi-HOS ^c, CO.-ltDt>1000rHl>00 g^ irir-cDcDcocDcocooo 3+++++++++ lOOiOOCMOOOOO ■^0>C0'-t'-iOiMC3iiMC0 loceoiojoiosoioioo 666666661-^^ eijlq gtrjW COOOICOCM'^iOn""? S+++++++++ r^iOiOOlNOOOOCO Tt<0JC0NOOCJi-*O«3 lOt-COOOQOOOt^XOOOO 6666666666 23 z iccou?o«^ooP^ - (cr^(N.-Hh-t-«t^oOr-* fe? ■* '*! -^ tH CO IC -^ -* "C OOOONiOOiOi^'nOiO •C'-<(NOlQ0t>'*CNC0"* W30000I>-I>l>Q0000OW 0000000000 in r^ . ^+§ z S'somP^mPnc? fc? (N00»Owt-iMTHio»O '=' 10 10 10 »0 -^ --^H 10 10 >0 5.+++++++++ OW5C»00»0»0000 ooco.-ioce»0(Noooo lOoocnciooooQOMoiai 0000000000 0) -^ . n m OOiCOOOOiOO O»0C0»0C0C0i0Tt2 ■ 10 15 20 Hours Figure 51. 25 35 effective in this regard. As in the case of fibrin or gelatin, the effect of any salt seems to be made up of the sum of the effects of its constituent radicals. Fig. 52 permits comparison of the action of different basic radicals. The eyes burst in both the pure hydrochloric acid solutions, but in none of those to which a salt had been added. The solutions containing salt were made by adding 20 cc. n/10 normal hydrochloric acid to 200 cc. m/6 solutions of the different chlorids. The bases arrange themselves in about the following order, in which that least effective in re- ABSORPTION, SECRETION— CELLS AND TISSUES 133 ducing the amount of swelling in an acid solution is placed highest in the series, and first in each group : 1. Lithium, Ammonium, Potassium, Sodium. 2. Barium, Strontium, Magnesium, Calcium, Copper (ic). 3. Iron (ic). We have small difficulty in discovering in this table the same grouping familiar to us from our discussion of the swelling of fibrin and gelatin. 134 (EDEMA AND NEPHEITIS In Fig. 53 are shown the effects of salts having different acid radicals. In each case 20 cc. n/10 hydrochloric acid are again added to 200 cc. of m/6 solution of the appropriate sodium salt. Figure 53. -'^■. Their order is as follows when that least effective in reducing the amount of swelling in an acid solution is placed first:' 1. Chlorid, Bromid, Nitrate, Acetate. 2. Phosphate, Sulphate, Tartrate. This table also is to all intents and purposes identical with that given in the discussion of water absorption by fibrin and gelatin. ABSORPTION, SECRETION-CELLS AND TISSUES 135 (e and /) Non-electrolytes do not share with electrolytes their marked power of decreasing the absorption of water by the eye. Fig. 54 shows this better than many words. The curves he very closely together, and in spite of the fact that the various non-electrolytes are present in amounts which are are osmotically more than equivalent to the powerfully acting electrolytes (20 cc. n/10 HC1-F200 cc. m/3 solution of the non- electroljrte), not one of the eyes has been kept from bursting. ig) The absorption and secretion of water by the eye is largely a reversible proc- ess. This is indicated in Fig. 55. Curve A shows how an eye which has reached the bursting point in a pure hydrochloric acid solution suffers a prompt loss of water if taken out of this solution and transferred to an equally concentrated one containing calcium chlorid in addition. Curve B shows the reverse of this. An eye which has gained but little weight in pure water is transferred to a dilute hydrochloric acid solution. Immediately the absorption of water is hastened, and becomes so great that the eye bursts. The eye, also, suffers somewhat permanently from every condition through which it has passed. Once, for example, an eye has been in an acid containinh a salt, it does not subse- quently swell as much in a pure acid solution (in the time allowed in these experiments) as it would have done had it been placed here directly. Figure 54. 136 (EDEMA AND NEPHRITIS 3. The Analogy between the Absorption of Water by Certain Protein Colloids and by Nervous Tissue The complete analogy between the absorption of water by certain protein colloids and by muscle and the eye as outlined in the preceding sections seemed to me to justify the conclusion that |the colloids and their state are chiefly if not entirely re- sponsible for the amount of water held by any cell, tissue or organ under different circumstances. Since this conclusion was first voiced, it has found generous acceptance and support from a ABSOKPTION, SECRETION— CELLS AND TISSUES 137 number of investigators.^ But it has also met with opposition. Here we must consider the objections of J. Baueb,^ who believes that in nervous tissues water absorption is not, in the main, a function of the protein colloids found in them. Before pointing out the obvious errors in Bauer's experiments and conclusions, let us first look at the following experiments made by Marian 0. Hooker '"^ and me, which, contrary to Bauer's view, show that the absorption and secretion of water by nervous tissues {brain and spinal cord) is entirely analogous to the absorption and secretion of water by such protein colloids as fibrin or gelatin/^ To obtain perfectly fresh nervous tissue in as unchanged a condition as possible, we used normal rabbits which had been on a generous mixed diet, killed them by a gentle blow behind the ears and then rapidly dissected out the brain and spinal cord. The dura and arachnoid membranes were removed and the pia was peeled off as well as possible. The nervous tissues were cut into pieces of approximateli^ the same size. In each series of experiments the pieces used were always taken from the same animal. This is necessary, for comparatively trivial things influence the initial state of the nervous tissue. If an animal is chased about its cage just before being used, or is ill, its nervous tissues show a different capacity for absorbing water, due, in the main, we think, to differences in their initial acid content, than when such things have not happened. In the same way the stale tissues from an animal dead some hours or days show different absorption curves (because of a higher initial acid content) than fresh ones. After weighing, the pieces of tissue were introduced into the different solutions. At various times they were taken out, re- iSee for example: I. Thaubb: Pfluger's Arch., 140, 109 (1911). K. Gedkoiz: Russ. Journ. f. exp. Landwirtsch., 11, 66 (1910). E. Przibeam: KoUoidchem. Beihefte, 2, 1 (1910). H. Klosb: Arch. f. Kinderheilk., 55, 43 (1910). H. Bbchhold: Kolloide in Biologie und Medizin, Dresden (1912). H. Klosb and H. Voigt: Beitr. z. klin. Chirurgie, 69 (1910). O. PoTZL and A. Schuiler: Zetschr. f. d. ges. Neurologie, 3 (1910). Rudolf Aenold: KoUoidchem. Beihefte, 5, 411 (1914). 2 J. Bauer: Arb. a. d. neurol. Inst. d. Wiener Univ., 19, 87 (1911); Kol- loid Zeitschrift, 9, 112 (1911). , . , „ - ^ i„ ' Marian O. Hooker and Martin H. Fischer: KoUoid Zeitschr., W, 283 (1912), where detailed weighings may be found. 4 For further evidence in this direction and independent criticism of Bauer's conclusions see Raphael Ed. Liesegang: Ergebnisse d. Neurol, u. Psych., 2, 157 (1912). 138 (EDEMA AND NEPHRITIS weighed and the increase or loss calculated in percentage of their original weight. The results are shown in the accom- panying curves, which were made by plotting time on the hori- zontal and changes in weight on the vertical. As they are all drawn to the same scale they may be compared directly with each other. Hours 12 18 24 Figure 56. 30 Hours 6 12 IS Figure 57. To indicate their complete analogy the following paragraphs on the absorption of water by nervous tissue are lettered to correspond with the similarly lettered paragraphs in the sections on fibrin and gelatin. (a and b) When nervous tissue (brain) is placed in distilled water it takes this up (gains in weight). We shall at once in- terpret this by saying that after removal from the body the tissue develops acid and this increases the capacity of the brain colloids for holding water. If the brain is placed in a dilute acid ABSORPTION, SECRETION— CELLS AND TISSUES 139 instead of in water it swells decidedly more. With every in- crease in the concentration of the acid the amount of water absorption is increased. These facts are brought out in the curves of Fig. 56. But the increased swelling with increase in concentration of acid continues only up to a certain point, when every further addition of acid only reduces the amount of water absorbed. This is shown in Fig. 57. The curves of Figs. 56 and 57 are based respectively upon the experimental data contained in Tables XLVII and XL VIII: TABLE XLVII Adult Rabbit Brain Hours in the 100 CO. 100 cc. Solution. 1/10000 n HCl. 3/10000 n HCI. % % % 1.228 (0) 1.142 (0) 0.865 (0) 1.20 1.825 (+ 49) 1.682 (+ 47) 1.346 (+ 55) 4.20 2.336 (+ 90) 2.247 (+ 97) 1.765 (+104) 22.35 3.775 (+208) 3.850 (+237) 1.775 (+221) 27.00 4.185 (+240) 3.950 (+246) 3.110 (+260) TABLE XLVIII Adult Rabbit Brain Houra in the Solution. 100 cc. 2/1000 n HCI. 100 cc. 1/1000 n HCI. 100 cc. 5/10000 n HCI. 100 cc. HjO. 1.20 4.20 22.40 27.00 % 0.630(0) 0.785 (+25) 0.866 (+38) 1.030 (+64) 1.090. (+73) % 0.650 fO) 0.926 (+43) 1.066 (+64) 1.205 (+85) 1.240 (+91) % 0.663 (0) 1.005 (+ 52) 1.210 (+ 83) 1.575 (+138) 1.680 (+153) % 1.228 (0) 1.825 (+ 49) 2.336 (+ 90) 3.775 (+208) 4.185 (+240) (c) The amount of acid developed in nervous tissue after removal from the body is so near that leading to a maximum of water absorption that none needs to be added from the outside in our further experiments. If we place the pieces of tissue in pure solutions of various kinds we shall in reality be observing the effect of these plus that of a certain amount of acid (produced by the tissue itself). Fig. 58 (as well as Figs. 59, 60, 61, 62, 63 and 64) shows that the addition of any electrolyte to the (acid) solution in which 140 CEDEMA AND NEPHRITIS brain tissue is swelling reduces the amount of water absorbed. This reduction is the greater the higher the concentration of the added electrolyte. Fig. 59 shows the same to be true for spinal 80 60 40 20 + !« ^^ NaCl- Solutions 3^ ^ ^^ ___^ /^ -^^ X X _-^ \^^ 1 — ■ 1 Hours 6 12 18 Figure 58. 24 cord. The experimental data contained in Tables XLIX and L form the basis of Figs. 58 and 59. TABLE XLIX Young Rabbit Brain Hours in the 120 CO. 120 CO. 120 CO. 120 00. Solution. m/2 NaCl in/4 NaCl m/a NaCl ra/8 NaCl % % % % 0.808(0) 0.860 (0) 0.915 (0) 0.982 (0) 2.45 0.822 (+ 2) 0.945 (+10) 1.000 (+ 9) 1.087 f+U) S.45 0.855 (+ 6) 0.980 (+14) 1.075 (+17) 1.212 (+23) 23.15 0.977 (+21) 1 . 141 ( +33) 1.267 (+38) 1.605 (+63) 29.25 1.005 (+24) 1.178 (+37) 1.410 (+56) 1.715 (+75) TABLE L Spinal Cord of Young Rabbit (Same Animal aa in Table XLIX) Hours in the Solution. 120 CO. m/2 NaCl 120 cc. . ro/4 NaCl 120 CO. m/8 NaCl 3.00 5.30 23.20 29.25 % 0.072(0) 0.057 (- 2) O.OSO (+11) 0.086 (+20) 0.090 (+26) % 0.081 (0) 0.092 (+13) 0.100 (+23) 0.112 (+38) 0.120 (+48) % 0.142 (0) 0.192 (+35) 0.200 (+41) 0.230 (+62) 0.242 (+70) ABSORPTION, SECRETION— CELLS AND TISSUES 141 (d) Equimolar concentrations of different salts reduce in very unequal degree the swelling of nervous tissue (in any acid solution). In Figs. 60, 61 and 62 some salts are compared having the same basic, but different acid radicals. In Fig. 60 the sodium Hours 6 Figure 60. salts are seen to arrange themselves in the following familiar order when that least effective in preventing swelling is given first: Chlorid, Nitrate, Acetate, Phosphate, Sulphate. In Fig. 61 the potassium salts assume the following order: Chlorid, Acetate, Citrate. Or in Fig. 62: Bromid, lodid, Sulphocyanate, Nitrate. The dehydrating action of salts having the same acid radical but different bases are compared in Figs. 63 and 64. In the 142 OEDEMA AND NEPHRITIS chlorid series when the salt least effective in inhibiting swelling is given first, we see the order: Ammonium, Sodium, Potassium, Strontium, Barium, Magnesium, Copper, or in the nitrate series similarly arranged: Ammonium, Potassium, Sodium, Strontium, Magnesium, Barium, Calcium, Iron. The dehydrating action of these salts on nervous tissue is prac- tically identical, therefore, with their dehydrating action on fibrin or gelatin. Houi-S 6 13 18 Figure 61. ^1^ 240 y 200 /HjO 180 - / 120 ■ / 80 / ^^S-^ — ^ 40 / /^^'^^^^^-■i;^--^^^^^^^^^^^^^ ni^i r^ 1 Hours 6 12 18 24 30 Figure 62. Figs. 60, 61, 62, 63 and 64 are based respectively upon the data contained in Tables LI, LII, LIII, LIV, LV. ABSORPTION, SECRETION— CELLS AND TISSUES 143 Hours 6 TABLE LI Brain op Rabbit Hours 100 en. 100 cc. 100 cc. 100 cc. 100 00. in the m/6 -m/6 m/6 m/6 m/6 Solution. Na2S04. Na2HP04. NaCzHiOz. NaNOi. NaCl. . % % % % % % 0.766(0) 0.761 (0) 1.220 (0) 1.200 (0) 0.752 (0) 1 . 120 (0) 2,00 0.766 (0) 0.741 (- 1) 1.251 (+ 3} 1.291 (+ 8) 0.799 (+ 3) 1.595 (+ 42) 3.00 0.752 (- 2) 0.755 {+ 1) 1 . 270 ( + 4) 1.310 (+10) 0.837 (+11) 1.780 (+ 59) 20.15 0.771 (+ 1) 0.827 (+10) 1.565 (+29) 1.639 (+36) 1.063 (+41) 2.821 (+152) 24.10 0.781 (+ 2) 0.826 (+10) 1.580 (+30) 1.640 (+37) 1.087 (+44) 2.814 (+151) 44.35 0.820 (+.7) 0.910 (+21) 1.710 (+40) 1.825 (+52) 1.202 (+59) 3.195 (+185) 68.15 0.855 (+10) 0.966 (+28) 1.825 (+50) 1.975 (+65) 1.265 (+68) 3.223 ( + 188) 144 (EDEMA AND NEPHRITIS ABSORPTION, SEGfiETION— CELLS AND TISSUES 145 TABLE LII Brain of Rabbit Houra in the 120 CO. m/6 120 CO. m/6 120 cc. m/6 Solution. Calcium citrate. Calcium acetate. Calcium chlorid. % % % % 1.086(0) 0.685(0) 1.072(0) 0.623(0) 2.30 0.992 (-8) 0.770(+13) 1.308(+ 22) 1.100(+ 78) 5.35 0.995 (-8) 0.815(+20) 1.410(+ 31) 1.450 (+133) 23.50 1.060 (-2) 1.033 (+51) 1.753(+ 63) 2. 075 (+233) 28.25 1.060 (-2) 1.050 (+54) 1.795(+ 68) 2. 165 (+247) 47.30 1 . 120 ( +3) 1.160 (+70) 2.005(+ 87) 1.858i(+198) 71.20 1.185 (+9) 1.245 (+82) 2.170(+102) 1.690i(+171) 1 Going into solution. TABLE LIII Brain op Rabbit Houra in 120 CO. 120 CO. 120 CO. 120 cc. 120 cc. ao. the solu- m/6 m/6 m/6 m/6 tion. KNOs. KCNS. KI. KBr. % % % % % 1.241(0) 1.105(0) 0.712(0) 0.773(0) 0.623(0) 2.30 1.388 (+12) 1.286 (+17) 0.863 (+21) 0.916 (+ 18) 1.100 (+ 78) 5.35 1.510 (+22) 1.363 (+23) 0.978 (+37) 1.047 (+ 35) 1.450 (+133) 23.30 1.822 (+47) 1.693 (+53) 1.235 (+73) 1.350 (+ 74) 2.075 (+233) 28.25 1.973 (+59) 1.713 (+55) 1.200 (+68) 1.395 (+ 80) 2.165 (+247) 47.30 2. 125 (+71) 1.867 (+69) 1.285 (+80) 1.475(+ 91) 1.8581 (+198) 71.20 2.245 (+81) 2. 070 (+88) 1.375 (+93) 1.590 (+106) 1.6901 (+171) 1 Going into solution. TABLE LIV Adult Rabbit Brain Hours in the 120 cc. m/6 120 cc. m/6 120 cc. m/6 120 cc. m/6 Solution. CuCk MgCla BaCh SrCl! % % % % 0.791(0) 0.685(0) 1.010(0) 0.878(0) 1.30 0.775 (-2) 0.692(+ 1) 1.010(0) 0.910 (+ 4) 18.15 0.753 (-5) 0.723 (+ 6) 1.155 (+14) 1.005 (+14) 26.00 0.747 (-6) 0.730 (+ 7) 1.217 (+20) 1.083 (+23) 47.15 0.749 (-5) 0.810 (+18) 1.297 (+28) 1 . 158 ( +32) 67.60 0.765 (-3) 0.887 (+30) 1.450 (+43) 1.230 (+40) Hours in the Solution. 120 cc. m/6 KCl 120 cc. m/6 NaCl 120 cc. m/6 - NH4CI 120 cc. H2O. 1.30 18.15 26.00 47.15 57.60 % 0.998(0) 1.023 (+ 3) 1.337 (+34) 1.430 .(+43) 1.600 (+60) 1.832 (+84) % 0.558(0) 0.625 (+12) 0.793 (+42) 0.815 (+46) 0.910 (+63) 1.017 (+82) % 0.788(0) 0.883 (+ 12) 1.385 (+ 76) 1.365 (+ 73) 1.530(+ 94) 1.710(+117) % 0.526(0) 0.875 (+ 63) 1.708 (+21S) 1.945 (+263) 1.8721 (+250) 1 . 642 1 ( +206) 1 Going into solution. 146 CEDEMA AND NEPHRITIS TABLE LV Adult Rabbit Brain Hours in the 120 CO. m/6 120 00. m/6 120 CO. m/6 120 CO. m/6 120 00. m/6 solution. Fe(N0s)3 Ca(N03)z Ba(N03)!. Sr(N03)!. Mg{NOa)!. % % % % % 0.575(0) 0.785(0) 0.535(0) 0.820(0) 0.742(0) 3.45 0.495(-14) 0.842(+ 8) 0.575(+ 7) 0.867 (+ 5) 0.765 (+ 3) 22.45 0.362 (-37) 0.955 (+22) 0.660 (+23) 1.035 (+26) 0.887 (+20) 27.45 0.332 (-42) 0.960 (+22) 0.673 (+26) 1.105 (+35) 0.965 (+30) 46.15 0.278 (-51) 1.020 (+30) 0.705 (+31) 1.158 (+41) 1.072 (+45) 70.15 0.240 (-58) 1.060 (+35) 0.763 (+43) 1 . 283 ( +58) 1.215 (+64) Hours in the 120 00. m/6 120 CO. m/6 120 CO. m/6 120 CO. solution. NaN03. KNO3. NHjNOs. H2O. % % % % 0.610(0) 0.758(0) 0.828(0) 0.370(0) 3.45 0.708 (+16) 0.975 (+ 29) 1.155 (+ 39) 0.810(+117) 22.45 0.970 (+60) 1.2S8(+ 70) 1.465 (+ 79) 1.505 (+306) 27.45 1.005 (+65) 1.350(+ 75) 1.525 (+ 84) 1.660 (+322) 46.15 1 . 135 ( +86) 1.485 (+ 95) 1.742 (+110) 1.745 (+372) 70.15 1.205 (+97) 1.563 (+105) 1.7321 (+109) 1.910 (+420) 1 Going into solution. (eand/) At the same concentration the non-electrolytes are far less powerful in reducing the swelling of nervous tissue (in an acid medium) than the electrolytes. Fig. 65 and Table LVI show this. Even though the three alcohols were present in a concentration osmotically equal to or higher than that of the electrolytes used in the experiments previously described, a decrease in the amount of swelling is either not evident at all or but slight. As both the figure and the table indicate, urea seems actually to favor the absorption of water. TABLE LVI Adult Rabbit Brain Hours in the solution. 4.00 23.45 28.00 46.30 70.30 120 CO. m/3 Glycerin. % 0.832(0) 1.153(+ 39) 1.902 (+130) 2. 188 (+163) 2. 595 (+210) 2.904 (+249) 120 cc. m/3 Urea. % 0.828 (0) 1.560 (+ 81) 2.347 (+183) 2. 3151 (+179) 2.510 (+203) 120 cc. HjO. or 1.227 (0) 2.200 (+ 79) 3.475 (+183) 3.840 (+213) 4.345 (+255) 4 . 2801 ( +248) 120 00. m/3 Methyl Alcohol. % 0.905(0) 1.572 (+ 73) 2.605 (+189) 2. 830 (+213) 3.200 (+253) 3. 288 (+263) 120 CO. m/3 Ethyl Alcohol. % 0.978(0) 1.782(+ 82) 2. 982 (+205) 3. 252 (+232) 3 . 695 ( +277) 3.935 (+303) 1 Going into solution. ABSORPTION, SECRETION— CELLS AND TISSUES 147 280 240 - 200 160 120 ■ 80 40 _ - ^^..-X"^ ^^^^^^*^ Ethyl Alcohol /Wate Methyl Alcohol X^ -Urea ^^-^^ //^ y ^^ ///^ Grlycerin ^ ; f/^ f 1 1 Hours 6 18 24 30 Figure 65. 42 TABLE LVII Adult Rabbit Brain Hours in the solution. ■ 120 CO. H2O 120 cc. m/6 NaCl. 120 cc. m/6 NaCl 120 cc. H2O. % 1.150(0) 2.030 (+76) 2.050 (+78) Transferred % % 1 . 100 (0) 1.260 (+15) 1.305 (+19) Transferred % 2.25 3 25 4 20 1.670 (+45) 1.630 (+41) 1.845 (+59) Transferred 1.480(+ 35) 1.760(+ 60) 6.00 24 00 2 830 (+157) 1.970(+ 71) 2.090 (+ 82) 2.467 (+114) Transferred 25.45 2. 075 (+89) 1.920 (+75) 1.700 (+55) 29 35 47.45 148 (EDEMA AND NEPHRITIS {g) The absorption and secretion of water by nervous tissue represents in large part a reversible process. This is brought out 160 120 80 40 + i« OHouis 6 /" HjO/ WaCl ^^^ A / k ■^'IhjO /hjOV/ NaCT_,,.-- / ^^^^ J /NaCl ' 1 , 1 1 1 18 2i FlGtJKB 66. 42 48 in Figs. 66 and 67 and Tables LVII and LVIII, upon which these drawings are based. If nervous tissue is placed in a dilute acid 120 ^ ^^^HCl ^ 80 - h /HCl / V / HCl+Naei,_^^^ ]( 7 / ^' — - ■ 40 HCH-NaCa + !« \ I /^Cl+NaCl 1. 1 J 1 1 1 1 ' Hours 6 12 18 24 30 36 42 Figure 67. 48 or in water (which amounts to placing it in a dilute acid) it swells. If, after any desired degree of swelling has been attained, the tissue is transferred to an equally concentrated acid containing ABSORPTION, SECRETION— CELLS AND TISSUES 149 a salt, the swelling ceases and a loss of water begins. If now the tissue is returned to the pure acid or to water, rapid absorp- tion again occurs. A reverse set of facts and curves is obtained if the tissue is first placed in acid plus salt, then in pure acid and then again in acid plus salt, as is also evident in Figs. 66 and 67. TABLE LVIII Adult Rabbit Brain Hours in the solution. 120 CO. m/6 NaCl+0.2co. n/10 HCl. 120 CO. H2O +0.2 00. n/10 HCl. 120 00. H2O +0.2 CO. n/10 HCl. 120 CO. m/6 NaCl+0.2cc. n/10 HCl. 2.30 % 1.517 (0) 1.730(14+) 1.765(16+) Transferred % % 1.280(0) 2. 180 (+70) 2.370 (+85) Transferred % 3 30 4 20 2.015(+ 33) 2.370(+ 56) 3.295 (+117) Transferred 1 . 950 ( +52) 6.00 1.890 (+48) 23 45 2 070 ( +62) 2.333 (+54) 2.205 (+45) 2.247 (+48) 25.45 2.217 (+ 74) 2.430 (+ 90) 2. 993 (+134) 29 45 47.45 These experiments dispose, we think, of Bauer's objections which have been taken up in detail elsewhere.'- His conclusion that acids only dehydrate nervous tissue was reached because he chose his acid concentrations too high (beyond those optimal for brain swelling), and because he worked with stale tissues (six to twenty-four hours old) so rich in postmortem acids that a further addition of acid from the outside could only decrease their power of swelling. In concluding these paragraphs, attention may be directed to Fig. 68, which shows how enormously brain tissue can swell. The figure shows the two halves of a young rabbit's brain and spinal cord split longitudinally. The half marked a was care- fully protected from evaporation, that marked b was kept in a n/ 10,000 lactic acid solution. At the end of twenty-four hours ' Marian O. Hooeeb and Martin H. Fischer: Kolloid-Zeitschr., 10, 292 (1912). 150 CEDEMA AND NEPHRITIS it had gained twice its original weiglit in water. Fatal brain cedemas show an increase in weight of but few per cent. How Figure GS. easily such is accounted for on the basis of colloid swelling can readily be seen. ABSORPTION, SECRETION— CELLS AND TISSUES 151 IV THE BIOLOGICAL SIGNIFICANCE OF THE ANALOGY BE- TWEEN THE ABSORPTION OF WATER BY CERTAIN PROTEIN COLLOIDS AND THE ABSORPTION OF WATER BY DIFFERENT TISSUES 1. Introductory Remarks The complete analogy from both a quantitative and a quali- tative point of view between the absorption of water by certain colloids and by the most different types of tissues seems to me to warrant the conclusion that the colloids and their state are the main factors concerned in deter^nining the amount of water held by a cell, a tissue, an organ or a whole individual under different physio- logical or pathological circumstances. That they might be of some importance in this regard has occurred to several observers, but careful study of their papers shows that for the most part they dismissed the thought with little more than mere reference to its possible role, or the added remark that its significance in the general problem could not be great. Interestingly enough W. Pfeffee,^ who worked so ear- nestly for the establishment of the importance of osmotic pressure as the great regulator of the water content of cells, seems to have been the first to regard the pressure of swelling (Quellungsdruck) as of use in explaining various exceptions to the laws of osmotic pressure as studied in botanical material. More recently Franz HoFMEiSTEK^ developed the same idea in his fundamental dis- cussions of the biological significance of the colloid state. Durig ^ expressed the behef that studies on the swelling of colloids might help to explain the exceptions to the laws of osmotic pressure noted in his experiments on the absorption of water by frogs in various solutions. Rudolf Hober* and E. Overton^ also ' W. Pfbffer: Pflanzen Physiologie, Leipzig, 1, 116 (1897); see also the first edition of 1881, 1, 26 to 29. 2 F. Kofmeister: Archiv f. exp. Path. u. Pharm., 28, 210 (1891). « Durig: Pfluger's Archiv, 85, 401 (1901). * R. Hober: Physikalische Chemie d. Zelle u. d. Gewebe, Second Edition, 61, 62 and 70, Leipzig (1906); Koranyi-Richtbr's Physikalische Chemie u. Medizin, 1, 294, Leipzig (1907); Hober's latest writings condemn the colloid-chemical view of water-absorption and my support of it entirely. Biol. Zentralbl., 31, 575 (1911). 'E. Overton: Nagel's Handbuch der Physiologie, 2, 744, Braunschweig (1906), where references to his earlier papers will be found. 152 (EDEMA AND NEPHRITIS considered the subject. Both, however, lay greatest stress on the osmotic conception of water absorption by cells, especially as modified by the belief that the osmotic membrane about cells is fat-Uke (Hpoid) in character. The r61e of the colloids is by both these authors not considered the fundamental factor in absorption, but is simply pointed to as one useful in explaining some of the many exceptions found to exist between the actual and the theoretical behavior of cells when these are regarded as osmotic systems. Much the same position is taken by H. J. Hamburger.^ The colloids as a factor in the regulation of the water content of organs have also been discussed by Wolfgang OsTWALD 2 and Wolfgang Paitli.^ Pauli pointed out that the swelling of red and white blood corpuscles in solutions of a dilute acid is not unlike the swelling of certain colloids, but in his discussion of the swelling of muscle he decided against this being essentially a colloid phenomenon because the analogy between the swelling of muscle and the swelling of certain colloids was not sufficiently close. His unfortunate conclusion was due to the fact that he compared the careful observations at hand on the swelling of gelatin with a group of inadequate observations on the swelling of muscle. The experiments just detailed on the analogy between the absorption of water by protein colloids and by muscle, eyes and nervous tissue constitute, so far as I know, the first attempt to estabHsh experimentally not only the quantitative, but the quali- tative importance of the colloids of the tissues in determining the amount of water held by them. From a quantitative stand- point, we found that certain (hydrophilic) emulsion colloids are readily able to absorb amounts of water which are larger than any we have to account for in protoplasm, and from a quali- tative standpoint, we found that the behavior of protoplasm toward various external conditions, so far as its water content is concerned, is no different from the behavior of some simple colloids toward the same external conditions. The individual cell which we call an ameba is in its pond water like a fibrin flake floating in a solution of some kind. Let us 1 H. J. Hamburger: Osmotischer Druck und lonenlehre. See especiallj- 3, 4 to 33, 50 to 54 and 108 to 144, Wiesbaden (1902 to 1904). ''Wolfgang Ostwald: Personal communication. ' Wolfgang Pauli: Ergebnisse der Physiologie, 6, 126 to 129 (1907). ABSORPTION, SECRETION— CELLS AND TISSUES 153 add acid to the medium in which the ameba lives and it swells as does the fibrin flake; or let us add salt, and both shrink. The aggregate of cells which we call tissues, or organs, behave simi- larly. Only for the words pond water we need to substitute blood or lymph, for this is the medium in which the cells of our body lie and from which they absorb as does the ameba an amount of water which is determined by the nature and the state of the colloids found in the cells. But to accept the ab- sorption of water by colloids as the most important factor in the absorption of water by the tissues is to arraign all the explanations which have thus far been given for the normal and abnormal variations in the amount of water held by protoplasm. We must, in consequence, study them for a moment in order to see if the conception which we have introduced of the variable capac- ity of the colloids for water merely adds to the forces already considered as active in protoplasm or whether the acceptance of the ideas here advanced necessitates a revision of our former beliefs. We can at once dismiss as purposeless all " explanations " which are of a " vitalistic," " neovitahstic " or " physiological " character — they are mere salves for the undiagnosed sore. Of clearly defined physical or physico-chemical explanations, two have assumed special prominence. The first of these originated with the plant physiologists and has been widely adopted by the animal physiologists. We may call it for short the osmotic theory of water absorption, and with it consider that modification of it which has come with the beUef that the semipermeable membrane assumed to surround cells is fat-like (lipoid) in char- acter. The second was born of the pathologists, and may be called the pressure theory. As we need to discuss it in detail later we dismiss it temporarily and here consider only the osmotic theory. 2. Criticism of the Osmotic Theory of Water Absorption by Protoplasm The osmotic theory has until rather recently represented the best attempt to analyze in physico-chemical terms the proc- esses of absorption and secretion by living cells. As espoused to-day by different workers, it has suffered strange and self- 154 cibEMA AND NEPHRITIS contradictory modifications from the original ' form in which it was put forward by Wilhelm Pfeffbe and Hugo de Vries, but the work of these two continues the foundation upon which the moderns have built, and if we would not get lost in termin- ology we must cojisider their work first. In order to account for the " turgor " (that is, the water content) of plant cells Pfeffer and db Vries held them to be surrounded by " osmotic " " membranes " of such character that while they gave passage to water they did not permit sub- stances' disscHveS* in this water to gb through. On such basis they explained the swelling of plant cells in water or dilute solu- tioi^ or their shrinkage in conegntrated ones by saying that in the former water is sucked into the cell, while in the latter it is sucked out. The moveiAent of water into or out of the cell occurs until the (osmotic) concentration of the dissolved sub- stances is the same- on' both sides of the membrane postulated to exist about the cells. But, in order to permit the water to move, this membrane must be impermeable to the dissolved substances (otherwise, of course, they would simply move from a region of higher concentration to one of lower concentration, and so osmotic differences could not come to pass, and con- sequently no movement of water). From these observations and theoretical views sprang the interest of the physical chemists in the whole problem of osmosis, and we see constructed the vasious " osmotic cells " that may be seen in any physico-chemical laboratory. Pfeffer was again pioneer here. He conceived the idea of supporting the " pre- cipitation membranes " that Moritz Traube had described before him in the walls of a poi^s pot in order to enable them to withstand pressure. Such- " precipitation membranes " may be made of many different substances, but the best and commonest is prepared by allowing the solution of a copper salt and the solution of a ferrocyanid to move into the wall of a porous pot from opposite sides. Where they meet a precipitate of copper ferrocyanid is deposited. The copper solution may now be washed out of the pot and the ferrocyanid rinsed off the outside. In the wall of the pot remains a " precipitation membrane " of copper ferrocyanid. This membrane allows water to pass through it easily, but it will not permit substances dissolved in this water to get through. The membrane is therefore " semi- ABSORPTION, SECRETION— CELLS AND TISSUES 155 permeable," and so identical with the " osmotic " membrane postulated by Pfeffer to surround the living cell. If the labora- tory cell is filled with a solution of any kind and placed in water, water is sucked into the cell; if it is placed instead into a stronger solution, water is sucked out. When equally concentrated solu- tions exist within and without, no movement of water occurs. As readily apparent, this behavior corresponds, when viewed super- ficially, with what Pfeffer and de Vries observed in living cells. Pfeffer made many osmotic measurements with his labora- tory cell, and on the basis of his observations Van't Hoff some years later formulated his famous laws. These are as follows: (1) At constant temperature the osmotic pressure of dilute solutions is proportional to the concentration of the dissolved particles. (2) At the same temperature equal volumes of all dilute solutions having the same osmotic pressure contain the same number of dissolved particles. (3) At constant volume the osmotic pressure of any solution increases as the absolute temperature. The work and conclusions of Van't Hoff and the physical chemists now became retroactive and the attempt was made to apply the laws of Van't Hoff not only to the biological facts that DE Veies and Pfeffer had furnished in their studies of plant cells, but also to those added by Hedin, Hamburger, Geyns, Koeppe, Loeb, Hober, Overton, Webster, etc., in their work with various animal cells. To this end the observa- tions made on plant and animal cells were compared with those made on the laboratory osmotic cell. When a solution of any electrolyte or non-electrolyte was found not to change the volume of hquid in a laboratory osmotic cell it was said to be " isosmotic " with its contents. Any series of solutions thus isosmotic with the contents of the osmotic cell were therefore isosmotic with each other (and therefore equally concentrated). Similarly, when a solution of any kind was found not to change the volume of a living cell it was said to be " isotonic " with the cell contents. In this way the solutions of many different substances were compared and their " isotonicity " determined. If the laws of osmotic pressure were active in living protoplasm it was to be expected that all "isotonic " solutions should prove to be " isosmotic." ? 156 (EDEMA AND NEPHRITIS When the first rough comparisons were made it was, in fact, thought that the isotonic solutions were isosmotic, but this conclusion could not stand the pressure of more careful and more numerous observations. To-day we may safely say that we do not know a single cell for which the laws of osmotic pressure are valid. We need not go into details to prove this. If cells obeyed the laws of osmotic pressure they ought always to have the same volume in isosmotic solutions of different substances. . Exceptions to this conclusion are the rule?^ This was first proved for the red blood corpuscles by Koeppb and is corroborated for muscle by the experiments of Loeb, Webster, Overton and my own, as described in the earlier pages of this book. Again, with every increase in the concentration of the medium surrounding a cell we should get a proportionate decrease in the volume of the cell. As a matter of fact, the shrinkage is always less than anticipated. KoEPPE found that red blood corpuscles always shrink less than expected when the concentration of the surrounding medium is raised. The same is true of muscle, hving frogs (Durig), enu- cleated eyes, nerve tissue and amputated frog legs. While in the osmotic cells of our physico-chemical laboratories electrolytes and non-electrolytes are equally active when the same number of dissolved particles are present in the unit volume, this is not the case in living cells. Generally speaking, the electrolytes are here active out of all proportion to the non-electrolytes. But aside from these physico-chemical facts which stand so immovably against any belief which sees in living cells a replica of the artificial osmotic cells of our laboratories, biological con- siderations make the whole conception impossible. To have the laws of osmotic pressure tenable for living cells we must have semipermeable membranes about them. Only as this is the case can changes in osmotic pressure become available for the movement of water into and out of cells. If now, for the sake of argument, we grant this assumption, then no dissolved substances can get into or out of the cell. Such a conception of the cell is impossible, for how under such circumstances could it get its necessary food, or how could it rid itself of its various metabolic products? Both processes are absolutely indispensable for the continuation of life. To get around the difficulty various ob- servers have made these osmotic membranes permeable to some or many dissolved substances. But the moment we grant this. ABSORPTION, SECRETION— CELLS AND TISSUES 157 then the dissolved substances can diffuse from regions of higher to regions of lower concentration, and so differences in osmotic pressure are equalized and no forces remain available for the movement of water. The adherents to the view that " osmotic " membranes exist about cells can take their choice, they can either utilize their conception to make water move or they can make their membranes permeable and so have dissolved substances move, but they cannot have both. Yet for life to go on in the cell both processes must be able to go on uninterruptedly. A further argument against any belief in semipermeable membranes about cells is found in the fact that in no cell studied has such ever been found by the examining eye. The morphological cell wall is admittedly not concerned in the osmotic activities of the cell. Usually the layer of proto- plasm just inside this is considered the so important semiper- meable membrane. This layer in plants differs in appearance from the rest of the cell protoplasm no more than the outermost edge of a leukocyte or an erythrocyte differs from the rest of the cell body. But in spite of this negative morphological finding such a semipermeable membrane might still exist. Such a sup- position, however, encounters trouble as soon as the fact is re- called that when a cell is cut in pieces or when the contents of a cell are squeezed out into a solution of any kind, these cell frag- ments (which assumes a spherical shape) behave just as the uninjured cell did before. This observation, which it seems to me points decisively against the existence of semipermeable membranes, has been accounted for by saying that the fragments form a new semipermeable membrane about them as soon as they come in contact with the solution into which they are dropped — supposedly in much the same way as new precipitation mem- branes may be 'formed in physico-chemical experiments. But in physical chemistry this formation of new precipitation mem- branes is not so universal an affair; it occurs only when two so-caUed membrane-forming solutions are brought in contact with each other, and it is hard to conceive of protoplasm being able to form a semipermeable membrane with just any solution with which it is brought in contact. The attempt may be made to meet this objection by saying that it is the universally present fat-like constituents (the lipoids) of the tissues which go to form the membrane when the cell fragments are dropped into any 158 (EDEMA AND NEPHRITIS watery solution, but as we shall soon see, the permeability oJ the lipoids to dissolved substances is far too limited to help us much toward an understanding of the phenomena that we are discussing. An enormous literature has sprung up about this question of " membranes " surrounding cells. From the original osmotic membranes of Pfeffer, which were semipermeable, we have come to those which are partially permeable and then to those which are permeable sometimes and then again not. But even these complicated notions encounter trouble, for there is so little connection between the kind of substances that enter cells and those that do not. Only the members of one group — that which has a ready solubility in the fats — have been recognized as having one property in common, and to account for their ready entrance into the cells the osmotic membrane about cells has been endowed with hpoid characteristics. The unfortunate part about this theory, which is in essence that of E. Overton, is that while it renders easier our conception of the absorption of these lipoid- soluble substances, it makes it impossible to get the ordinary salts and water into cells, for these are not particularly soluble in these lipoids. Overton has, in fact, come to such a con- clusion. And yet we know from physiological and pathological facts that all these types of substances must be able to. get into cells. Moreover, what do we gain when we have succeeded in getting any dissolved substance or water through any kind of membrane postulated to exist about a cell? It would collect here and we should still have to account for the movement of the dissolved substance or the water into or through the rest of the cell proto- plasm. There are no membranes about cells. All the phenomena which are difficult to explain when we assume membranes to exist about cells are readily interpreted without recourse to such postu- lates on the basis of the colloid constitution of protoplasm. In answer to these arguments sonie of my critics have retorted that a " membrane " exists whenever two phases come in contact with each other. At this point we have to stop and define terms, for here the argument begins to become academic. A drop of any fluid, a drop of any colloid solution, a drop of protoplasm or a cell has a " membrane " about it, but this " membrane " is simply a surface tension film; it has nothing in common with ABSORPTION, SECRETION— CELLS AND TISSUES 159 the " osmotic " membranes that in turn the botanists, the physical chemists and the original animal physiologists who worked in this field talked about. These surface tension films are chemically identical with the rest of the cell protoplasm and (except as colloid particles tend to collect in these surface films and so raise their concentration here) as such behave toward water or dissolved substances exactly as does the rest of the cell protoplasm. These facts indicate clearly that there is little reason for accepting the osmotic theory as of paramount or even great importance in ex- plaining the ways and means by which tissues absorb or secrete water. . I would like to be correctly understood in this matter. I am not maintaining that the laws of osmotic pressure may not account for some of the phenomena observed in at least some cells. This is a question which on the basis of the experimental data now available cannot be decided, but the biological significance origi- nally attributed to these laws has certainly been much overrated. Nor does such a decision against the r61e of osmotic pressure in these biological phenomena minimize in the slightest the value of the work of that score of investigators who have busied them- selves with this problem — they have made not only the best effort to analyze physico-chemically the forces active in the absorption and secretion of water by the cell and its myriad associated prob- lems, but they have laid down the experimental data upon which all subsequent workers on this problem must build. ^ 3. Criticism of the Lipoid Membrane Theory In an attempt to meet the inadequacies of the osmotic theory of water absorption by protoplasm, E. Overton ^ assumes that the surface of cells is made up of a substance which in its prop- erties as a solvent is not unhke ether or the fatty oils. He was led to this conclusion in attempting to account for the fact that many substances when dissolved in water are unable to " plas- molyze " cells. For example, while the various salts, at a suitable concentration, lead to a shrinkage of plant cells, a large number 1 For new and striking evidence against the osmotic notion of water absorption by cells see Wade W. Oliver: Science, 40, 645 (1914). ' E. Overton: Vierteljahresschr. d. naturt. GeseUsch. zu Zurich, 40, 1 (1895); 44, 88 (1899); Zeitschr. f. physik. Chem., 22, 189 (1897); Pfliiger's Archiv, 92, 261 (1902); Nagel's Handbuch der Physiologie, 2, 2te Halfte, 744 to 896 (1907). 160 (EDEMA AND NEPHRITIS of other chemical compounds, such as urea, glycerin, various sugars and alcohols do not do so. Believers in the osmotic theory of de Vries and Pfeffer explain this exceptional behavior by saying that the membranes assumed to exist about cells are permeable to this group of sub- stances. Overton has tried to define the nature of this per- meability by pointing out that all these substances have the property in common of being soluble in fats and fat-like bodies, and as such are universally present in protoplasm (as the lipoids — lecithin, cholesterin, protagon, cerebrin), he has tried to account for the permeability of cells to these substances by saying they go through these cell membranes because they are soluble in them. Most salts, he says, do not go through because they are insoluble in this surface film, in consequence of which they may extract water from the cell and so lead, at the right concentra- tion, to its plasmolysis. The difficulty with Overton's explanation is that in account- ing for the entrance of the lipoid-soluble substances, he makes it impossible to explain the entrance of that much larger group, of which the acids, alkalies and salts are representatives, the vast majority of which are not soluble in fat-like bodies. Even our foodstuffs and the products of cell metabolism belong in good part in this group. Yet, judging from physiological experi- ments, we know that these must be able to enter cells, otherwise how could we account for the normal life or the marked varia- tions in it which we are able to produce by means of these very substances? It is not enough to say that any or all of these substances move only through the intercellular substances. That certain substances in certain tissues may move more easily through the intercellular substance than through the cells them- selves is not questioned — salts, for example, do not diffuse with the same ease through different colloids — but that does not alter the main contention that salts and many other substances not soluble in the lipoids can and do pass into and through the cells themselves. Again, the assumption that cells are surrounded by a fat-like membrane makes it impossible to account for the entrance or exit of water from the cell. Water is not soluble in the fats (except theoretically), and hence cannot pass through a layer of it, and yet we know that in exceedingly short periods of time cells are ABSORPTION, SECRETION— CELLS AND TISSUES 161 capable of absorbing or secreting enormous amounts of water. The attempt might be made to explain this absorption of water by calling attention to the colloid properties of at least some of the Upoids — lecithin, for example, which is capable of absorbing water when dropped into it (Hober). But as soon as we accept this as true then our reasons for the non-entrance of the salts fall away, for when a lipoid absorbs water it loses at the same time its property of being solvent only for hpoid-soluble substances. What, again, do we attain when such have penetrated the lipoid surface membrane? We accomplish only an accumulation of the absorbed substance within the membrane itself, and just inside of this. We have then to explain how it gets through the rest of the cell. In the attempt to harmonize these conflicting notions, Na- THANSOHN ^ has assumed the surface of cells to represent a sort of mosaic, a part of which is formed by fat-like substances, another by " protoplasmic material " possessed of the properties of a semi- permeable membrane. The objections that must be raised against Nathansohn's conception are clearly a combination of those that were formerly raised against the osmotic conception alone, plus those that can be lodged against Overton's modification of it. The view of Nathansohn is, however, valuable because it brings out the idea of a mixture in protoplasm of substances having fat-like characteristics with such as do not possess this property. But to confine this mixture ' to the surface of cells is wrong because too limited. We encounter no difficulty in ex- plaining the various experimental facts at our disposal by ignoring altogether the existence of semipermeable or partially permeable membranes about cells. The substance of a cell consists of a mixture of different colloids. A part of these are colloid proteins with physical and chemical properties like those of fibrin, gelatin, etc.; a part colloid lipoids, which, while sharing some of the properties possessed by the proteins, as their power of swelling in water, have specific properties, such as their power of taking up substances soluble only in the fat-like bodies.^ ' Nathansohn: Pringheim's Jahrbiicher, 39, 607 (1904). A review is found in Hobeb: Physikalische Chemie d. Zelle u. d. Gewebe, Zweite Auflage, Leipzig, 176 (1906). ^Tfee idle of the colloid carbohydrates is ignored in this discussion because little of immediate interest to us has as yet been done with them. They are unquestionably of tremendous importance, especially in the plants. 162 OEDEMA AND NEPHRITIS 4. Adequacy of the Colloid-chemical Theory of Absorption and Secretion Let us now see what sort of a substitute for, or addition to, our present conceptions regarding the forces active in absorption and secretion is found in the r61e of the (hydrophiUc) emulsion colloids. What can the colloid-chemical theory of water absorp- tion do with the unexplained physiological facts detailed above? Two substances have always stood out prominently as ex- ceptions to the laws of osmotic pressure as considered active in protoplasm, the acids and the alkalies. The various tissues which have been examined in their dilute solutions all show an absorption of water vastly greater than can be accounted for on the basis of osmotic pressure. In fact, the amount that muscle can swell in dilute acids has been employed by Overton as a conclusive argument against the ordinary osmotic conception of water absorption by different tissues. He has pointed out that were all the proteins, carbohydrates and fats contained in muscle split into their simplest products, a sufficient jdeld of molecules and ions would not be obtained to furnish an osmotic pressure adequate to account for the amount of water absorbed. We have no trouble in explaining this behavior of acids and alkalies on a colloid basis. The acids and alkalies are among the sub- stances most powerful in increasing the hydration capacity of protein colloids. In this way we can account for the large amounts of water absorbed in the presence of traces of acid or alkali by red and white blood corpuscles, spermatozoa, muscle, the epithehal cells of the bronchi, intestine, bladder or esoph- agus, etc. There is also no difficulty in accounting for the unequal swelling of cells in osmotically equivalent solutions. This is true of the swelling of such simple proteins as fibrin, gelatin and gluten. In fact, the same substances which exhibit an exceptional be- havior in the " osmotic " study of cells show a like behavior in the case of simple proteins. [i j To find an analogue for the failure of muscle, red blood cor- puscles and cells in general to shrink the calculated amount vnth every unit increase in " osmotic " concentration is also simple. We need only to refer once more to the swelling of fibrin or of gelatin in which we found that here, too, doubling the con- ABSORPTION, SECRETION— CELLS AND TISSUES 163 centration did not halve the volume — the amount of decrease was always less than anticipated. In making the colloids responsible for the amount of water held by the tissues we escape all need for membranes. We can dispense with them in considering the absorption and secretion of water by cells just as we can in considering the ab- sorption and secretion of water by powdered fibrin or gelatin. Nor are we surprised when fragments of a cell behave toward external conditions as did the whole; in fact, we expect this, for colloids constitute the body of the cell, and just in so far as the colloids in the different parts of the cell do not differ from each other, in so far also do we not expect the processes of ab- sorption and secretion in these various parts to differ. The absence of a visible membrane does not annoy us — it simply indicates homogeneity of the protoplasm. The presence of a visible (not simply " osmotic ") membrane (such as a cellulose cell wall) interests us much more. It introduces another col- loid, and with it all the possibilities arising therefrom, for all colloids, so far as water absorption is concerned, do not neces- sarily react in the same way either quantitatively or quali- tatively toward any given set of external conditions. For this reason the protoplasm of a plant cell shrinks away from the surrounding cellulose wall when immersed in a concentrated salt solution, and is limited in its subsequent expansion if removed to wa,ter, for the colloids constituting the cellulose wall are not affected in so marked a way by low concentrations of acid, alkali or salt, as is the protoplasm within it. The possibility of ex- plaining the whole problem of inequahties in the amount of water held by different parts of the same cell therefore evidences itself here. What holds for the single cell holds also for different cells, in consequence of which we are not surprised when with variation in the colloid constitution of different cells we find a correspond- ing difference in their behavior when subjected to the same set of external conditions. Under normal circumstances different cells contain unequal amounts of water, and in states of ex- cessive turgor (oedema) neighboring, but morphologically differ- ent cells may show most unequal degrees of swelHng. Whether we deal with different parts of the same cell or with different cells does not matter, we need no semipermeable or other kind 164 (EDEMA AND NEPHRITIS of membrane to explain this. The absorptive powers of the various (hydrophilic) emulsion colloids are simply not the same or a colloid common to all the cells has been made to swell more in one place than in another by changes in its surroundings. Let it be added that we are now able to explain the variations in the water content of the much neglected intercellular sub- stances. In discussing water absorption by cells the intercellular substances are all too often overlooked, and this in spite of the fact that under physiological conditions some of the largest amounts of water are stored in tissues containing few cells (as the bell of the jelly-fish or Wharton's jelly of the umbilical cord), while in pathological states the very tissues in which cells are fewest, and intercellular substance most conspicuous, may grow richest in water. We need but recall the intense cedemas of connective tissue or the pathological changes characteristic of myxoedema. If we bear in mind that these intercellular substances are, like the cells themselves, but mixtures of different (hydrophilic) emulsion colloids, none of this surprises us. 5. Absorption and Secretion of Dissolved Substances by Protoplasm Thus far we have discussed only the absorption and secretion of water. We have now to consider the dissolved substances in the water. - To emphasize what should be self evident, we cannot and must not consider the absorption or secretion of water and the absorption or secretion of a substance dissolved in the water as iden- tical processes. Workers in biology make this mistake constantly. The processes of the absorption of water and of the absorption of dissolved substances do not parallel each other in simple physico- chemical experiments, and so need not, and do not, in living cells. The two are frequently associated, and may at times lie so closely together that they give the impression of running parallel with each other, but they are at all times independent of each other and may even take place in opposite directions at the same time. As we shall see later, a tissue may be absorbing a salt while it is secreting water, or vice versa. For the absorption of water by tissues we have made the (hydrophihc) emulsion colloids and changes in their state chiefly responsible. The colloid proteins appropriate the lion's share ABSORPTION, SECRETION— CELLS AND TISSUES 165 in this matter, but the colloid lipoids and the colloid carbohy- drates/ in so far as they have the capacity for holding water, must not be ignored. The absorption of dissolved substances is quite independent of the amount of water absorbed (except as the absorbed water retains the characteristics of ordinary water and . so increases the bulk of solvent available for water-soluble substances in the cell). Since in our criticism of the osmotic theory of water absorption we have incidentally destroyed the mechanism which different authors have used to explain the pecuharities noted in the absorption and secretion of dissolved substances by proto- plasm, we need to state how on the colloid basis we are going to account for them. The troublesome element in the whole problem is summed up in the observation that when any soluble substance is intro- duced into a living organism it does not distribute itself uni- formly throughout that organism. When we drop a crystal of some dye into a cylinder of water we know that after a while the dye by a process of diffusion comes to have the same con- centration in all parts of the liquid. The same dye (or any other substance, be this oxygen, a salt, a foodstuff or a medicinal agent) introduced into a living animal spreads through its tissues by a process of diffusion also, yet in the end one organ or one type of cell or different parts of one and the same cell may be stained to different degrees. Supporters of the osmotic theory have tried to account for such phenomena by saying that the osmotic membranes about cells possess a " selective permea- bility " which lets some substances through while it holds others out. In this way they believe dissolved substances to be kept apart in neat but differently concentrated packages throughout the living organism. But since we found it necessary to give up our belief in such membranes, we have to seek an explanation on another basis. Concentration differences can be maintained in different parts of the same cell, between different cells or between cells and their surrounding media even in the absence of " mem- branes " because of inequalities in distribution, determined by solubility, adsorption or chemical differences, or all three together. What this means may be illustrated as follows: ^The (hydrophilic) colloid carbohydrates are responsible for many of the peculiarities of water absorption as observed in plants. 166 CEDEMA AND NEPHRITIS (a) Inequalities in Distribution Due to Inequalities in Solubility When a solution of iodin in water is covered with a layer of ether and the whole is shaken, we can see even with the naked eye that the iodin is ultimately present in different con- centrations in the two liquids. While scarcely any remains in the water, the ether assumes a deep color from the iodin. The process is simply a homely illustration of the everyday chemical procedure of " shaking out with an immiscible hquid." The extraction of the iodin from the water depends upon the fact that iodin is soluble in ether, and so decidedly more so in this than in water that practically all moves over into the ether phase. The ultimate state of equilibrium attained, characterized by this very unequal distribution (partition) of the dissolved substance between the water phase and the ether phase, is in this case simply due to the difference in the relative solubihties of the iodin in the two solvents. The proportion of iodin dissolved in each of of the two phases — in this case a concentration of iodin about nine times as high in the one as in the other — is always constant. We call this the distribution coefficient or coefficient of partition. In discussing the living cell we have so far spoken of its solvent powers chiefly from the standpoint of its water content. If the cell had solvent powers determined only by its water content, it is obvious that dissolved substances could never appear in it in higher concentrations than those in which these substances are present in the media surrounding the cell. But such a con- ception of the cell is too limited. In addition to water, most of the various cells of all living organisms contain fat, and the already mentioned fat-like bodies known as lipoids (lecithin, cholesterin, cerebrin, protagon). We can see in advance that living cells containing fats or lipoids must be able to take up (that is, dissolve or absorb) many substances which are better soluble in such fats and lipoids than in water, in greater amounts than the media surrounding these cells which are not so rich in or lack these compounds entirely. We are indebted to Hans Mbyek ^ and E. Oveeton ^ for recognizing the great physiological importance of the facts here ^Hans Mbyee: Archiv f. exp. Path, und Pharm., 42, 109 (1899); ibid., 46, 338 (1901). ^ E. Overton: See reference on page 159. ABSORPTION, SECRETION— CELLS AND TISSUES 167 outlined. By methods which we need not discuss here, they found it possible to differentiate between substances which pass into or through cells but slowly and those which do this rapidly. To the compounds which diffuse rapidly into protoplasm belong the monatcmic alcohols, aldehydes and ketones, the hydrocarbons with one, two and three chlorin atoms, the nitro- alkyls, the alkylcyanids, the neutral esters of inorganic and many organic acids, anilin, etc. The diatomic alcohols and the amins of monatomic acids pass into cells more slowly, and still more slowly glycerin, urea and erythrite. The hexatomic alcohols, the sugars with six carbon atoms (hexoses), the amino-acids and the neutral salts of the organic acids diffuse into cells only very slowly. A glance at this list shows that we have to deal with all manner of chemical substances. Some are relatively simple in composi- tion while others are very complex; some are of physiological importance and found normally in the living cell; others are entirely foreign to the living organism. What physico-chemical character have they in common which allows them to penetrate living cells with more than usual ease, and so to stand out from the great group of the ordinary neutral salts, for example, which do so only slowly? They are all more soluble in fat solvents than in water, and therefore pass into and through cells contain- ing fats and lipoids more rapidly than into and through such as do not. With a given cell the rapidity and the absolute amount of any compound ultimately absorbed must depend upon its relative solubility in water and in the fat or fat-like bodies con- tained in the cells. In other words, it depends upon what is its distribution coefficient between the two phases whether any dis- solved substance will enter a cell slowly or rapidly, and whether it will ultimately be found in the cell in a greater, in the same or in a lower concentration than in the medium surrounding it. The importance of these simple facts is self evident. In order that a substance may produce any physiological effect it must first get into the cell. Other things being equal, a quicker and more powerful effect will be produced by a lipoid-soluble preparation than by one not thus soluble. The marked effects of the anesthetics (chloroform, ether, alcohol, ethyl chlorid) and of various alkaloids (morphin, cocain, atropin) is associated with their fat and lipoid solubility. The 168 (EDEMA AND NEPHRITIS nervous tissues are high in fat and fat-like bodies, and so take up these substances with special avidity. Because of his greater stock of fat solvent, the fat individual demands more anesthetic before going to sleep than does a lean one. Anesthesia, like all intoxication, is a matter not of absolute amount of anesthetic present, but of concentration. The various grades of anes- thesia go hand in hand with definite concentrations of anesthetic in certain cells of the nervous system, and it must evidently take longer to attain this concentration in a fat man than in a lean one. (b) Inequalities in Distribution Due to Inequalities in Adsorption Not- only may a living cell come to contain in the unit volume a greater or less amount of any dissolved substance than does the surrounding medium because the cell contains better or worse solvents for it, but the cell may do this because of its adsorptive powers. These adsorptive powers are associated with the fact that the cell is largely colloid. The general problem of adsorp- tion may be illustrated as follows: When a dye is dissolved in distilled water a uniformly colored solution results. If the solution is divided and to one-half is added a little finely-powdered charcoal while nothing is done with the other, we find after shaking both that while our control solution remains entirely unaltered, the color largely disappears from the other. The decolorization has not been chemically induced; the pure carbon does not react chemically with any of the constituents in the tube. The powdered charcoal has a great surface, and the action of this upon the dissolved particles of dye has made them accumulate (condense) upon it. The theory of how this surface action is accomplished need not interest us here. What has been described is an example of adsorption. The charcoal is the adsorbent, the dye, the adsorbed substance. Any number of substances could be cited as acting under various conditions as adsorbents; and almost any substance may act as the material capable of being adsorbed. Finely- divided kaolin, precipitates of various kinds and inorganic or organic colloids may take the place of carbon in the above ex- periment, and acids, alkalies and salts can be adsorbed in the same way as the readily visible dye. All adsorbents do not. ABSORPTION, SECRETION— CELLS AND TISSUES 169 however, behave quaUtatively or quantitatively in exactly the same way toward any given material to be adsorbed, and dif- ferent external conditions modify markedly the adsorption ex- hibited by any given adsorbent. Examples of adsorption are familiar to everyone. The commercial decolorization of beers, sugars, etc., by animal charcoal; the removal of color from a bath by dipping wool, cotton, etc., into it (dyeing) ; the staining of histological specimens are all examples of adsorption. The adsorption of any substance by an adsorbing agent is never complete. Charcoal never takes all the dye out of a bath; some always remains behind. The distribution of the dye be- tween the solvent and the adsorbent is governed by the laws of equiUbrium. After the charcoal has taken up as much of the dye as possible, if the supernatant liquid is poured off and pure water is substituted for it, then some of the dye leaves the char- coal and goes back into solution in the water. In this way we can again wash all the dye out of the charcoal. Conversely, when the charcoal has taken up as much dye as it will from a given dye bath, it will proceed to take up an additional amount if more dye is added to the supernatant liquid. The relationship between the concentration of the substance to be adsorbed and the amount taken up by the charcoal is an interesting one and may be thus stated: From relatively dilute solutions the adsorbent will take up much, from more concentrated solutions relatively less, of the substance to be adsorbed. In other words, if at a certain concentration we can take four-fifths of the dye present in a solution out of this with a given amount of charcoal, then if the dye has a higher concentration we can take out only less than four-fifths, or if it has a lower concen- tration, more than four-fifths. Protoplasm behaves toward substances dissolved in a medium that surrounds it in an entirely similar way. Upon this depends the fact that it may contain the same, a higher or a lower con- centration of any dissolved substance than the medium sur- rounding it. Since the protoplasm (adsorbent) of different cells is not the same, it comes to pass that while they are all bathed by the same blood and lymph they nevertheless do not all adsorb the same amount of the proffered materials. In other words, equihbrium is not attained between the protoplasm of different cells and the medium surrounding these at exactly the same 170 (EDEMA AND NEPHRITIS point. Hence it comes to pass that the salt content of the blood, or its content of a dye, a chemical or an immune body, may not only be unlike that of the cells, but that it need not be the same in different cells. The adsorption properties of protoplasm are markedly in- fluenced by various external conditions ^ exactly as are those of a laboratory adsorbent. Thus, if acid is introduced into proto- plasm, its adsorption powers change markedly. In this way a cell or tissue which, under normal circumstances, acts as an ex- cellent adsorbent for a dissolved substanbe, may practically lose this, or, conversely, one which before adsorbed a given sub- stance only poorly may now take this up with avidity. (c) Inequalities in Distribution Due to Specific Chemical Differences A third reason why a cell may contain substances in a higher (or lower) concentration than the medium surrounding it resides in the fact that it contains substances capable of combining with the proffered dissolved substance. Thus, if a cell contains iron, it may be expected to take up more of a proffered substance capable of combining with this iron (say a ferrocyanid) than a cell devoid of it or containing it in less amount. We need not multiply such illustrations, for the list is as long as the list of chemical reactions capable of ensuing between the various sub- stances found in any living cell and the substances that come normally or abnormally in contact with it. The " specific absorption " and consequent " specific effect " of various pharmacological preparations, of " toxins," of " fer- ments," etc., is generally regarded as an expression of such in- equalities in distribution due to specific chemical differences.^ This point of view is largely correct, but it is well to emphasize that it is Ukely to be carried too far. We are still too strongly under the influence of the " purely chemical " point of view in this matter. We have already learned that many of the " spe- cific " immune reactions are not so intensely " specific "; and the ' See page 514, and such works as L. P^ilet-Jolivet: Die Theorie des Farbeprozesses, Dresden (1910). ^ Paul Ehelich: Sauerstoff-Bediirfnis des Organismus, Leipzig (1885); Deutsch. med. Wochensohr., 697 (1898); Collected Studies on Immunity (translated by Bolduan, New York, 1907). Heinbich Zangqbb: Viertel- jahresohrift d. naturforsoh. Gesellsch. in Zurich, 53, 408 (1908). ABSORPTION, SECRETION— CELLS AND TISSUES 171 whole realm of colloid-chemistry is dotted with examples of reactions that were looked upon as " chemical," when further analysis has shown that the " specific " in these reactions did not depend so much upon the presence of certain chemical com- pounds as upon the physical states in which the components entered into the reactions. Were we at this point to sum up our conception of the structure of protoplasm as thus far developed, we could liken it fairly accu- rately to a mass of protein intimately mixed with more or less fat- like material {the fats and lipoids), the whole being under physio- logical conditions immersed in a liquid {pond water in the case of an ameba, or lymph and blood in the case of our body cells) from which the protein-fat mixture soaks up a certain amount of water as well as a certain amount of the various dissolved substances found in the water. The water absorption is governed by the state of the {hydrophilic) emulsion colloids. The absorption of dissolved sub- stances is a matter of equilibrium between the concentration of those found in the medium outside the cell and that of the same sub- stances found in the cell itself. We have indicated how solubility characteristics, phenomena of adsorption and chemical combination influence the point at which equilibrium is reached. This simple picture of the cell furnishes to our minds an adequate con- ception of its main structure. PART THREE (EDEMA PART THREE (EDEMA INTRODUCTION The foregoing pages have brought evidence indicating that the colloids of the tissues and their state are chiefly if not entirely responsible for the amount of water they hold under various conditions. The problem of cedema is also but a problem in colloid- chemistry, the problem of the ways and means by which the normal hydration capacity of the body colloids is heightened. We have had occasion to discuss the great contributions first made by the plant physiologists and later adopted by the animal physiologists to this general question of excessive turgor as observed in various cells. We have now to consider the con- tributions made by the pathologists to this same question, which they, however, discuss under the heading of oedema. I deem it purposeless to review in detail all their theories. Such a task would not be easy, for the contentions of the various authors cannot always be stated in brief and do them justice. Too often this is because they have mixed their excellent experi- mental findings with the particular hypothesis which they were attempting to make into a theory; and too often also do we find good attempts to account for cedema on a mechanical basis mixed with the vague conceptions of the activities of " living " protoplasm. Any effort, therefore, to analyze the contributions 175 176 (EDEMA AND NEPHRITIS of an author to the general subject of the nature and the cause of oedema must distinguish carefully between the value of that which he may have contributed to the experimental side of the subject and to the theoretical. Richard Bright made one of the first attempts to account for oedema when he tried to find in the loss of albumin from the body in nephritis a cause for the thinning of the blood (hydremia). Such hydremic blood, he reasoned, would then pass easily through the blood vessels and into the tissues, and so cause the latter to swell. An experimental investigation of Bright's hypothesis forms the basis of the much-discussed work of Julius Cohnheim and LuDWiG Lichtheim.^ These authors found that the injection of enormous quantities of sodium chlorid solution into the veins of various animals did not bring about an oedema similar in dis- tribution to that observed in Bright's disease, and so decided that hydremia alone, or a hydremia connected with an increase in the amount of blood circulating in the blood vessels (hydremic plethora), could not be responsible for the oedema of nephritis. Experiments carried out by a number of authors since the work of Cohnheim and Lichtheim confirm, in the main, their findings. For the older hydremic theory, Cohnheim and Lichtheim substituted what we may call for short the pressure theory of oedema so widely accepted by pathologists to-day. Briefly formulated, it holds that variations in the pressure of circulating liquids (such as the blood or lymph) are chiefly responsible for the variations in the amount of water held by the tissues, in that through changes in pressure the circulating liquids are supposed to be forced through the vessel walls into the tissues. It is not strange that this beUef should have originated and seemed espe- cially acceptable to those biological workers whose interest centers particularly in animals which possess the conspicuous feature of a circulatory system. But the very school which originally laid most stress upon this force — the school of patholo- gists and the earlier physiologists — has found its efforts to increase the amount of water held by tissues through an increase by experimental means of the blood and lymph pressures, to result in failure. 1 Julius Cohnheim and Ludwig Lichtheim : Virchow's Archiv, 69, 106 (1877); also Cohnheim: AUgemeine Pathologie, Zweite Auflage, 1, 430, Berlin (1882). (EDEMA 177 To meet the deficiency there was therefore added to the changes in pressure a second element — a change on the part of the involved tissues themselves. Thus Cohnheim himself recog- nized that severe oedemas occur in animals when no alteration whatsoever in blood pressure is apparent. To account for them under such circumstances he invoked an "increased permea- bility of the blood vessel walls." Such a belief has never been proved experimentally, and, indeed, of what consequence would it be from a pathological standpoint? To force liquids through blood vessel walls is not to force them into the tissues. And the fluid of an oedematous tissue is very decidedly in the tissues themselves. Cohnheim's hypothesis would simply squeeze the cedema fluid as far as the outer wall of the capillaries. A series of other observers have expressed the tissue factor in yet other terms. We read that in addition to changes in blood pressure there are also necessary for the development of oedema a " loss of elasticity " on the part of the tissues, a " decreased ability to hold back water," a " heightened imbibition," etc. But none of these authors had any clear-cut notion of just how these factors were operative. In 1898 Jacques Loeb ^ tried to explain oedema in the terms of the osmotic theory of water absorption by defining it as a state • of increased water absorption due to an increased osmotic pressure of the cell contents. This was the same explanation which Hambukger,^ von Ltmbeck,^ GiJRBER* and Eijkman^ had previously used to account for the excessive swelling of various cells when subjected to the action of acid or various other changes in their surroundings. Its inadequacy was pointed out by E. . Overton ^ when he showed that were all the proteins, carbo- hydrates and fats contained in our tissues split into their simplest molecules, enough molecules and ions would not result to yield a sufficient osmotic concentration to account for the amounts of 1 Jacques Loeb: Pfliiger's Archiv, 69, 1; 71, 457; 75, 303 (1898). 2 H. J. Hambuhger: Arch. f. (Anat. u.) Physiol., 513 (1892); 153 (1893); Zeitsohr. f . Biol., 35, 252 and 280 (1897), where references to his earlier papers are found. See also Arch. f. (Anat. u.) Physiol., 31 (1898). 'C. VON Limbeck: Arch. f. exper. Path. u. PharmakoL, 35, 309 (1894). * Gtjkber: Sitzberich. d. med. phys. Gesellsch. Wiirzburg, Feb. 25 (1895). s C. Eijkman: Virohow's Arch. f. path. Anat., 143, 448 (1896), where references to his earlier papers will be found. 8 E. Ovebton: Pfluger's Archiv, 92, 115 (1902). 178 OEDEMA AND NEPHRITIS water absorbed by such swollen (cedematous) cells. Following this criticism Loeb in republishing his papers struck out all his views on oedema.^ But while the theory which Loeb tried to support cannot be upheld (any more than the general osmotic theory of water absorption) he threw out with his rejected oedema views an experimental fact which is of permanent scientific value, and which, if it had been properly appreciated by pathologists and clinicians, would have spared us much of the late literature on this subject. I refer to his experiments carried out with the intention of establishing the fact that the cause of oedema resides in the tissues. II THE CAUSE OF (EDEMA RESIDES IN THE TISSUES A simple experiment proves this. If one leg of an ordinary frog (Rana), a tree frog (Hyla) or a toad (Bufo) is hgated just above the knee as tightly as possible, so that the ligature shuts off not only the venous flow, but also the arterial, and the animal is then placed in sufficient distilled water to cover the legs, the ligated leg develops an intense oedema, while the unligated one remains normal. To explain this result recourse cannot be had in this experiment to the pressure of any circulating liquids, for none such exists, and so all the conceptions of cedema which regard the pressure, per se, of circulating liquids, as one of the causes, or the chief cause, in the development of this condition, are robbed of their most fundamental support. The choice of animals for these experiments was not entirely a random one. It seemed desirable to deal with such in which there exists normally an outside source of water for the tissues, one separate from the ordinary blood or lymph current. Such conditions are satisfied in any of the amphibians. Nevertheless, ' Loeb's original paper discussing oedema appears in Pfliiger's Archiv 71, 457 (1898). His collected papers in English appear under the title, Studies in General Physiology, I and II, Decennial Publications of the Uni- versity of Chicago, Chicago (1905). What is left of the original article appears on page 501. (EDEMA 179 an absorption of water may be obtained through the skin of all animals, for my toads developed just as intense oedemas of the leg as did the frogs, and it is a well-known fact that the bodies of dead land animals swell (become oedematous) if kept in water.^ The oedemas which these frogs and toads develop are in every way a counterpart of the most intense forms observed chnically. The tissues are boggy, pit on pressure and when incised allow the escape of fluid. The rate at which the oedema develops in the three types of animals is not the same. It develops and passes away most rapidly in tree toads (Hyla). For toads (Bufo) and ordinary frogs (Rana) the following holds: An oedema of the ligated leg is readily discernible at the end of eighteen hours, and is marked at the end of twenty-four. Within forty-eight hours the sweUing approaches its maximum, and may at times be so great that the skin of the ligated leg is ruptured. This maximal swelling is usually maintained some two days, when it begins to diminish. The diminution in the size of the leg is at first merely due to loss of water, dependent upon changes in the tissues which we shall discuss later. But in the entire absence of a circulation the leg below the ligature cannot, of course, continue to live, and so anywhere from one to two weeks after the ligature has been tied, the skin peels and splits and the tissues below it become soft and disintegrate. This loss of substance becomes pro- gressively greater until at the end of three to five weeks only a bony stump covered with tags of tissue may be left. A number of accessory phenomena are deserving of mention. Twenty-four to forty-eight hours after the ligature has been tied a number of small vesicles usually begin to develop upon the oedematous leg. They are found earliest and most commonly in the tissues of the web of the foot, but they may occur anywhere in the skin below the Ugature. The small vesicles, which appear early, gradually increase in size until forty-eight to ninety-six hours after the ligature is tied they become great blebs, which in place of the original water-white or faintly straw-colored fluid ' It is needless to point out that the bloaiing of bodies in consequence of the development of gas through bacterial action in the gastro-intestinal tract or in the tissues proper is, of course, not referred to in this remark. 180 (EDEMA AND NEPHRITIS found in the vesicles, are likely to contain (especially in toads) a blood-stained serum. After these have persisted a day or two, they rupture and allow the escape of their contents. The color of the skin of the ligated legs also suffers change. Within twelve hours after the hgature is tied this is usually seen to fade somewhat, and to lose the luster of the healthy skin. At the end of forty-eight hours the color markings, characteristic of the particular species of frog under observation, are always much blurred. Late in the experiment (in the second or third week), the ligated leg assumes the gray or grayish-black look of necrosis. What has been said is illustrated in Figs. 69, 70, 71, 72 and 73. Fig. 69 is a photograph of a frog (Rana), kept in a little water, forty-seven hours after a ligature has been tied as tightly as possible about the left leg. The increase in the size of this leg over the normal right is clearly apparent. Figs. 70 and 71 illus- trate the same fact in another frog treated the same way. The tense skin with the blurring of surface markings is easily noted in all three pictures. Later photographs of the frog of Fig. 69 are shown in Figs. 72 and 73. These were taken ninety-five hours after the ligature was tied. Some small bhsters which formed between the toes have increased in size to constitute the large bleb seen in the photographs. The oedema in the leg and foot generally is still evident. While these photographs show us that an oedema develops in a frog's leg even in the total absence of a circulation, they tell us nothing of the severity of these oedemas; in other words, simple inspection of the illustrations does not yield conclusive evidence that the cedemas are as severe as any ever observed clinically. To settle this a few experiments are given in which the oedematous legs were amputated at various periods after their ligation and their weight compared with that of the normal leg of the opposite side. Experiment 1. December, 1907. — One leg is ligated with silk just above the knee in each of four toads (Bufo), and they are placed in separate dishes, each containing enough distilled water (50 cc.) to cover the legs. The ligated legs are found visibly (Edematous at the end of twenty hours. The toads are left in the dishes for fifty-four and one hundred and sixty-eight hours, when they are killed, and the two legs are amputated (the hgated one just above the hgature, the other at a corresponding point on the opposite leg), and weighed. (EDEMA 181 Figure 69. 182 (EDEMA AND NEPHRITIS Figure 70. CEDEMA 183 Figure 71. 184 CEDEMA AND NEPHRITIS FiGnRB 72. (EDEMA 185 FiGUBE 73, 186 (EDEMA AND NEPHRITIS The difference in weight, with the gain on the part of the ligated leg, expressed in percentage of the weight of the unhgated leg, is shown in the table. K4 1, rr J A /Ligated, 0.436 (+26%) 54 hours, Toad A | Un^at^d, . 346 (0%) KAX. rr ,Tj /Ligated, 0.371 (+25%) 54 hours. Toad B (u^Ugated, 0.296 (0%) KAu rr. ,„/ Ligated, 0.444 (+45%) 64 hours. Toad C | Unjjgated, 0. 306 (0%) ifisi,.„.= T„«^ n /Ligated, 2.152 (+82%) 168 hours, Toad D ( Unligated, 1.180 (0%) Experiment 2. March, 1908. — Ligatures are tied as tightly as pos- sible just above the knee about the left hind legs of six toads (Bufo). They are placed in finger, bowls containing 100 cc. distilled water. After varying periods they are taken out, killed, and the weights of their hind legs compared as already outlined. The results are given in the table: A ofM. T J A (Ligated, 1.080 (+33%) 4.30 hours. Toad A | u^gat;d, 0.810 (0%) ■I -Ton I, rr J Tj (Ligated, 3.350 (+58%) 17.30 hours, Toad B (unUgatk, 2.120 (0%) 28 OOhours Toad C /Lighted, 4.137 (+80%) 28.00 hours, ioadO junligated, 2.295 (0%) -lo ^..1. T J T, /Ligated, 7.840 (+56%) 43.45 hours, Toad D { uAigat;d, 5.020 (0%) CO nni- rr ,„ f Ligated, 7.262 (+130%)? 63.00 hours, ToadE {u^gatk, 3.152 (0%) 10A AKV. rr J v /Ligated, 7.160 (+23%) 124.45 hours, Toad F (unligated, 5.810 (0%) ^ The toes of the sound leg are missing. The cedematous leg is practically covered with large blebs. These experiments prove that the severest grades of CEdema may develop in toads and frogs in the entire absence of a circula- tion. Their validity to do so has, however, been questioned. The objections raised come to this, that in spite of the ligature some sort of a blood or lymph circulation with its ever-adherent " pressure " still exists in the leg. One should, of course, be convinced that no ordinary circulation can continue through the soft tissues of the leg when it is remembered that the ligature is tied as tightly as possible about the leg at a point where musculature is practically lacking. The only other possibility for a circulation would have to be found through the lower end of the femur, and the tissues in and about the knee-joint, whereby ? 1 CEDEMA 187 a connection between the thigh above, and the leg below, might be conceived to be continued. These objections are answered by two facts: (1) If the ligature is tied about one leg of a frog, and the animal is not kept in water, but in a dry vessel, the ligated leg dries up entirely, and this member is carried about in a mummified condition for as long as the experiment is continued. The rest of the frog dries out more slowly than the ligated leg. (2) If after ligating the leg the member is amputated and placed in a little distilled water, it shows the same series of changes as though it had been left united to the frog. We will find abundant evidence of this fact in experiments to be described later. Ocular demonstration of it may be found in Fig. 74. In this are shown anterior and posterior views of two frogs' (Rana) legs forty-nine hours after they had been ligated, amputated, and placed in distilled water. The spreading toes, bulging webs, and swollen leg muscles betray the oedema. Its severity is made apparent when it is stated that both legs have gained over 50 per cent in weight. It would be difficult to conjure up the existence of any orthodox circulation in this experiment with amputated legs. Experiments 3 and 4 may serve in further illustration of what has been said. In these, tree toads were used. Similar experiments with frogs will be described later and need not be dealt with separately here. ExPEiiiMENT 3. December, 1907. — ^A ligature is passed about the left leg above the knee in each of two tree toads. The one toad is kept in a dry vessel, the other in one containing a little distilled water. Twenty hours later, oedema is well marked in the ligated leg of the frog kept in water, while that of the frog kept dry is already beginning to shrivel. At the end of fifty-eight and one-half hours the toads are killed, the legs amputated and weighed, with the following results: rr, rp J . , ^. / Ligatcd, . 071 ( "38 . 8%) Tree Toad A, kept dry | uSigat^d, . 116 (0%) ^ T, ,„, , ../Ligated, 0.279 (-1-78.8%) Tree Toad B, kept moist | UnUgated, 0. 156 (0%) Experiment 4. June, 1908. — Three tree toad legs are amputated close to the pelvis. The skin is puUed over the femoral stumps and ligated tightly. The legs are weighed and placed in separate finger bowls each containing 110 cc. distilled water. The first figure in each of the columns is the weight of the tree toad's leg at the beginning of the experiment. After each of the subsequent weighings is given in 188 (EDEMA AND NEPHKITIS B Figure 74. (EDEMA 189 parentheses the percentage of increase in weight over the original weight of the muscle. Hours in tbo solution. 110 cc. H2O. llOoc. HjO. 110 CO. H2O. 0.486(0)% 0.473(0)% 0.363(0)% 0.30 0.560 ( + 15.2) 0.540 (+14.1) 0.417 ( + 14.8) 1.30 0.600 (+23.4) 0.590 (-+24.6) 0.465 (+28.1) 2.30 0.630 (+29.6) 0.620 (+31.0) 0.498 (+37.2) 4.30 0.712 (+46.5) 0.703 (+48.6) 0.557 (+53.4) 6.10 0.795 (+63.5) 0.770 (+62.8) 0.620 (+72.7) 17.35 0.944 ( +94. 2)x 0.799 ( +68. 9)x 0.662 ( +82. 3)x 22.25 0.842 (+73,2) 0.772 (+63.2) 0.627 (+74.7) 28.45 0.780 (+60.5) d 0.755 (+59.8) d 0.598 (+64.7) d, d, represent opposite legs of the same toad. X. At this point the legs are found blistered. Fig. 75 is based upon the calculations contained in Experiment 4, and represents graphically the course of water absorption Figure 75. as observed in these three amputated tree toads' legs. The curves show that the initial increase in weight is followed later by a decrease. This corresponds with the ocular observations already detailed on the development of cedema in ligated legs left in situ. 190 (EDEMA AND NEPHRITIS After what has been said it will not seem strange that these oedematous changes in a ligated leg occur in a toad or frog just as readily and rapidly when the animal has its central nervous system destroyed as when this is not done. If, however, the animal dies, the difference between the weights of the two legs does not develop. But this is not because the oedema does not develop in the ligated leg — it does just the same — but an equally intense absorption of water occurs in the other leg which through the death of the animal has been deprived of its circulation. These experiments already enable us to cast aside all those explanations of oedema which attribute its development to pressure changes per se of circulating Uquids. The cause of oedema resides in the tissues themselves, and these become oedematous not because water is forced into them, but because changes take place in them whereby they are enabled to absorb water from any available source. In the case of the experiments on toads and frogs this available source of water is the water contained in the dishes in lohich the animals are kept. In clinical cases of oedema, it is found in the fluids which pass through or about a tissue. Ill ON THE NATURE AND CAUSE OF (EDEMA We are now in a position to attempt an analysis of the nature and the cause of oedema. In order to render clear the argument that follows and the purpose of each experiment, we will at once state our conclusion. A state of oedema is induced when- ever, in the presence of an adequate supply of water, the capacity of the colloids of the tissues for holding water is increased above that which we are pleased to call normal. Any agency capable under the conditions existing in the body, of thus increasing the hydration capacity of the tissue colloids constitutes a cause of oedema. The accumulation of acids within the tissues brought about either through their abnormal production, or through the inadequate removal of such as some consider normally produced in the tissues, is chiefly responsible for this increase in the hydration capacity of the colloids, though the possibility of explaining at least some of it through the, production or accumulation of substances {of the type of urea, pyridin, certain amins, etc.) which can hydrate CEDEMA 191 colloids as can acids, or through the conversion of colloids having but little capacity for water into such as have a greater capacity must also be borne in mind. It was the purpose of Part Two in this volume to prove that in the colloids of the tissues and in their variable capacity for holding water we have an adequate explanation for the largest amounts of water ever held by tissues under phys- iological conditions or in states of excessive swelling (excessive turgor, plasmoptysis, oedema). We need now to discuss how the degree of hydration characteristic of normal cells may be so increased that they are judged cedematous. Of the several agencies active here we shall discuss in greatest detail, because we consider it most important, the question of an abnormal production and accumulation of acid in the involved tissues. Our proof for the truth of the general conclusion stated above will take three directions: 1. An abnormal production or accumulation of acids, or con- ditions predisposing thereto, exist in all states in which we encoun- ter the development of an oedema. 2. Conversely, any means by which is rendered possible the abnormal production or accumulation of acids in the tissues is accompanied by an cedema. 3. The development of an oedema is antagonized by the same substances which decrease the capacity of the (hydro- philic) emulsion colloids for holding water and is unaffected by substances which do not do this. We will consider these separately: 1. An Abnormal Production or Accumulation of Acids or Condi- tions Predisposing Thereto Exist in all States in which We Encounter CEdema § 1 We are especially prone to see states of oedema develop in conjunction with circulatory disturbances. Thus, when the function of the heart is sufficiently impaired an oedema which is more or less general affects the body. It is ordinarily said that because of the disturbance in the circulation an increased capillary pressure results, in consequence of which fluid is 192 OEDEMA AND NEPHRITIS squeezed out into the tissues. And yet everyday clinical experi- ence shows that such reasoning is entirely wrong, for if we give our patient digitalis or some other heart " stimulant " which increases the blood pressure the oedema gets better, not worse. What we have said regarding a disturbance in the general circula- tion holds also for the local interferences with the circulation as through thrombosis, embohsm, or ligation of an artery or vein supplying any part of the body. If a good collateral circula- tion exists, the thrombosis, embolism, or ligation may be entirely without effect, or if such a collateral circulation is gradually established the oedema may gradually pass away; but if neither of these is possible, then the oedema persists. A patient tends to develop a general or a localized oedema whenever an insufficient amount of arterialized blood is being pro- pelled through his tissues, and any general or local condition which produces such a state or aggravates an existing one, aggravates the oedema, and vice versa. This is why postural changes, rest in bed, drugs which increase the effectiveness of the heart's work, and measures which tend to restrict those physiological functions which we know normally to be followed by an increased demand for blood all help to improve an oedema, while the reverse does the opposite. But how does insufficient flow of normal blood through a tissue lead to an oedema? Is such accompanied by an abnormal production or accumulation of acid? As already stated, this is what we need and know from our previous experiments to be potent in increasing the capacity of the tissue colloids for water. That this is the case is a well-known and long-established fact. When the blood is not carried away from a tissue at its normal rate there tends to accumulate in it and in the tissues drained by it the carbonic acid which is constantly being produced in our cells. It is this carbonic acid which under normal circum- stances accounts for the swelling of the red and white blood corpuscles whenever the arterial blood changes to venous,^ and this tendency is greatly heightened when the normal blood is replaced by the highly venous blood encountered in circulatory disttirbances. What happens in the cells of the blood happens also in the tissues cells drained by that blood. They all tend ^ See the experiments of Hamburger, von Limbeck, Grtns, Eijkman, etc. CEDEMA 193 to swell just as do fibrin flakes in water when the carbonic acid tension in it is increased. The observation of Strassburg and EwALD that the carbonic acid content of oedema fluids and of tissues deprived of a circulation runs very high is therefore one of the factors to be considered in trying to find a cause for the increased capacity of the tissues for holding water in states of dis- turbed circulation. There exists, however, a second and more powerful factor which leads to an abnormal acid production when the circulation is disturbed. This is brought about through the inadequate supply of oxygen to the affected parts. As first proved through the striking experiments of Trasaburo Araki,^ dogs, rabbits, and frogs excrete lactic acid in their urine in addition to various other abnormal substances whenever subjected to oxygen want by any means whatsoever. Under ordinary circumstances lac- tic acid is not found in the urine, but let the oxygen supply to these animals be sufficiently reduced (through confinement in a closed box, through carbon monoxid poisoning, or through the injection of curare, amyl nitrite, or cocain and the acid appears. Such acid is also found in human beings when through accident or disease they are compelled to suffer from oxygen want. Lactic acid is not the only acid that may be or is produced under such circumstances. E. Mendel ^ found the phosphoric acid content of the urine increased after epileptic seizures and in apoplexy, and F. Hoppe-Seyler ^ found various oedema fluids to contain valerianic, succinic, and butyric acids, besides lactic. The lactic acid found in the urine in conditions associated with a lack of oxygen is produced in the tissues, enters the blood, and is excreted by the kidneys. This has been proved by Araki's later work and through Hermann Zillessen's * experi- ments. ZiLLESSEN found that when the oxygen supply to a muscle or to the fiver is shut off for a variable number of hours through ligation of the arteries supplying these parts, ar increased production of lactic acid occurs. If the ligature is loosened and the first blood returning from the oxygen-starved tissues is ' Trasabttro Araki: Zeitschr. f. physiol. Chemie, 15, 335 and 546 (1891) ; ibid., 19, 422 (1894). ^E. Mendel: Archiv f. Psychiatrie u. Nervenkrankheiten, 3, 636. 3 F. Hoppe-Setler: Zeitschr. f. physiol. Chemie, 19, 476 (1894). ^Hermann Zillessen; Zeitschr. f. physiol. Chemie, 15, 387 (1891). 194 (EDEMA AND NEPHRITIS analyzed, this is found to be particularly rich in lactic acid, and if the blood is titrated, it is found to have a diminished capacity for neutralizing a standard oxalic acid solution.^ In consequence of circulatory disturbances, whether general or local, an abnormal production and accumulation of carbonic, lactic and other adds occurs which increases the hydration capacity of the colloids of the involved tissues, because of which they then suck water out of the blood and lymph streams bathing them, §2 In place of interfering mechanically with the circulation we may make the tissues suffer from a lack of oxygen, and thus from an abnormal production and accumulation of acid, by interfer- ence with the normal oxygen-carrying power of the blood. We find in this a ready explanation of the oedemas so frequently noted in the severe anemias, no matter what their cause. It is of interest, therefore, that Felix Hoppe-Seyleb ^ was able to isolate lactic acid from the urine in two cases of severe anemia. As additional evidence in this direction may be cited R. von Jaksch's 3 findings, . amply verified by subsequent workers, that the blood shows a distinctly diminished power of neutraliz- ing acid in pernicious anemia, leukemia, and chlorosis. That this decrease in the ability to neutralize acids really means that an abnormal production of acids has occurred in the tissues of the anemic individual, is evident not only from the work of Abaki and Zillbssen already cited, but from von Jaksch's own finding that in carbon monoxid poisoning in which the ab- normal presence of lactic acid in the urine has been indisputably shown by Abaki, there also exists a distinct decrease in the normal capacity of the blood to neutralize acids. §3 An oedema, often of a severe grade, is the almost constant accompaniment of various states of inanition. It is observed ' Araki, Zillessen and most of the earlier observers speak of a "decreased alkalinity " of the blood. Because modem physico-chemical conceptions have changed our old notions of what constitutes alkalinity, it is best to state the experimental findings of these authors as above. 2 Felix Hoppe-Seyler: Zeitschr. f. physiol. Chemie, 19, 473 (1894). ^ R. VON Jaksch: Khnische Diagnostik, Ftinfte Auflage, Berlin (1901). (EDEMA 195 not only in starvation, but in the various forms of scurvy that are observed clinically, and the experimental types that may be induced in animals. What evidence have we for the abnormal production or accumulation of acids in all these conditions? We are, first, not without clinical evidence. The observations on hmnan beings undergoing a voluntary fast all agree in showing that the urine grows progressively more acid with each day of starvation. The only exception to this rule was noted by LuiGi LuciANi ^ in his study of Succi during a thirty days' fast. For the first six days of Succi's fasting there was a gradual increase in the acidity of the urine; for the rest of the period it remained very high in spite of the fact that he consumed large amounts of alkaline mineral water .^ A. E. Weight^ has made a further observation of interest. He noted a diminution in the capacity of the blood to neutralize acid in seven cases of scurvy. Yet more convincing are the experimental studies on starv- ing animals. The normally acid urine of the carnivora becomes more intensely so with progressive starvation, and in herbivora, the normally alkaline urine becomes highly acid. The same occurs if animals (especially herbivora) are fed an exclusive diet of any sort. An exclusive oat diet, which is high in acid salts and low in calcium, is quickly fatal. H. Weiske ^ found that when certain mineral salts, especially calcium salts (which are peculiarly powerful in neutrahzing the effects of acid), are added to the pure oat diet the animal fares better. A thorough study of starvation and such one-sided diets, rich in acids and poor in calcium, has been made by Axtel Holst and Theodor Frolich ^. They describe as constant findings in their experi- ments the occurrence of oedema. " A pronounced universal anasarca " was noted in all the starved animals, while those fed exclusively on oats, barley, wheat, or some of their derivatives showed various degrees of oedema up to such universal anasarcas. ^LuiQi LuciANi: Das Hungern, 164, Hamburg und Leipzig (1890). ^ Succi's fasting period was long enough to have allowed of the develop- ment of an oedema, and yet none is noted in Luciani's account of the case. I attribute this to the beneficent effects of the mineral water he consumed. See the effects of alkali and salts on cedematous states as given below. 'A. E. Wright: Lancet, 2 (1900). . (M C^ N . to -* IM s - «0 CO U3 tJ* t> 00 00 W 1 i 1 1 1 ^ 00 »o »o lO t> ■^t* CM 00 O CO CO CI gg++++ ■»J< -!)< t^ 00 C3i CO O O O f-< CM »0 ^, lO CQ CM 1-1 O M ■* Oi r + + + + CO "* lO tH CO £? Co' co' 00 O) I> o ! 1 1 1 1 o-j 00 00 o r-l O 00 ^ CO CO CO CO N CO 6S S S" ^ ;:? "* o rH lO W CO " 00 00 00 o 1 1 i 1 1 1 CO 1> IC CO Tt* CO o ^ IQ Tjf O O O CO CO CO CO CO CO CO Is ^ + fe? CO Ih o o o ?? i-H i-H CO CO CD t^ S' + + + + + + rH CD lO ■^ Hi 3 u rf i9, fi aw 6 Is iC "S "5 *"" $ + 3 ffi a + >> o y B K g O s + 3 o ^ OJ 00 CO b- ffi CO CO P.J- 1 r* 00 1 1 b- r* 1 1 CO OJ CO Tj- C^ lO ^ CO CO CO CO CO CO CO CO CO CO CO 6? o 00 C^I CD ■* O i-< O (N CO O O t> ^ (N M + + + + + + o 00 CO OS lO O CD CO 00 00 o N in 00 CO CO CO ■*■*■* ■* 1-H 00 O >-< CD O C^- i-i O Tf 00 O CO S++++++ CO O C31 CO CO C^ O CO ■<*< CO lo CO r- 00 CO CO CO CO CO CO CO 6? s ;:;• ^ 00 ;:: s ^ Tt< ■* Tj* lO lo o 1 1 1 1 1 1 o CD 05 o 03 Oi M 00 00 1> 00 GO CO (M N (M (N (M CQ 6S o CO >0 CO OS (N -! O CO « m o + 7.0 +19.0 +28.7 +33.6 > 65 % 67 (+ 3.0) 79 (+21.5) 84C+29.2) 87 (+33.3) > 52 % 64.5(+ 4.8) ■ 59 ( + 13.4) 64 (+23.0) 65 (+26.0) Dead > 42.5 % 46 (+ 7.6) 51 (+20.0) 53 (+24.7) Dead »— 1 l-l l-H 42 % 44(+ 4.6) 49 (+16.6) 55 (+30.9) 67 (+36.0) " 39 % 42.5 (+ 8.9) 47 (+20.6) 62 (+33.3) 62 (+33.3) h-l 33.6 % 38 (+13.4) 41 (+22.4) 44 (+31,3) 47 (+40,3) Dead e 3 o M o lo lo o CO w T-H CO o 00 d ■* 00 rt ^ CO 00 220 (EDEMA AND NEPHRITIS Incidentally these experiments show that contrary to much chnical teaching, administration of sodium chlorid does not increase an existing cedema. We shall have occasion to return to this question later, but we already see that sodium chlorid is no exception to the general rule that the presence of salts leads to dehydration of protein colloids. The effect of sodium chlorid in reducing an oedema is evident to mere inspection. In Fig. 84, a and h, is shown the untreated Frog 3 of Experiment 11, photographed at the time of injection and forty-two hours later. Fig. 85, a and 6, shows Frog III treated with sodium chlorid and similarly photographed. 4. On (Edema Due to Other than Acid Causes Unfavorable criticism of the colloid-chemical theory of water absorption as applied to the problem of oedema has spent its force but little upon the fundamental contention that the hydro- philic colloids and their state of increased hydration character- izes this pathological entity, but rather upon the entirely sub- sidiary one of the mechanism by which the normal hydration capacity of any hydrophilic tissue colloid is raised to the degree accepted as characteristic of " oedema." In discussing this lesser phase I have laid chief stress on the importance of an abnormal production and accumulation of acid in the affected part. My critics have for various reasons, adequate in their judgment, denied the effectiveness of this factor, but, with the exception of W. J. Gies, none has contributed any new sugges- tions as to what might be responsible for the increased hydra- tion if a production or accumulation of acid were not. Let me first reemphasize that I have never held an acid production and accumulation to constitute, of necessity, the only factor responsible for the increased hydration which charac- terizes cedema. I pointed out even in my first papers ^ that a conversion of colloid material of one type into another of a more highly hydrated type might lead to oedema, and empha- sized the importance of Wolfgang Ostwald's findings in this direction, according to which Beta-gelatin swells more than the ordinary kind. Since Beta-gelatin is a partly hydrolyzed one, and since this change has much in common with the first changes ^ See the bibliography at the end of this volume. (EDEMA 221 P a 222 (EDEMA AND NEPHRITIS S! p a (EDEMA 223 of protein digestion under the influence of proteolytic ferments, it was but natural that the possible importance of these in the production of oedema should come to mind. William J. Gies ^ has since insisted upon this point anew. The ultimate acceptance or rejection of the importance of this element in the problem must depend upon experiment. Thus far it has not been demon- strated that the ferments play any great role in increasing the water content of proteins, and there are absolutely no sugges- tions at hand as to how in oedema the chemistry of normal cellu- lar activity becomes so upset as to allow the ferments then to do things which they do not do normally. The quantitative experiments of Grover Tracy, Frank R. Elder and William J. Gies 2 — and such only can tell us of the relative importance of different agencies in hydrating colloids — show proteolytic ferments to increase water absorption under optimal conditions only some three to six parts, when acid alone has already made the proteins take up from seventy to eighty times their original weight. But what is of greater significance in the biological aspects of the problem are the more recent experiments of Edgar G. Miller, Jr., and Gies^ which confirm an older observation of my own* that tissues exposed to autolytic changes swell no more than fresh tissues subjected to the same external conditions. Beyond this my critics have been entirely barren of any suggestions regarding other agencies which in biological material might be capable of increasing its hydration capacity. In col- loid chemistry we know, of course, a number which increase the hydration of protein colloids, but it is quite another matter whether these ever appear in living material in sufficient amounts or are there sufficiently active to give rise to such cedemas as are encountered clinically. The only way in which such questions can be settled is by direct animal experiments. Such are simple enough in themselves and it would seem an easy matter to determine whether or not certain substances which increase the hydration of proteins in vitro do this also in living 1 William J. Gies, Biochem. Bull. 1, 312 (1911). 2 Groveb Tracy and William J. Gies: Biochem. Bull., 1, 467 (1912); Frank R. Elder and Gies: ibid., 1, 640 (1912). 'Edgar G. Miller, Jr., and William J. Gibs: Biochem. Bull., 1, 475 (1912). *See page 209; or page 111 of the first edition of (Edema. 224 (EDEMA AND NEPHRITIS animals. But it is not always easily proved that in the latter case the given substance did this directly, and not indirectly by interfering with the normal oxidation chemistry of the body cells, or by interfering with normal cardiac or respiratory activity. Tentatively, however, the following facts may prove of interest in connection with the general problem of cedeina production in animals by other means ^ than by the introduction, produc- tion or accumulation of acids in them. It does not surprise us that frogs develop an oedema if kept in solutions containing alkali in toxic amounts. Alkali effectively hydrates protein in vitro and it does this in vivo. To shut out the factor of acid production (through muscular work) conse- quent upon the movements of the frog in the alkaline solution it is well to destroy the brain. The accumulation of urea (and other nitrogeneous products) in the blood and tissues of patients likely to show oedema (as in the nephritides) and its proved capacity of increasing the hydra- tion of protein, suggested that an accumulation of urea might be one of the factors in determining the development of oedema in certain clinical cases. Experiments on frogs as thus far per- formed do not, however, lend any support to such a conclusion. If urea is of importance in leading to a development of oedema its role from a quantitative standpoint is rather small. Its biological importance from a qualitative standpoint because of the peculiar type of hydration which it produces may, however, be great. Both pyridin and some of the amins (ethyl and diethyl amin) lead to marked oedema in frogs. The amins are peculiarly powerful in their action, exceeding both in rate of development and in the intensity of the oedema produced all alkalies of the same concentration. This observation seems to me of funda- mental importance in connection with the oedemas accompany- ing some of the infections, for the amin character of many of the toxins is a proved matter. A series of important observations made by Allan Eustis ^ fit in here. Eustis has for several years insisted on clinical ^See Martin H. Fischer and Anne Sykes: Science, 38, 486 (1913); Jour. Med. Soc, New Jersey, 11, 116 (1914). 2 Allan Eustis: Am. Jour. Med. Sci., 143, 862 (1912); New Orleans Med. and Surg. Jour., 66, April (1914). (EDilMA 225 grounds, as have others, upon the importance of protein deriv- atives in the production of certain oedemas such as bronchial asthma and urticaria. His clinical experience has, however, been supported by experiment, in that he has shown that beta- imidazolylethylamin (a putrefaction product of the histidin normally produced in digestion) when applied to the slightly broken skin is followed by an intense urticarial-like eruption. IV ON THE PASSIVE CONGESTION (EDEMAS OF THE KIDNEY AND THE LIVER It is our next problem to discuss some of the special aspects of oedema, as determined by the fact that the oedema affects cer- tain organs or constitutes a particularly prominent feature of special pathological or clinical states, etc. Thus, nephritis is in good part an oedema of the kidney, glaucoma an oedema of the eye, uremia an oedema of the brain; while a deficient urinary output, an increased intraocular tension or coma are but the physiological parallels if not the direct expressions of these oedemas. Because of their great clinical interest, the questions of glaucoma and of nephritis with its associated problem of uremia receive detailed discussion later. Here we want to touch upon some special aspects of oedema which are not discussed there. In addition to bringing further evidence for the colloid-chemical view of oedema these remarks will serve to illustrate how I think we may correctly interpret certain well- known and long-recognized pathological and clinical facts. How, on the basis of the foregoing discussion, are we to regard the so-called passive congestion oedemas of the kidneys and the liver? In consequence of interference with the outflow of blood from these organs, be this due to a merely local dis- turbance, such as pressure upon the efferent vein, or to a more general one, such as heart disease, they become overfilled with blood and a general increase in the size of the organs is noted. This increase in size is independent of the accidental presence of an excessive amount of blood in the organ; it is due, in other words, to an increase in the size of the individual cells and tissues themselves — an oedema. For this oedema of the parenchymatous 226 OEDEMA AND NEPHRITIS organs the same factors of increased blood pressure, increased permeability of blood vessel walls, etc., that are so familiar to us from our previous considerations, have again been held respon- sible. In fact, the deductions made from consideration of the well-defined passive congestion oedemas of various organs as observed clinically or produced experimentally may be said to have colored our whole conception of the essential nature of all classes of oedema. There exists no dearth of isolated experimental and clinical observations on the passive congestion oedemas of the liver and the kidney, but the attempts that have been made to correlate them can hardly be said to be successful. Adherents of the pressure theory of oedema, for example, cannot meet the gross facts that evidence of any increase in blood pressure is all too often absent in patients with marked " congestion " of the kid- neys or liver; that swollen, passively congested organs decrease in size after the use of drugs whose chief action makes for an increase in blood pressure; and that enormous experimental increases in blood pressure in animals do not lead to oedemas of these organs. On the other hand, believers in the increased permeability of blood vessel walls have never proved their point physico-chemically; nor have those who have recently resur- rected the role of hydremia forty years after Cohnheim buried it experimentally. On the colloid-chemical basis we interpret the phenomena that characterize the passive congestion oedemas of kidneys or liver as follows: The cause of the oedema is again to be sought in the tissues. The circulatory disturbances leading to an cedema of these organs all have this in common: they bring about a state of oxygen want in the tissues in consequence of which acids are produced in them. These increase the capacity of the tissue colloids for holding water, whereby they are enabled to absorb an increased amount from any available source. This idea is supported by the following: § 1 It is a well-known fact that when the efferent (renal) vein of the kidney is tied in animals, the organ becomes filled with blood, while the kidney tissues proper swell and become pro- gressively firmer in consistence. This is the typical picture of (EDEMA 227 a passive congestion sufficiently severe to permit of the develop- ment of an oedema in the congested area. We need not repeat that what happens in this experiment is usually interpreted as an oedema due to an increased blood pressure, alterations in vascular permeability, etc. All these explanations fall as soon as it is stated that ligation of the renal artery leads to the same scries of changes in the kidney as ligation of the renal vein (with f •tt^ !> A B Figure 86. — A, normal right kidney; B, oedematous left kidney of the same rabbit twenty-three hours after Ugation of the renal artery. Experi- ment "IV" of May 13, 1909. the exception of the overfilling of the blood vessels). (See Fig. 86.) An abstract of a few experiments carried out with Gertrude Moore ^ on rabbits may serve to illustrate this point. In a series of nine Belgian hares we Ugated the left renal vein in three and the left renal artery in the remaining six. The operations were made under morphin anesthesia, and in no case consumed more than five minutes. None of the operations was complicated by infection. At various periods after the operations the 1 Martin H. Fischer and Gertrude Moobe: Kolloid-Zeitschr., 5, 286 (1909). 228 OEDEMA AND NEPHRITIS animals were killed and the two kidneys of each animal weighed. As clearly apparent from the following tables, the increase in the weight of the kidney after Hgation of the artery is quite as great as after ligation of the vein. The extra amount of clotted blood found in the kidney when the veins are ligated easily accounts for the somewhat higher values found in Table LXIV over those found in Table LXV. TABLE LXIV Ligation of Left Renal Vein Rabbit. Weight of rabbit. Hours after ligation. Weight of kidneys. Gain in weight in % of Normal. Ligated. normal. May 13, '09, "VI". .. May 1, '09, "C" May 13, '09, "V" 930 1500 811 19.20 22.10 42.40 4.20 7.50 3.80 7.00 12.70 13.95 1 66.6% 69.3% 267.1% « 1 This unusually high figure was due to the fact that an enormous extravasation of blood into the capsule with oedematous swelling occurred in this case. TABLE LXV Ligation of Left Renal Artery Rabbit. Weight of rabbit. Hours after ligation. Weight of kidneys. Normal. Ligated. Gain in weight in % of normal. May 13, '09 "I". . May 13, '09, "III" May 13, '09, "IV" May 1, '09, "A". May 13, '09, "II". May 1, '09, "B". 970 1127 1115 1500 734 1560 4.25 19.00 23.00 23.00 42.15 48.00 5.60 5.07 4.75 7.43 3.63 7.40 6.50 7,95 7.42 11.70 6.65 1 10.65 16.1% 56.8% 56.4% 57.3% 83.2% 1 43.9% 1 This high value was due in part to an extravasation of blood into the capsule with cedematous swelling. Note that the escape of blood occurred after ligation of the artery (diapedesis without blood pressure!). A decrease in blood pressure is, therefore, quite as effective in bringing about an oedema of the kidney as an increase. While such a result is unexplainable on the basis of the widely accepted pressure theory of oedema it is not surprising to us. In fact, such an experimental result was anticipated. Ligation of the vessel which carries the arterial blood to an organ must of necessity lead to a state of oxygen want in the tissues quite as readily as hgation of the vessel which carries the venous blood away. (EDEMA 229 We turn now to the liver, where, owing to the anatomical peculiarities of its vascular supply, valuable conditions are offered for experiments with which to test further this colloid concep- tion of oedema. The kidney is supplied with blood through the renal artery, which it wiU be recalled is very large as compared with the size of the organ. The physiological purpose of this anatomical arrangement is not far to seek. Through this artery there must pass to the kidney not only enough blood to supply the kidney tissues with oxygen, but all that blood from which the kidney separates the urine. In the case of the liver the blood supply is quite different. Through the venous portal blood the char- acteristic functions of the liver are subserved; through the arterial blood furnished by the hepatic artery the parenchyma is supplied with its necessary oxygen. Both these streams unite to leave the hver through the hepatic vein. When now a liver becomes decidedly cedematous from passive congestion, say in consequence of a heart lesion or pressure upon the hepatic vein, how is this result to be interpreted? After our remarks on the essential role played by oxygen want, and not mere blood pressure changes in the production of oedema in the kidney, the well-known fact that ligation of the portal vein is not followed by an oedema of the liver does not surprise us — the portal vein carries only venous blood to the liver, and so changes in its parenchyma due to a production of acids are not to be expected. Quite a different picture is obtained when the hepatic artery is ligated. In spite of the fall in blood pressure brought about by this means the liver rapidly develops an intense oedema. This result is quite expected on the basis of our theory, and indicates clearly that the real reason why a passive conges- tion leads to an oedema of the liver is because it interferes with the necessary flow of arterial blood through the organ via the hepatic artery. The following four experiments show how quickly ligation of the hepatic artery in rabbits leads to oedema of the liver, and how severe this is. The oedema follows ligation the more rapidly and is the more intense the more perfect the ligation of the various branches that constitute the hepatic artery in this animal. The operations were again made under morphin anesthesia, in ten to fifteen minutes and without infection. The 230 CEDEMA AND NEPHRITIS increase in the size of the liver, while readily apparent to the eye, can be expressed numerically only by indirect calculation of the weight of the liver in percentage of the body weight of the operated animal. In a series of six normal rabbits we found the Uver to constitute 2.9 per cent, 3.2 per cent, 3.5 per cent, 3.7 per cent, 3.7 per cent, and 3.7 per cent (an average of 3.45 per cent) of the total body weight. Table LXVI shows how much the liver is increased in size when the hepatic artery is ligated, TABLE LXVI Weight Hours Weight of Per cent Rabbit. of after liver at of body Remarks. ' rabbit. ligation. autopsy. weight. May 14, '09, "VIII" 845 13 48.5 5.7 One well-defined artery to the liver. May 1, '09, "XX" 694 16 37.3 5.3 One well-defined artery to the liver. Mayl4, '09, "IX".. 867 18 35.7 4.1 Artery has several branches, one only ligated. May 10, '09, "Red". 764 23 33.2 4.3 Several small branches li- gated. § 2 Having shown by these simple means that the kidneys and liver become oedematous in consequence of various circulatory disturbances only because these disturbances lead to a state of lack of oxygen in the tissues, we have now to say what are the consequences of such a state of affairs. We are especially interested in evidence which shows that under such circum- stances an abnormal storage or production of acids occurs in these tissues. Thp accumulation of carbonic acid in the cells and tissues is a necessary consequence of any interference with the outflow of blood from (or inflow to) a part. The abnormal accumula- tion and production of other acids in both kidneys and liver in consequence of a disturbance in oxygen supply to these viscera has, however, been proved directly for these organs by the already quoted experiments of Trasaburo Araki ^ and Hermann ZiLLESSEN.^ Ligation of the renal artery or vein, or ligation of the hepatic artery, is always followed by the production of 1 T. Abaki: Zeitsohr. f. physioL chem., 15, 335 and 346 (1891); ibid., 19, 422 (1894). See also Hoppb-Seylbr, ibid., 19, 476 (1894). 2H. Zillessbn: Zeitschr. f. physiol. chem., 15, 387 (1891). (EDEMA 231 lactic and other acids in the kidneys and the liver. In these acids, the (hydrophilic) protein colloids of the tissues, and an available source of water, we have, therefore, all the conditions necessary for the development of an oedema. When the renal vein is tied, the available source of water is found in the blood that attempts to enter the kidney through the renal artery and stagnates in the kidney; when the artery is tied the increased hydration capacity of the tissue colloids is satisfied by absorbing water from the blood that backs into the kidneys through the veins. In patients with passive congestion of the kidneys a ready source of water is, of course, found in such a circulation as continues through these organs. When the hepatic artery is tied, or in clinical cases of passive congestion of the hver, a plentiful source of water is found in the portal circulation and the blood from the hepatic vein. On the basis of these colloid conceptions of water absorp- tion we have also now no difficulty in understanding the well- known physiological fact that an accumulation of carbonic acid in the arterial blood supply to the kidney, or any interference with the normal oxygen-carrying power of the blood unaccom- panied by any changes in blood pressure, leads to an increase in the size of the kidney, in other words to, an " csdema," while an abundant oxygen supply brings about the reverse result. We can also understand why the kidney enlarges in the various toxic forms of nephritis. The toxic agents — various toxins, snake venoms, cantharidins, metallic salts, etc. — all belong either directly in the group of the reducing bodies or can be shown experimentally to interfere with the normal oxidations of living cells. But interference with these oxidations is followed by the same consequences as ligation of an artery or a vein, so that a swelling of the kidney cells is a logical result. The enlarge- ment of the liver and the kidneys in phosphorus poisoning can be explained on the same ground. The increase in the size of the liver in phosphorus poisoning is not due primarily to an excessive deposition of fat, but to an increased amount of absorbed water. §3 The colloid theory seems able to harmonize satisfactorily observations which on the basis of other theories seemed con- 232 OEDEMA AND NEPHRITIS tradictory, while it correlates at the same time a series of apparently unattached clinical and experimental facts. Much interest attaches itself in this connection to a series of observa- tions by H. J. Hamburger ^ on the kidney and the Hver; and by Waichi Hirokawa 2 on the kidney. In studying the " osmotic " behavior of the kidney, Ham- burgee found all the diameters of the isolated kidney to increase when he perfused it with blood serum to which an acid had been added, and to decrease when he replaced the acid with an alkali. Isolated kidney cells behave similarly. They swell in water and in weak salt solutions. In sufficiently strong solutions of neutral salts they keep their normal size or even shrink. In these experiments the action of acids and alkalies is entirely unintelhgible on the osmotic basis of water absorption, and close scrutiny reveals unexpected disparities between observed and calculated effects of the different salt solutions. Our colloid theory fares better. The perfusion with acidified serum leads to an increased hydration capacity of the tissue colloids. When the serum is alkalinized the acid present in the kidney is neu- tralized and through the simultaneous reduction in acidity and the production of salts in the tissues the hydration capacity of the colloids is decreased, hence shrinkage of the organ. The isolated kidney cells suffer oxygen want and become acid after removal from the body. For this reason they swell when placed in distilled water. Salt solutions counteract this swelling, and this the more the higher the concentration of the salt, as experiment proves. The whole series of phenomena is identical with that previously described in our experiments on the swelling of fibrin. Hirokawa also studied the " osmotic " behavior of kidney cells. He found blocks of kidney tissue progressively more capable of absorbing water from increasingly stronger solutions of sodium chlorid the longer the blocks of tissue had been out of the animal. Hirokawa correctly attributes this finding to the postmortem production of acid in the tissues, which he believes to increase the capacity of these tissues for holding water in the 1 H. J. Hambukqeb: Osmotischer Druck und lonenlehre, 3, 52, and ibid 50 and 54, Wiesbaden (1904). ^ W. Hirokawa: Hofmeister's Beitrage 2. chem. Physiologie, 11 458 (1908). (EDEMA 233 same way that K. Spieo found the capacity of gelatin plates for holding water raised through the addition of a little acid. Kidney cells, therefore, can become " cedematous " when entirely removed from the body; in other words, when entirely away from any vestige of a circulatory system. It only remains for us to connect the behavior of these " dead " kidneys with that of the passively congested " living " ones in an animal. This is done as soon as we recall the fact that the postmortem production of acids in the tissues, and that which occurs in the absence of an adequate oxygen supply represent identical processes. Experimental data are at hand which show that in the entire absence of any circulation, liver cells also may show all the signs of an oedema that we are accustomed to look for in an autopsy. H. J. Hamburger ^ found isolated liver cells and blocks of liver tissue to swell in water and in dilute salt solutions, and to maintain their volume or even shrink if stronger salt solutions or dilute alkalies were added. Dilute acids, including carbonic, markedly increased the sweUing. These remarks again parallel the effects of acids, alalikes and salts on the swelling of fibrin or gelatin in a faintly acid solution. V ON THE NATURE AND CAUSE OF PULMONARY (EDEMA The most generally accepted explanation of pulmonary oedema is that of William H. Wblch,^ according to whom it is due to " a disproportion between the working power of the left ventricle and of the right ventricle of such character that, the resistance remaining the same, the left heart is unable to expel in a unit of time the same quantity of blood as the right heart." It is readily seen that this theory is a mechanical one which assumes that through a heightened pressure of blood within the pulmonary circulation fluid is squeezed into the tissues of ' H. J. Hamburger: Osmotischer Druck und lonenlebre, 3, 50 and 54, Wiesbaden (1904). 2 William H. Welch: Virohow's Archiv, 72, 375 (1878); the quotation is transcribed from a letter to S. J. Meltzer, American Medicine, 8, 195 (1904). 234 (EDEMA AND NEPHRITIS the lung and out into the alveoli and bronchi. According to this conception pulmonary oedema is placed in the general group of Julius Cohnheim's ^ congestion oedemas. Welch's ideas have not gone unchallenged. Through the observations of various authors, particularly H. Sahli^ and M. LowiT^ it has been proved beonyd doubt that the severest grades of pulmonary oedema may • exist chnically and be produced experimentally without any evidence of an increased pressure in the pulmonary circuit. Welch's theory has in consequence been variously modified or cast aside entirely. We hear again of " increased permeabiUty of blood vessel walls," of " hydremia," of " secretory " disturbances, of still more vague " irritations," and, when all these fail, of changes in the pecuhar " life " of the cells themselves. The views held by the various authors are so divergent and at times so flatly contradictory that a detailed discussion of them is purposeless. The vagueness of these theories stands in sharp contrast to the really excellent experi- mental and clinical observations that are available. A unifying interpretation of these is still lacking. Toward such the follow- ing is offered: The problem of pulmonary oedema is identical with the problem of oedema in such an organ as the liver. The reason for this is at once apparent when we call to mind that the vascular arrangement in the lungs is similar to that which we previously discussed for the Hver. Just as the liver, so is the lung supplied with two blood streams — with a venous stream through the pulmonary artery, which only passes through the lung for purposes of oxygenation, and an arterial stream through branches from the thoracic aorta, the bronchial arteries, which supplies the parenchyma of the lung with oxygen. The blood brought by these nutrient arteries leaves the lung in part through the bronchial veins, in part admixed with the blood of the lesser circulation through the pulmonary veins. The facts at hand on the experimental production of pulmonary oedema are easily interpreted by saying that an oedema results whenever the oxygen supply to the parenchyma of the lung is sufficiently interfered with. 1 Julius Cohnheim: AUgemeine Pathologie, Zweite Auflage, 1, 501, Berlin (1882). 2 H. Sahli: Arch. f. exp. Path. u. Pharm., 19, 431 (1885). 1 M. Lowit: Ziegler's Beitrage, 14, 401 (1893). (EDEMA 235 If the pulmonary artery passing to one lung is ligated, no oedema results. If, in addition, most of the branches passing to the opposite lung are similarly treated, we still get no cedema. Enough circulation needs only to be maintained through the lung to keep the animal ahve. This result is, to our mind, entirely to be expected, for such ligations do not interfere with the oxygen supply to the parenchyma of the lung. Ligation of the pul- monary veins may lead to oedema of the lungs, but only if suf- ficiently extensive to shut off inost of the blood as it returns from the lung. In other words, it is not an easy matter to dam back the blood in the bronchial arteries (which discharge in part into the bronchial veins, in part into the pulmonary veins) by hgating only the pulmonary veins. These experiments show that interferences with the pulmonary circulation itself are, on the whole, scarcely able to lead to an oedema of the lung. The most effective way to bring about a pulmonary oedema is to disturb the systemic circulation. Compression of the left ventricle leads to pulmonary oedema, as does ligation of the aorta either at its root or not lower than the point of origin of the left subclavian artery. Ligation of the thoracic aorta low down, or of the abdominal aorta, does not lead to pul- monary cedema. These undisputed experimental facts are hard to understand on the basis of any pressure theory. While a rise of blood pressure in the pulmonary circuit may well be present in all these experiments, why should it be more effective when induced through ligation of the aorta than through direct hgation of the pulmonary vein? And why should ligation of the aorta to just below the left subclavian artery lead to a pul- monary oedema, and hgation just a little lower down be inef- fective? Only a few small arteries are given off by the thoracic portion of the descending aorta. We experience no difficulty in interpreting all these findings when we recall that the bronchial arteries leave the aorta just below the left subclavian. Compression of the left ventricle and ligation of the aorta to just below the sub- clavian all spell a lack of oxygen for the lung parenchyma, and hence an oedema. A ligation just below the bronchial arteries is without effect in this regard. These experiments show that a pulmonary oedema develops under the same conditions as an oedema anywhere else — whenever the lung parenchyma is placed in a state of lack of oxygen. This 236 CEDEMA AND NEPHRITIS state of oxygen want we always discovered to be important in other organs because it led to an abnormal accumulation or production of acids in the tissues. That such conditions prevail when the lungs become cedematous is borne out, not only by the fact that a pulmonary oedema is never induced in any animal by the various ligations described above without gross evidences of improper aeration of the blood, but by the following facts regarding chemically induced cedemas, and the oedemas of excised lungs. PoKROWSKY, Fkiedlander, and Herteb ^ found that rabbits and dogs which had breathed for some time an atmosphere rich in carbon dioxid showed grades of pulmonary oedema at autopsy which varied from such as were scarcely recognizable to such sufficiently intense to kill the animals. (Edemas have also been noted after inhalation of the fumes of various other acids. Other chemical methods of inducing a pulmonary oedema lead to a state of lack of oxygen and acid production in the tissues in a more indirect way. Under this heading come hydrocyanic acid, various ethers and anesthetics, carbon monoxid, adrenalin, and iodin — all of them substances which we know interfere markedly with the normal oxidations of living cells. The clinical evidence that pulmonary oedema is more often an accompaniment of the oedema of nephritis than of the oedema of heart disease is also easily understood on the basis of this chemical origin of pulmonary oedema. In nephritis we have the toxic bodies which are responsible for the oedema of the body tissues generally, more or less uniformly distributed throughout the various tissues and the blood. The parenchyma of the lungs is therefore as likely to be affected by these toxic bodies as the parenchyma of any other organ of the body. (But the pulmonary oedemas seen in patients with blood vessel disease are not at once to be attributed to a nephritis which they may show. They represent more probably the direct consequence of vascular disease involving the bronchial arteries.) In heart disease, on the other hand, the severity of the oedema of any organ is dis- tinctly dependent upon the quality of the circulation going through this organ which in turn determines the amount of oxygen furnished the organ and the readiness with which the carbonic ' Cited from Cohnheim, Allgemeine Pathologie, Zweite Aufl., 1, 502, and 2, 273, Berlin (1882). OEDEMA 237 acid formed in it is carried away. Generally speaking, the greater the distance of an organ from the left ventricle, the poorer must be its oxygen supply, and in consequence the greater its oppor- tunity to develop an oedema. In heart disease the lung is, there- fore, of all the organs, in the best position to be supplied even to the last, not only with the best oxygenated blood available, but with that lowest in carbonic acid. This explains why, in spite of much embarrassment in the pulmonary circulation, an oedema of the lung need not develop. It does not until the lung parenchyma itself suffers from lack of oxygen, a state not reached until an inadequate amount of blood, or an inade- quately aerated one is supplied through the bronchial arteries. Hence the so common terminal pulmonary oedema. Cohnheim has well said, " man does not die because he develops a pul- monary oedema, he develops a pulmonary oedema because he is dying." The gradually developing lack of oxygen and the accumulation of carbonic acid in the lungs in consequence of a gradually failing circulation and respiration account for it with- out difficulty. This conception of pulmonary oedema can be tested in yet another way. If the lung becomes oedematous from any con- dition which interferes with a normal oxygen supply to the parenchyma, then it ought to be particularly easy to produce an oedema in a lung when removed from the body. The most intense cedemas which simulate in every way those observed at the autopsy table may be produced in lungs removed from the body, and in the entire absence of any such blood pressures as are con- sidered active in the current theories of pulmonary oedema. The entire uninjured lungs of sheep freshly obtained from a nearby slaughter house, and with the heart intact, served for material in these experiments. As injection fluids, I have thus far used water, various salt solutions, dilute acids and these mixed with salts. As the experiments are not yet complete I describe only the effects of injecting water or m/6 (0.975%) sodium chlorid solution into the pulmonary arteries. With use of either of these fluids an intense pulmonary oedema results. The experiments are carried out in the following way: A cannula is first tied into the pulmonary artery; a ligature is next thrown about the heart below the cannula, and the heart cut off below this ligature. After adherent tags of tissue are removed. 238 CEDEMA AND NEPHKITIS the lung is weighed and hung up by a Hgature drawn through the trachea. If, now, a sodium chlorid solution or distilled water is simply allowed to trickle into a funnel connected with the glass cannula inserted into the pulmonary artery, the lung takes up enormous amounts of the fluid in a very short time. A lung weighing 500 grams will take up two to three liters of either of these fluids in an hour or two. What becomes of them is interesting. The lung tissue itself is first affected. It swells up enormously (more than doubling in weight after infusion for an hour or two), and in the earlier periods of the experiment, if the influx of fluid into the pulmonary artery is stopped, the lung may be turned upside down and not a drop of fluid flow out of either the blood vessels or the trachea. If the injection of fluid is continued the pleural surface after a time becomes moist, and soon a drop of fluid falls from the lower edge of the lung. This is soon followed by another and another until a steady drip is established which may amount to several hundred cubic centimeters of " pleural exudate " in the course of an hour. At the same time the lung can no longer be turned upside down without obtaining a bloody, frothy fluid from the trachea. This fluid gradually rises in the trachea, and if not removed, overflows. The overflow continues as long as the infusion of water or salt solution into the pulmonary artery is kept up (several hours) . Let it be noted that all this time not a drop of fluid comes out of the veins, even though these have not been ligated. If the infusion is properly regulated the tissues take up all the fluid that passes into the artery, absorb much of it themselves, and then allmo it to pass over into the alveoli and bronchi and through the pleura. Even after the infusion of liquid has been kept up for several hours, only a few cubic centimeters can be recovered from the blood vessels. From the experiments that have been carried out thus far it can be said that the longer the lungs have been out of the animal, the more quickly do these signs of a pulmonary oedema develop. Of the various injection fluids used, water leads to the greatest cedema of the parenchyma of the lung itself. When any salt solu- tion is used this is not so great, but the evidence of fluid in the bronchi is obtained earlier, and this " secretion " is more intense. Sodium citrate and sodium sulphate are more powerful in this regard than sodium chlorid. In other words, the salts which (EDEMA 239 dehydrate various protein colloids most are also most powerful in dehydrating the pulmonary tissues, and thus of permitting the greatest accumulations of fluid in the alveoli. We have thus far spoken of pulmonary oedema as a patholog- ical entity in the sense in which this term is ordinarily used in pathology. But for purposes of discussion and for the ultimate solution of the problem I believe that we will have to distinguish between the mere presence of an increased amount of fluid in the tissues of the lung proper, and the presence of fluid in the alveoli. While in the ordinary pulmonary oedema evidence of both is found, greatest weight is usually laid on the occurrence of fluid in the alveoli and bronchi. When this is present it undoubtedly repre- sents the extreme of what we are pleased to call a pulmonary oedema. But very severe oedemas of the lung may exist without any fluid in the .alveoli (as in the earlier periods of the oedemas produced in excised lungs) . The presence of an excessive amount of fluid in the lung tissues proper and the presence of abnormal amounts of fluid in the alveoli are rather to be regarded as asso- ciated, though not identical processes. ^Ve have no difficulty in interpreting all the phenomena of the oedema of the lung tissue itself on the basis of our colloid theory of water absorption. The tissues of the lung in pulmonary oedema come to hold an increased amount of water because acids are produced in them. Whether the possibilities for such an abnormal accumulation of acid are offered the lung by hgating various blood vessels in the body or by taking it out of the body is immaterial. That water absorp- tion really represents but an excessive hydration of certain protein colloids is again proved by the fact that all salt solutions inhibit the development of the oedema of the lung tissues proper, not only according to the concentration of the salt employed, but according to the character of the salt. The citrate and sulphate of sodium, for example, inhibit the absorption of water by the lung tissues themselves more than the chlorid. Yet just the reverse holds regarding the giving off of fluid into the bronchi. The explanation of the mechanism by which this water is given off is discussed, in part in the next paragraphs which consider the syneresis of colloids, in part in later chap- ters dealing with secretion. Why the different salts behave as they do we shall learn there. 240 CEDEMA AND NEPHRITIS VI SYNERESIS AND THE ACCUMULATION OF FLUID IN THE BODY CAVITIES IN (EDEMA As familiarly known, it is characteristic of the oedemas occurring in the higher animals, for fluid to accumulate in the body cavities. In the cedemas of heart disease, for example, we observe not only excessive quantities of fluid in the tissues themselves, but the pleural, pericardial and peritoneal cavities come to contain an abnormally great amount. This fluid is not water, but a colloid solution in which the proteins appear in lower concentration than in the normal body fluids (blood and lymph). Similar serous accumulations occur within the tissues themselves. It is generally said that a " transudation " of fluid occurs into the tissue spaces, such a space being regarded by many as a kind of miniature serous cavity. In truth no such cavities of course, exist; they are made by the serous fluid as this separates from the more solid (cedematous) tissues. How are these accumulations of fluid brought about? The explanation has been given by Wolfgang Ostwald ^ in directing attention to the syneresis exhibited by colloids. As first noted by Thomas Graham, hydrated colloids which were previously " dry " separate off liquid on standing. The separated fluid is not the pure solvent, but a dilute solution of the colloid. The classic example of this sort of change is seen in Fig. 87, where a silicic acid gel which originally showed no free fluid has, on standing, liberated the large amount shown in the photograph. In doing so the originally more highly swollen gel shrinks, as indicated by the space between the edge of the sohd colloid and the flask. What is important to us biologically is that pro- teins show the same type of change. Solid gelatin as well as other protein media, as the familiar blood serum of the bacteri- ologists, all squeeze off fluid containing protein on standing. The bacteriologists call this " water of condensation," but this is incorrect, for the fluid is really squeezed out by the protein. The more highly hydrated the protein colloid, the more fluid is squeezed off. This is shown in Fig. 88. Each of the flasks from left to right contains respectively 200 cc. of a 5, 4, 3, and 1 Wolfgang Ostwald: Personal Communication (1913). (EDEMA 241 2 per cent solid gelatin. The photograph was taken after the flasks had stood for 2| days in an ice chest. Separation of a dilute gelatin solution is evident in the flask on the extreme Figure 87. Figure 88. right and some has been freed in the flask next to it. No separation of liquid occurred in the more concentrated gelatin contained in the two flasks on the left. 242 (EDEMA AND NEPHRITIS The accumulation of fluid in the serous cavities and in the so-called tissue spaces in aedematous states represents the separa- tion of a dilute liquid protein colloid from the more solid, heavily hydrated ones making up the oedematous tissues themselves. It is the analogue of syneresis as observable in hydrated colloids. As degree of hydration and the time element are of importance in determining the amount of fluid that is thus squeezed off from laboratory colloids, so also do the high hydration char- acteristic of oedema and the time element, as determined by the chronicity of the agencies leading to the oedema, play important parts in the development of its accompanying " transudations." Incidentally, these remarks may suffice to answer the critic- ism first raised by W. J. Gies ^ and more recently repeated by Felix Maechand,^ Rudolf Klemensiewicz,^ C. Ziegler * and others according to which the colloid-chemical theory of oedema is inadequate because there is nothing in the behavior of col- loids to explain the mechanism of " transudation." VII CONCLUDING REMARKS It would be a task of purposeless length to review the myriad contributions of various authors to the facts and theories of oedema and attempt their reinterpretation in the terms of colloid chemistry. To satisfy some of my critics who insist on such kindergarten methods it may suffice to indicate the road which any such reinterpretation must follow. Suppose we choose for comment so simple a fact as that the injection of large quantities of " physiological " sodium chlorid solution is likely to be followed by some oedema in an animal. Does this prove that " increased blood pressure," " plethora " iW. J. Gies: Biochem. Bull, 1, 124 and 279 (1911 and 1912); other criticisms by Gies as well as my answers to them (ibid., 1, 444 (1912)), are also found here. See also F. G. Goodbidge and W. J. Gies: Proc. Soc. Exp. Biol, and Med., 8, 106 (1911), and my answer in the first edition of my "Nephritis" (page 184). 2 F. Marchand : Verh. d. deut. Naturforsch. u. Arzte, 1912. ' Rudolf Klemensiewicz: Verh. d. deut. Naturforsch. u. Arzte, 1912; see also, M. Kornbk: Transfusion, newly edited by Klemensiewicz, Leipzig (1913), where the latter's criticisms are stated more moderately. ^C. Zibglbb: Verh. d. deut. Naturforsch. u. Arzte, 1912. CEDEMA 243 and " hydremia " are the cause of oedema, as some insist to this day, forty years after Cohnheim and Lichtheim and their followers showed that no reasonable amount of injected fluid ever did this? I think not. In a long series of experiments on rabbits, made with an entirely different object in view, I found it necessary to inject intravenously such amounts of sodium chlorid solution as were used by these authors. I found invariably that if the injec- tions were only continued long enough the rabbits always devel- oped intense general cedemas. The cedema is in other words more a function of the time than of the amount of fluid injected. How are these oedemas to be interpreted? Simply by noting this: Rabbits subjected to such prolonged and great sodium chlorid injections suffer from lack of oxygen. In the later hours of the experiments this becomes so great that the animals are dis- tinctly cyanotic. As soon as we have such a state of lack of oxygen we have the conditions at hand that increase the capacity of the tissue colloids for holding water, as our previously detailed experiments have shown, and so they are in a position to absorb water from the circulating hquid in the blood and lymph vessels. Just why in such experimentally induced cedemas, the abdom- inal organs, for example, should develop the oedema sooner than the subcutaneous tissues is a matter that needs separate investigation. Predilection for certain regions of the body is characteristic also of various clinical forms of oedema (oedema of nephritis, oedema of heart disease). The colloids of different tissues are different, the demand for oxygen is greater in the glandular organs than in the connective tissues, etc. Just how the sodium chlorid injections produce the lack of oxygen also needs analysis. Simple dilution of the blood, the increased work thrown on the heart in pumping this blood, that thrown on the various glandular organs in separating the salt solution from what is the normal blood, the effect on respiration, etc., all have to be considered. Another fact is constantly overlooked in experiments on oedema made on the higher animals — the necessity of furnishing an adequate supply of water to the tissues. This is not easily controlled in mammals, and it is for this reason that I chose to do most of my experimenting with frogs, which may be dropped into water and so be allowed to absorb all they can 244 (EDEMA AND NEPHRITIS take up through the skin. As mammals cannot be relied upon to drink voluntarily as much water as we might like to have them consume, one is always in the predicament of wondering just how much water ought to be injected through the stomach tube, and in experiments in which only one part of an animal is supposed to become csdematous, an inadequate water supply means too often that the affected part, in order to become cedematous, must first rob some other tissue with a lesser affinity for water before it can satisfy its own needs. After our remarks on the r61e of the coltoids in oedema, it is, of course, self-evident that the over-consumption of water could not increase an oedema after the capacity of the tissue colloids for holding such has once been satisfied. There is also no difficulty in understanding why Cohnheim's experiments, in which he combined the infusion of sodium chlorid solution with moderate injury to a part, always led to the development of an CEdema in the part more promptly than infusion alone. The moderate injury (heat, sunburn, iodin application), simply brought about by indirect means, the so necessary change in the colloids of the tissues, and the increased capacity for holding water once established, the water of the sodium chlorid infusion quickly satisfied it. The increased swelling of protoplasm after mechanical injury, for example, goes down into the very elements of living matter. No more brilliant proof of this can be furnished than the observation of G. L. Kite ^ who found an immediate localized swelling (oedema) to follow the track of his glass needles when pushed into the protoplasm of isolated living cells when observed under the highest powers of the microscope. The interpretation of another experimental observation of CoHNHEiM 2 seems to me to need revision. Cohnheim found that an animal which had been bled repeatedly, and injected after each bleeding with a sodium chlorid solution, finally developed a general oedema, and interpreted this as an oedema of cachexia, caused through an increased permeability of the blood vessel walls, determined primarily through a hydremia. Would it not be simpler to say that through these frequent ' G. L. Kite, Personal communication, 1913. 2 J. Cohnheim: AUgemeine Pathologie, Zweite Auflage, 1, 498, Berlin 1882. ' (EDEMA 245 bleedings the animal became anemic — that is to say, its organs got into a state of lack of oxjgen — and when a supply of water was furnished the tissues, whether through a sodium chlorid infusion, or in any other way, they took this up? We need not further discuss the inadequacy of all blood or lymph pressure theories of oedema. While Cohnheim regarded blood pressure as one of the two great factors concerned in the production of oedema, he also recognized that severe oedemas occur when no change whatsoever in blood pressure is apparent. To account for them under such circumstances he had recourse to an " increased permeability of the blood vessel walls." If in the light of modern physico-chemical conceptions we try to say just what is meant by this, we have to define the blood vessel wall as a colloid membrane. From physico-chemical observations we know that the permeability of such colloid membranes is alterable, so this far Cohnheim is on safe ground. But of what consequence would an increased permeability of the blood vessels be from a pathological standpoint? To force liquids through the blood vessel walls is not to force them into the tissues. And the fluid of an cedematous tissue is very decidedly in the cells themselves. Cohnheim's hypothesis would simply squeeze the oedema fluid as far as the outer walls of the capillaries. If we try to aid Cohnheim's conception of permeability and make it extend to all protoplasm, then we get the cause of oedema right where we have tried to say it is, namely, in the tissues themselves; and then our problem is simply that of how tissues hold their water. In this the forces that have been suggested as active — not only the variable hydra- tion capacity of colloids, but even the previously suggested one of osmotic pressure, with or without Overton's conception of lipoid surface layers — are so infinitely greater than the highest grades of blood pressure that pathologists have ever registered that the two cannot be compared. The more recent experiments of Magnus have added much to our knowledge of the experimental side of oedema. His results, too, are usually interpreted as lending support to Cohnheim's conception of the increased permeability of blood vessel walls as a factor in the production of oedema. How well they support the belief that the cause of oedema is to be sought in a change in the colloid constitution of the tissues is readily evidenced by 246 CEDEMA AND NEPHRITIS the following. Magnus found that animals which are trans- fused after death always develop a general anasarca. Living animals do not do so as readily, but they do if deeply chloroformed or etherized or injected with arsenic. In place of these words we could write, placed in a condition of lack of oxygen with an adequate supply of water. With these remarks, which have been introduced simply to illustrate how I think the experimental results of the score of workers who have busied themselves with this problem of oedema should be interpreted, we will close our discussion. It is readily apparent that through experimental analysis the part played by the blood and the lymph circulations has gradually become less prominent. From having been looked upon as alone deter- mining the amount of water held by the tissues, we have come to find that the tissues are largely their own masters in this regard. The blood and lymph circulations carry fluid to the tissues and away from them, but what the tissues will take up or give off rests with them. Oiily as these circulatory systems carry to the tissues substances which directly threaten their existence, or fail to remove such as the tissues have produced, which if allowed to accumulate will overcome them, only in so far are the circulatory systems masters of the tissues. PART FOUR ABSORPTION AND SECRETION IN THE COMPLEX ORGANISM PART FOUR ABSORPTION AND SECRETION IN THE COMPLEX ORGANISM THE GENERAL PROBLEM The previous pages have dealt with the absorption and secre- tion of water from the point of view of the isolated cell, tissue, or organ. Our general conclusion has been that the tissues simply soak up a certain amount of water from the fluid medium in which they lie (the blood and lymph in the case of the higher animals), and that this amount is determined by the state of the colloids found in the tissues. Before we can advantageously proceed with a discussion of the special aspects of cedema we need to consider this problem of absorption and secretion from the viewpoint of the organism as a whole. How can we utilize the teachings of colloid-chemistry in this direction? The absorption and secretion of water by a multicellular organism seems at first sight to be decidedly different from the absorption and secretion of water as observed in a single cell — say an ameba or a muscle cell. It is easy to think of an ameba as a spherical mass of colloid material saturated with water, and under cha,nges in its physico-chemical surroundings or through direct changes in its own chemical composition so altering this colloid material as to make it take up or give off water. As I view it, this simple conception does constitute the heart of the entire problem of water absorption and secre- tion as observed in this animal. 249 250 (EDEMA AND NEPHRITIS But in a multicellular organism biological facts confront us which do not at first seem to be interpretable on any such simple basis. In the mammal, for example, we find whole organs set apart, seemingly endowed with powers of absorption only, while others function seemingly only as secretory organs. It becomes hard, for example, to see just what relationship ex- ists between a mucosal cell of the small intestine concerned almost exclusively with an absorption of water from the lumen of the gut, or a kidney cell concerned equally exclusively with a secre- tion of urine, and the ameba or muscle cell which now absorbs and now secretes water either in response to its own physiological demands or under the conditions with which experimentally we are pleased to surround it. And yet on closer analysis the difference between the two is not so striking. First of all, we need to appreciate that the mucosal cell is an absorbing cell only so long as we look at it from the side of the lumen of the gut. If we regard it from the blood vessel side, the mucosal cell is a secreting cell, for what it absorbs from the gut it gives up to the blood. Similarly, the kidney cell is a secreting cell only because we usually look at it from the point of view of being a producer of urine — as a matter of fact, every- thing that goes to make up the normal urine was absorbed from the blood. But even if we look at the matter from the nar- rower point of view, the intestinal cells under certain circumstances become secreting cells in that they secrete substances into the lumen of the intestine; and according to the judgment of some authors, certain kidney cells may reabsorb materials that have been secreted by others. In essence, therefore, secretion and absorption in the higher animals is not different from absorp- tion and secretion as observed in an ameba or any isolated tissue cell. That which remains, therefore, to characterize absorp- tion and secretion in the higher animals is merely this, that under normal circumstances and viewed from the point of view of the organ- ism as a whole, absorption and secretion occur predominantly in one direction. What requires special analysis in the higher animals is, therefore, not absorption and secretion per se, but the condi- tions existing in the multicellular organism which make it possible for certain organs to act chiefly as absorbing systems, while others act predominantly as secreting systems. This is what creates all the problems that are conveniently grouped ABSORPTION, SECRETION— COMPLEX ORGANISM 251 under the general heading of the special physiology of absorption and secretion, as observed in the higher animals. Let us see, first of all, if we cannot define in general terms what must be the conditions which lie at the bottom of this predominant functioning of certain cells and tissues in one direction, and this on the basis of our belief that the colloid con- stitution of the living cell is primarily responsible for the phenom- ena of water absorption and secretion by the cell. An ameba or an isolated cell or tissue derived from a higher animal and kept in a solution of any kind is surrounded by this solution on all sides. Could we imagine the chemical processes within these cells held in abeyance, then we see how they would after a time succeed in getting into a state of equilibrium with their, surroundings. When such an equihbrium has been established, the cells neither absorb nor secrete water. Only as this equilibrium is disturbed, either through changes in the surroundings of these cells or through the specific chemical changes occurring in the cells, can we expect a renewed absorption or secretion. Under quite different conditions do we find the individual cells of the multicellular organism existing in the intact living body. While in a certain sense the internal activities of the ameba may be compared with those of the individual cells mak- ing up, say the intestinal mucosa, and there exists a certain analogy between the fact that both are surrounded by a liquid medium, here the analogy stops. For while the ameba is sur- rounded on all sides by the same liquid medium, the cells of any of the absorptive or secretory organs, found for instance in a mammal, are through different portions of their cell protoplasm in contact with entirely different media. The cells constituting the intestinal mucous membrane, for example, are bathed on one side by intes- tinal contents; on the other by blood or lymph, or both together. Such cells, like any other absorptive or secretory cells similarly situated, find themselves, therefore, in the predicament of try- ing to get into equihbrium with as many different media as surround them. It is in trying to do this that all the phenomena that we call absorption and secretion in the higher animals are produced. It is in the attempt to get into equilibrium with the intestinal contents on the one side, and the blood on the other, that the 252 (EDEMA AND NEPHRITIS mucosal cell (better, the colloid membrane separating the intes- tinal contents from the blood), absorbs the intestinal contents and transfers them to the blood. How this is accomplished will now be discussed. II ON ABSORPTION 1. General Remarks on the Physico-chemical Structure of an Absorbing System in the Complex Organism It follows as a necessary conclusion from our argument that in the resting state the body of a multicellular living organism — a mammal, for example — is built up of a system of different Qiydrophilic) emulsion colloids saturated with water. To be counted in with the structures that make up this water-saturated colloid system and composing an integral part thereof, are the blood and the lymph. It may at first sight seem somewhat surprising that the blood and lymph are included, but the relation between the col- loid and the water of fluid (hydrophilic) colloids, (sols), is identical with that of the relation between colloid and water in solid colloids (gels) such as fibrin. This identity is not only demanded by physico-chemical theory, but has been proved experimentally by the work of Wolfgang Patjli and Hans Handovsky ^ on blood serum .^ That the entire mixture of solid and liquid colloid ma- terial constituting the organism is saturated with water is evidenced by the fact that we cannot make it, as a whole, take up any more water or give up any except as chemical changes are first produced in it which either increase or decrease the capacity of its colloids for holding water. In consequence, an organism not subject to any marked changes from without or within njaintains a constant weight over long periods of time. We need but recall how all the secretions of a man undergoing absolute starvation drop to practically nothing, and how, on the other hand, the consumption of even enormous amounts of water by the normal individual does not lead to the development of ' Wolfgang Pauli and Hans Handovsky: Biochem. Zeitschr., 18, 340 (1909). ^ See page 106. ABSORPTION, SECRETION— COMPLEX ORGANISM 253 the slightest oedema. We are accustomed to say that the kidneys quickly rid the body of any excess of water. Just why this is done will be discussed shortly. The chemical changes capable of altering the hydration capacity of the body colloids may be of a character to affect the constitution of the entire mass that composes the body of a multicellular organism; or they may affect only smaller parts. In the former case we get either an absorption or a secretion of water by the organism as a whole; in the latter, only a limited or localized absorption or secretion of water by the parts involved. It is even possible for chemical changes to be going on in one part which lead to an absorption of water, while other changes are going on in another part which lead to a secretion of water. Thus, conditions are so arranged in the body under normal circumstances as to favor almost constantly the absorption of water from the gastro-intestinal tract, while at the same time they favor the secretion of the urine from the kidneys. An absorption system consists in essence of three parts or phases: (1) A material to be absorbed. (2) A membrane which does the absorbing, and (3) The blood or lymph into which the absorbed substance finally gets. In the case of the gastro-intestinal tract these general terms are synonymous with (1) The gastro-intestinal contents. (2) The gastro-intestinal mucosa with its specific cells and all their supporting structures, and (3) Tlie blood and lymph. Let us consider for a moment their physico-chemical properties. (1) The gastro-intestinal contents from a chemical point of view represent a widely varying mixture. Expressed physico- chemically, however, they are a mixture of colloids and crys- talloids in water. In the process of digestion the colloids are for the most part converted by a series of chemical cleavages into crystalloids. Thus, the proteins are broken into amino- acids, the fats into fatty acid and alcohol (glycerin), the car- bohydrates, when digestible, into the simple hexose sugars. In the end, the absorption of the gastro-intestinal contents becomes 254 CEDEMA AND NEPHRITIS really, therefore, the problem of the absorption of a watery solu- tion of various crystalloids. (2) The membrane through which the intestinal contents are absorbed into the blood and lymph is made up of all the cellular and intercellular elements found between the gastro- intestinal contents on the one side and the circulating blood and lymph on the other. From a histological standpoint we know that this membrane is very different in different parts of the gastro-intestinal tube. With the different cells we must always in our considerations count in the intercellular substance that binds them together. From a physico-chemical stand- point this membrane is colloid in constitution. It is made up, in the main, of a mixture of various (hydrophilic) emulsion colloids. As a whole it is more or less solid in nature like a leaf of gelatin soaked in water. But in no sense are the different portions of the gastro-intestinal tube piade up of exactly the same colloid material, either in a chemical or a physical sense. We know this to be true because its different parts take up dyes, for ex- ample, with very different avidities when these are injected into the circulating blood. We have now to point out that while the colloid membranes with which we busy ourselves in the laboratory are made up of dead material, that separating the gastro-intestinal contents from the blood and lymph is alive. This does not imply, how- ever, that we must at once become vitalists. It only means that it introduces a series of more or less independent chemical and physico-chemical reactions into our general problem of absorp- tion which demand additional care and study to analyze. The physico-chemical state of this living membrane is dependent upon the chemical changes that occur in the cells constituting this membrane, and these chemical changes are in turn intimately connected with the changes that occur in the blood supplying these cells. It is readily apparent, therefore, that the introduc- tion of a single variable into the circulation may upset the entire chemistry of the cells of the absorbing membrane, and so their physico-chemical state. This is why what looks like such a small change in our entire absorption system may be followed by a most profound effect upon absorption. (3) The blood represents a mixture of various formed elements with a liquid menstruum. The formed elements are colloid ABSORPTION, SECRETION— COMPLEX ORGANISM 255 bodies (nucleated and non-nucleated cells) which react toward changes in their environment (acid, alkahes, salts, non-elec- trolytes, etc.), in the very definite ways already described. The liquid portion of the blood (the plasma) is a hquid colloid mix- ture of various proteins. It behaves like a solution of gelatin, obeying laws identical with those governing the behavior of the more solid proteins which we have already discussed. Mixed with it are normally a number of different salts and varying amounts of different non-electrolytes. Absorption from the peritoneal cavity presents a problem fairly identical with that from the lumen of the gut. We need in the above paragraphs but to make the absorbing membrane consist of the peritoneal cells with their supporting elements instead of the mucosal cells with theirs. The blood and lymph remain the same, and the material to be absorbed from the peritoneal cavity may have under experimental conditions any composition we choose to give it. As absorption from the peritoneal cavity represents a somewhat simpler problem than absorption from the gut we shall consider it first. 2. Absorption from the Peritoneal Cavity ^ In view of the excellent running accounts of absorption that may be found by consulting R. Heidenhain,^ Ernest H. Starling,^ E. Waymouth Reid,* H. J. IIamburger,^ E. Over- ton,® Otto Cohnhe[m,^ or Rudolph Hober,^ it is needless to attempt any detailed definition of the present status of our iSee Mabtin H. Fischer: KoUoidchem. Beihefte, 2, 304 (1911). ' R. Heidbnhain: Hermann's Handbuch der Physiologie, 5, Leipzig (1883); Pfluger's Archiv, 56, 579 (1894). ' E. H. Stabling : Schaf er's Text-Book of Physiology, 1, 285, London and Edinburgh (1898); Oppenheimer's Handbuch der Biochemie, 3, 206, Jena (1909). * E. Waymouth Reid: Schafer's Text-Book of Physiology, 1, 261, London and Edinburgh (1898); Phil. Trans. Royal Soc, 192, 231 (1900); Journal Physiol., 26, 436 (1901). 'H. J. Hamburger: Osmotischer Druck und lonenlehre, 2, 95, Wies- baden (1904). ^ E. Overton: Nagel's Handbuch der Physiologie, 2, 774, Braunschweig (1907). '0. Cohnhbim: Nagel's Handbuch der Physiologie, 2,607, Braunschweig (1907). ' R. Hober: Koranyi-Richter, Physikalische Chemie und Medizin, 1, 295, Leipzig (1907). 256 (EDEMA AND NEPHRITIS knowledge of absorption. This is shown in a masterly way by these authors. Depending upon whom we consult we find sug- gested, as the forces active in this matter, variations in hydro- static pressure, filtration, or the two combined; diffusion and osmotic pressure, with modifications of both as determined by different media, different membranes and different solutions; imbibition; and when these physical forces are found wanting, then the " peculiar " forces of living matter are called upon for help. How unsatisfactory are all these explanations is clearly evidenced by the divergence of scientific opinion and the mutual criticism that finds expression in the individual writings of these authors, and this in spite of the fact that the experi- mental grounds upon which they base their opinions agree very well with each other. My own experiments referred to below were extremely simple in character, made purposely so in order to eliminate the many and great errors that creep into these absorption experiments as soon as anesthetics, operations, animal boards and elaborate pieces of apparatus are employed. Had it not been for the use of these, one might have contented himself with mere interpreta- tion of the experimental facts already found by previous authors. How some of these procedures affect absorption will be pointed out at the proper place. I used healthy guinea pigs which were kept on a liberal diet of timothy hay, corn and oats, with water ad libitum. In order to permit comparison with each other, the animals in each set of experiments were taken from the same cage and treated exactly alike. No anesthetic being necessary, none was given. The various solutions and the water, after warming to 38° C, were injected into the peritoneal cavity by means of a hypo- dermic needle. The animals were held only during the few moments necessary for the injection, after which they were allowed to run about in their cages. At the end of a specified time they were killed by a blow on the head, immediately opened, and the unabsorbed liquid contained in the peritoneal cavity aspirated, by means of a pump, into small flasks. The amount of fluid recovered was then measured. Let it be noted that what is discussed primarily in these pages is the absorption of water from the peritoneal cavity. A priori no one would be inclined to look upon the absorption ABSORPTION, SECRETION— COMPLEX ORGANISM 257 of any solution as representing a single process, and yet, in practice, this is done and has been done constantly. On all sides we see discussed the absorption of a solution as such. The absorption of a solution represents the composite of the absorption of the solvent, and the absorption of every individual substance dissolved in that solvent. Absorption of solvent and absorption of dissolved substance may mutually affect each other (see below), but this does not make them identical, nor does it make the absorption of the solution a single process. Excellent experimenters have gone so far as to look upon the distribution of a dissolved sub- stance (such as a dye) in a tissue as evidence that the solvent in which that substance was originally dissolved was present there, or at least had passed that way. This is a most serious mistake. § 1 When any liquid is injected into the peritoneal cavity and we find that after a time it has been absorbed, we know from anatomical considerations that it must have passed into the lymph and the blood streams after having traversed the cells and inter- cellular substance which originally separated these two circulat- ing fluids from the liquid that was injected. If we try to for- mulate the problem in terms of physical chemistry, then our purpose is to discover how the absorption of a solution that has any composition we may choose to give it, is accomplished by two colloid, circulating liquids (which, for the sake of brevity, we will regard as sols) that are separated from this solution by a soUd colloid membrane (a gel) . It is of interest for our further discussion first to call to mind which of these two hquid colloids plays the more important role in this absorption. As the per- itoneal cavity is usually regarded as an immense lymph space, one might on a priori ^'ounds be inclined to look upon the lymphatic circulation as that chiefly concerned in absorption from this cavity. And yet that the lymph plays a subordinate part and the blood circulation the chief r61e is indicated by E. H. Starling and Tubby'sI finding that dyes appear in the urine after injection into the peritoneal cavity before the lymph coming from the thoracic duct shows any color; by the observ- 1 Starling and Tubby: Journal of Physiol, 14, 140 (1894). Starling: Schafer's Text-Book of Physiology, 1, 304, London and Edinburgh (1898). 258 (EDEMA AND NEPHRITIS ation of Orlow,' who noticed no increase in lymph flow after intraperitoneal injections of salt solution; and by that of H. J. Hamburger^ who found peritoneal absorption unim- paired after ligation of the thoracic duct. But, after the point is estabhshed that absorption from the peritoneal cavity is brought about chiefly through the agency of the blood, we have yet to say why this is the case. It is evident that the answer to this question bears both a quantitative and a qualitative element. In the higher animals the lymph circu- lation stands quantitatively far behind the circulation of the blood. Other things being equal, the blood would therefore absorb more than the lymph in proportion as the blood flow through a part exceeds quantitatively the lymph circulation through the same part. But chemical differences between the two play, to my mind, an equally important part. The total colloid content of the blood is higher than that of the lymph.^ But, beyond this, the blood suffers rapid temporary changes in chemical composition that the lymph does not. Chief among these are the quantitative variations in the content of oxygen and (especially) of carbonic acid as induced through respiration. Further changes in the composition of the blood are wrought through the diffusion of metabolic products into and out of it, as when the blood passes through the kidneys, active muscles, the liver, etc. While somewhat similar changes are induced in the lymph when this passes through various organs, the rapid varia- tions that we find in the blood are for obvious reasons lacking. But these more marked and rapid changes in the blood, com- bined with its more rapid circulation, mean at the same time more marked and rapid changes in the surroundings of the various tissue cells about which this circulating medium passes. The equilibrium with their surroundings which these cells endeavor to establish is therefore continually being disturbed because of these changes in their surroundings; and, so long as this happens, so long must the cells absorb. And hence the greater 'Orlow: Pfluger's Archiv, 59, 170 (1895). 2 H. J. Hamburger: Arch. f. (Anat. u.) Physiol., 281 (1895). ^Almost one-fourth of the blood is protein. Blood plasma contains to each 100 parts almost 9 parts of protein. Lymph contains 3.4 to 4.1 parts of protein. See C. Schmidt: Vierordt's Daten und Tabellen, 97, Jena (1888); J. MuNK and Rosenstein: Arch. f. Physiol., 376 (1890). ABSORPTION, SECRETION— COMPLEX ORGANISM 259 importance of the blood circulation over the lymph circulation in this problem of absorption in the higher animals. § 2 Let us now turn to the problem of the absorption of water from the peritoneal cavity. When water is injected at body temperature into the peritoneal cavity of guinea pigs it is rapidly absorbed, as the following table shows: TABLE LXVII Guinea pig. Weight in grams. Amount of water injected in cc. Amount of fluid in cc. recovered after one liour. a b c d 413 535 544 460 20.8" 20.8 20.8 20.8 5.4 6.4 4.8 4.9 There is nothing new about this observation. Where we encounter difficulty is in saying why the water is absorbed. Against the generally accepted belief that water is under such circumstances absorbed because the osmotic pressure in the cells lining the peritoneum is higher than that of the distilled water, serious objections can be raised. We know that the peritoneum does not retain this water, but gives it up to the blood (chiefly) . On the osmotic basis this secretion into the blood could there- fore occur only because the blood has a higher osmotic concen- tration than the cell contents. As a matter of fact we know that body cells, lymph and blood have, to all intents and purposes, the same osmotic concentration. The still more seri- ous objection that this osmotic conception of water absorption ignores entirely the important part played by the intercellular substances need not be discussed here. That an injection of water into the peritoneum makes the cells here take up water because of an increased hydrostatic pressure directly induced by the injection, or aided by the contractions of the abdominal muscles, etc., is also scarcely tenable. The injection, first of all, does not appreciably increase the intra-abdominal pressure; and secondly, absorption occurs when the abdomen is opened, or in a dead animal (see below). 260 (EDEMA AND NEPHRITIS We have no difficulty in interpreting the absorption of water from the peritoneal cavity as a colloid phenomenon. In order that the absorption of water may occur, the hydrophilic colloids of the peritoneum must only be unsaturated with water. But when we consider the fact that after a few cubic centimeters of water have been absobred from the peritoneal cavity, more may be absorbed if the injection is repeated and this almost without limit, then we have to conclude that under normal circumstances the tissues composing the peritoneum are constantly unsaturated with water. What we really have to discuss, therefore, are the conditions that combine to keep the colloids of these tissues unsaturated with water in the living animal. The first of these is the continuous production of acid (carbonic acid) in the tissues composing the peritoneum. In consequence of this the capacity of the tissues for holding water is increased, and they absorb it from any available source. If water is present in the peritoneal cavity they will take it up. But this would only lead to a swelling of the peritoneal tissues. As in this process an upper limit would soon be reached and absorption cease, this alone cannot lead to the continuous absorp- tion which is observed. A second variable must exist, and this is found in the circulation of the blood and the lymph. Through these the carbonic acid produced in the cells is constantly carried away. But to carry this away from the tissues is to reduce the capacity of the colloids of the peritoneum for holding water, which, in consequence, they now give up. As long as the circula- tion is maintained in a normal way absorption from the peritoneal cavity must therefore be continuous, for while the tissues of the peri- toneum are on the one side busy in absorbing water they are, on the other, busy in giving it up to the blood along with their carbonic acid. The blood also carries all water contained in it in com- bination with the colloids found in the blood. As the arterial blood, low in carbonic acid, (representing, as we have said, a liquid colloid solution saturated with water), enters the capillaries, there diffuses into it the carbonic acid that is being produced in the cells. Through this the capacity of the blood colloids to hold water is raised, they find themselves in an unsaturated condition, and so are able to absorb water from any available source. This could be water directly, though in the living body ABSORPTION, SECRETION— COMPLEX ORGANISM 261 it means that the blood robs any tissue of its water that is holding it with less avidity than that represented by the colloids of the blood. In the case under discussion the blood absorbs water from the tissues composing the peritoneum. The peritoneum, in its turn, takes water from the peritoneal cavity, if any is present there. The now venous blood, with its higher water-content, passes to the lungs, where its carbonic acid escapes. When this happens the blood colloids are unable to retain longer the water absorbed previously, and this becomes " free " in the blood. It is this " free " water that under normal conditions the kidney extracts from the blood, and, by a process the reverse of that which we have described for peritoneal absorption, secretes as urine.^ To this question we return below. §3 We can immediately apply an experimental test to this proc- ess of reasoning. If the blood coursing through the peritoneal tissues and the tissues themselves take up water only because their colloids are unsaturated with it, then clearly they should be unable to take it up when offered to them in their own form. In other words, they should be unable to absorb water from any " solution " in which all the water is held in combination with colloids in the same way as they hold it. As a matter of fact, they cannot, as proved by the following experiments with egg albumin, which represents a liquid in which all the water is bound to colloid material. TABLE LXVIII. Guinea pig. Weight in grams. Amount and character of solu- tion injected. Amount of fluid in cc. recovered after one hour. a b c d 533 537 555 563 20.8 cc. water 20 . 8 cc. white of egg (natural) 31.2 cc. water 31 . 2 cc. white of egg (natural) 5.4 18.4 7.6 27.7 Similar experiments with blood are described later.^ It is because the water is thus united to colloid material that blood and lymph remain so long in the peritoneal cavity; in fact they ' See section on Secretion, p. 281. 2 See page 609. 262 (EDEMA AND NEPHRITIS cannot be absorbed from here or from the intestinal tract until ferments or other conditions have first so affected the colloids that they yield up their water in a " free " form. §4 The effect on water absorption when the same amount of differently concentrated sodium chlorid solutions are injected intraperitoneally instead of plain water is indicated in Table LXIX. TABLE LXIX Amount of fluid Guinea Weight in Amount and character of solu- pig- grams. tion injected. after one hour. a 417 20 . 8 oc. water (control) 5.4 b 397 20.8 cc. m/12 NaCl 11.8 c 419 20.8 cc. m/8 NaCl 13.0 d 488 20.8 cc. m/6 NaCl 14.6 e 445 20.8 cc. m/4 NaCl 19.9 The interpretation of these findings is about as follows: When, in place of pure water, a sodium chlorid solution is introduced intraperitoneally we may assume that the water of this solution tries to diffuse into the cells just as though the salt were not there. But, at the same time that this is occurring, the salt is also diffusing into the peritoneum. Just why this happens is discussed below. But the presence of the salt in the colloids of the peritoneal tissues will make these tend to give up water. The salts will, therefore, tend to counteract the effect of the carbonic acid in the cells. The normal fall which tends to make the water move from the peritoneal side of the peritoneal absorbing membrane to the vascular side will now be counteracted by one occurring in the opposite direction. The normal streaming of water which tends to make for an absorption from the peritoneum will be met by a counterstream which tends to make for a secretion from this structure. The end result, so far as absorption of water is concerned, will represent the algebraic sum of the two. If this second stream is not a great one, there will be only a slight reduction in the rapidity with which the water is absorbed. This is what happens with the dilute salt solution. But, as the concentration of the salt ABSORPTION, SECRETION— COMPLEX ORGANISM 263 increases, this counter-current must become more and more manifest, so that, as in the last experiment (e) of Table LXIX, practically no absorption (of water) occurs within the time limits set for the experiment. §5 In equimolar concentrations different salts affect to very unequal degrees the absorption of water by colloids swelling in the presence of an acid. So also, and in the same general order, do they affect the absorption of water by the peritoneum. TABLE LXX Guinea pig- Weight in grams. Amount and character of solu- tion injected. Amount of fluid in cc. recovered after one hour. a b c d e f 643 594 551 409 496 492 20 . 8 cc. m/8 sodium chlorid 20.8 oc. m/8 sodium acetate 20 . 8 cc. m/8 sodium nitrate 20 . 8 cc. m/8 sodium sulphate 20 . 8 cc. m/8 sodium citrate 20 . 8 cc. m/8 disodium phosphate 7.0 10.0 12.6 20.0 23.4 26.6 TABLE LXXI Guinea pig. Weight in grams. Amount and character of solu- tion injected Amount of fluid in cc. recovered after one hour. a 343 20 . 8 cc. m/8 potassium iodid 3.4 b 335 20 . 8 cc. m/8 potassium bromid 8.8 c 322 20.8 cc. m/8 potassium chlorid 11.0 d 290 20.8 CO. m/8 potassium sulpho- cyanate 13.4 e 355 20 . 8 cc. m/8 potassium nitrate 13.4 f 354 20.8 oc. m/8 potassium acetate 16.8 g 363 20 . 8 cc. m/8 potassium tartrate (died 40 minutes after injection) 18.9 h 386 20.8 cc. m/8 potassium citrate (died 15 minutes after injection) 20.5 TABLE LXXII Guinea pig. Weight in grams. Amount and character of solu- tion injected. Amount of fluid in cc. recovered after one hour. a 6 c d e 452 396 484 476 502 20.8 cc. m/8 potassium chlorid 20.8 cc. m/8 ammonium chlorid 20 . 8 cc. m/8 magnesium chlorid 20.8 CO. m/8 calcium chlorid 20.8 cc. m/8 strontium chlorid 12.8 13.7 19.4 24.2 24.4 264 (EDEMA AND NEPHRITIS As Tables LXX, LXXI, and LXXII show clearly, every one of the salts employed markedly retards the absorption of water from the peritoneal cavity. This harmonizes entirely with the fact that the presence of every salt inhibits the absorp- tion of water by such an emulsion colloid as fibrin, gelatin or serum albumin which is swelling in the presence of an acid. But the parallelism between the two processes is even closer than this. We note in Table LXX, for example, where the effects of equimolar solutions of sodium salts are compared, that the sulphate, citrate and phosphate have an effect far above that of the chlorid, acetate or nitrate in preventing absorption. In Table LXXI, where the effects of a series of potassium salts are compared, the order of the acid radicals is again the familiar one observed in studies on pure protein colloids. Table LXXII brings out the same fact for a series of different basic radicals. That the results should be so nearly identical with the effects of various salts on pure colloids is really somewhat surprising when we remember that in such experiments as these one is compelled to work with a very considerable experimental error, arising from the fact that in each of these series several animals are used, that we cannot control the amount of water consumed by them just before being used, that we cannot escape the specific poisonous effects exerted by the different salts employed, etc. Nevertheless the experimental results are point for point almost identical with the findings on pure colloids. This indicates to my mind how predominant is the colloid element in this prob- lem of absorption. §6 As compared with the effect of electrolytes, various non- electrolytes affect the absorption of water by colloids in the presence of any acid only slightly. Table LXXIII gives the results obtained when solutions of various non-electrolytes in concentrations osmotically about equivalent to the solutions of the various salts used above, are injected intraperitoneally : It is clear from this table that ethyl and methyl alcohols do not delay the absorption of water from the peritoneal cavity. On the other hand, glycerin and the two sugars used produce a decided inhibition. The sugars even produce a secretion of ABSORPTION, SECRETION— COMPLEX ORGANISM 265 fluid. The effects again agree with the findings on pure colloids where only the last-named produce in the higher concentrations a dehydration. TABLE LXXIII Guinea Weight in Amount and character of solu- Amount of fluid pig. grams tion injected. in cc. recovered after one hour. a 425 20.8 cc. m/4 ethyl alcohol 5.8 b 434 20.8 CO. m/4 methyl alcohol 2.1 c 464 20.8 cc. water (control) 6.6 d 583 20.8 cc. m/4 urea 11.7 e 569 20 . 8 cc. m/4 glycerin 18.2 f 687 20.8 cc. m/4 glycerin 17.4 g 521 20 . 8 cc. m/4 cane sugar 25.7 h 725 20.8 cc. m/4 cane sugar 27.0 t 522 20.8 cc. m/4 dextrose 26.3 ; 710 20.8 cc. m/4 dextrose 29.0 §7 Both alkalies and acids when injected intraperitoneally delay the absorption of water, as indicated in the following table: TABLE LXXIV Guinea Weight in Amount and character of solu- Amount of fluid pig. grams. tion injected. after one hour. a 544 20.8 cc. water (control) 4.7 b 545 20.8 cc. n/100 NaOH 6.2 c 543 20.8 cc. n/50 NaOH 11.0 d 568 20.8 cc. n/25 NaOH 10.6 e 460 20 . 8 cc. water (control) 4.8 f 460 20.8 cc. n/100 HCl 7.4 g 447 20.8 cc. n/50 HCl 12.4 h 450 20.8 00. n/33 HCl 12.0 In explanation of these results the following is offered. In the concentrations employed both acids and alkalies produce an excessive swelling of the peritoneal tissues. This excessive swelling delays absorption, not alone by occluding the lumina of the capillaries supplying the peritoneum and so decreas- ing the absolute blood flow through these tissues, but by so increasing the avidity of the peritoneal tissues for water that the blood passing through them is not enabled to take the water away from them with its usual ease. 266 OEDEMA AND NEPHRITIS § 8 Table LXXV shows how water and various salt solutions are absorbed from the peritoneal cavities of dead animals. The guinea pigs were killed by a blow on the head, and injected subsequently in the same way as in the experiments with living animals. After injection, the animals were turned about a few times to allow the liquids to spread through the peritoneal cavity, and were then laid quietly on their bellies for one-half hour, after which they were turned on their backs for one-half hour. TABLE LXXV Amount of fluid Recovered from second pouring in of 20.8 CO. water after Guinea pig- Hours dead. Weight in grams. Amount and character of solution injected. recovered in cc. after 1 hour. 1 hour. a just dead 331 20 . S cc. water 7.0 b just dead 396 20.8 cc. water 9.0 c 1.00 333 20.8 cc. water 9.4 15.3 d 2.30 351 20.8 cc. water 9.3 15.0 e 7.30 395 20.8 cc. water 8.0 f 24.00 375 20.8 cc. water 12.5 a 48.00 353 20 . 8 cc. water 17.0 h 0.15 267 20.8 cc. m/8 NaCl 13.2 i 0.15 295 20.8 cc. m/8 NazSOi 10.0' i 0.15 299 20. 8 cc. m/8 sodium citrate 11.4 J 1 A part of the peritoneal fluid was accidentally lost. The table shows that water is readily absorbed from the peritoneal cavities of dead animals. How is the result to be explained? The answer is not essentially different from that given for the living animal. An acid production in the tissues is again responsible for increasing the capacity of the tissue col- loids for holding water. Only, while we attributed this to car- bonic acid in the living animal, we can attribute it in the dead animal not only to this acid, but in addition to lactic and the other acids that we know are produced postmortem. The longer an animal is dead, the higher we may assume becomes the con- centration of the acids in the various tissues. On this basis we might expect a progressively greater absorption of water with every increase in the length of time that an animal is dead. But this could hold only within certain limits, for in pure colloids we know that with a progressive increase in acid concentration ABSORPTION, SECRETION— COMPLEX ORGANISM 267 the absorption of water increases only up to a certain point, after which a decreased absorption is noted. The same is evident in Table LXXV, where animals long dead (/ and g), show a decidedly lower absorption of water than animals dead only a short time. As is sufficiently well indicated by the results obtained with animals h, i and j, various salts retard the absorption of water from the peritoneal cavity of dead animals as they do in living animals, and, we would add, for the same reason. 3. Absorption from the Gastro-intestinal Tract The foregoing paragraphs, which show that the same con- ditions that retard the absorption of water by such an emulsion colloid as fibrin or gelatin, retard in almost identical fashion the absorption of water from the peritoneal cavity, prove, it seems to me, that the two processes are in essence the same. What is next in order is to compare this process of peritoneal absorption with the processes of absorption as observed in other regions of the mammalian organism to see if the conclusions drawn regard- ing absorption as observed in the peritoneal cavity cannot be extended to cover at least some of these. Of chief interest in this connection, because of its physiological importance, is absorption from the intestinal tract. To anyone conversant with the wealth of experimental data on alimentary absorption that has been accumulated by VoiT and Bauee,i R. Heidenhain,^ Franz Hofmeistee,^ H. J. Hamburger,* R. Hober,^ G. B. Wallace and A. R. CusHNY,^ Otto Cohnheim,^ E. Waymouth Reid ® and G. KovESi,^ the following are familiar facts : ' VoiT and Bauer: Zeitschr. f. Biologie, 5, 536 (1869). 2R. Heidbnhain: Pfluger's Arch., 56, 579 (1894); 62, 331 (1896). 3 Franz Hofmeister: Arch. f. exp. Path. u. Pharm., 28, 210 (1891). ^ H. J. Hamburger: Osmotischer Druck und lonenlehre, 2, 168, Wies- baden (1904), where references to his earlier papers will be found. ^Rudolph Hober; PflUger's Arch, from 70 on; see his many papers during the years 1898 to date. ^ G. B. Wallace and A. R. Cushny: Am. Journal of Physiol., 1, 411 (1898); Pfluger's Arch., 77, 202 (1899). 'Otto Cohnheim: Zeitschr. f. Biol., 36, 129 (1897); 37, 443 (1899). 8E. Waymouth Reid: Journal of Physiol., 21, 85 (1897); 22, 56 (1898); 26, 427 (1901). 9G. Kovesi: Centralbl. f. Physiol., 11, 553 (1897). 268 (EDEMA AND NEPHRITIS When water is introduced into a segment of intestine it is rapidly absorbed. All salt solutions, so far as the water in them is concerned, are absorbed less rapidly than the pure water. The concentration of the salt solution is an important factor in this phenomenon. When sodium chlorid solutions of different concentrations are compared, they are found to be absorbed the more slowly the higher the concentration of the salt. If sufficiently strong solutions are employed there may first result a pouring out of liquid into the lumen of the gut, so that the solution becomes diluted, after which it is slowly absorbed. When the absorption of equimolar (or better, osmotically equivalent) solutions of different salts is studied, it is found that these are absorbed at very different rates. The effect of any salt in a solution upon the absorption of water from that solution may be thus stated: With a given base, the acid radicals arrange themselves in the following order, where that which retards least is given first: Chlorid, bromid, iodid, nitrate, sulphate, phosphate. With a given acid, the order of the basic radicals is as follows (R. Hober), that least effective in preventing absorp- tion being given first: Potassium, sodium, calcium, magnesium, barium. It is easy to see that the order of the various salts is practically identical with that found above in the experiments on peritoneal absorption. The position of the acetate, tar- trate and citrate, not given in the above hsts, can be deter- mined by consulting the tables of Wallace and Cushny, when it is found that they occupy a place in the absorption of water from the gut which is the same as that occupied by them in the case of peritoneal absorption. With any of these salts as with ordinary sodium chlorid, the delay in the absorption of the water grows with the con- centration of the salt. A point is finally reached where such water as is introduced into the intestine is not only not absorbed, but water is secreted into the gastro-intestinal tract. This concentration point hes high in the case of sodium chlorid, sodium bromid, etc., but very low in the case of sodium sul- ABSORPTION, SECRETION— COMPLEX ORGANISM 269 phate, phosphate, tartrate, citrate, etc. This is one of the chief reasons why the last named are known as " sahne cathartics." Point for point, therefore, the absorption and secretion of water by the gut is identical with the absorption and secretion of water by the peritoneum, and both are comparable to the absorption and secretion of water by simple protein colloids when placed in like surroundings. The identity of the processes of absorption from the per- itoneal cavity and from the intestinal lumen goes even further. The rapid absorption of aqueous solutions of various alcohols from the intestinal tract shows that these non-electrolytes do not interfere with the absorption of water even when present in concentrations osmotically equivalent to those of the active salts. Sugar solutions and glycerin also behave in the intestinal tract, so far as the absorption of water from their solutions is concerned, as they do when introduced intraperitoneally. The slow absorption of water, or, in response to the introduction of a solution of sufficiently high concentration, the actual secretion of water into the gut, is evidenced not only by direct experiment, but by everyday clinical experience. Are not the sugars, when consumed in any considerable quantities, capable of producing watery stools (independently of any previous fermentation with the production of organic acids), and do not glycerin enemas produce the same secretion of water into the bowel that results when enemas containing any of the saline cathartics are employed? We have interesting parallels also of the peritoneal experiments which showed that water when united to a (hydxophilic) emul- sion colloid is incapable of being absorbed without first being freed. Thus, protein solutions (such as egg white, blood, or blood serum) are not absorbed from the intestinal tract unless proteolytic ferments are present which, by acting on the proteins chemically, destroy their markedly (hydrophilic) colloid character, and so liberate the water held by them. In this way also can we understand the behavior of cellulose and, especially, agar-agar in preventing constipation. One of the commonest causes of con- stipation resides in the too perfect absorption of water from the gastro-intestinal contents. It is a time-honored custom to suggest the addition of vegetables to the diet of such individuals. In addition to the action of the salts (citrates, tartrates, etc.) obtained from vegetables and the effects of the production 270 (EDEMA AND NEPHRITIS (through fermentation) of certain organic acids in the bowel which alone tend to prevent a too great absorption of water from the gastro-intestinal contents, the high cellulose content of such a diet (that is to say, its high emulsion colloid content) makes it impossible for the mucosa to get the water out of it. Since cellulose is not changed (except very slightly by certain bacteria) in its passage through the gastro-intestinal tract, it retains all the water with which it was saturated before being consumed, or with which it saturates itself in its course through the alimentary tract. The same explanation holds for agar- agar or the feeding of any of the Japanese sea weeds from which this is prepared. Agar-agar is a (hydrophilic) emulsion colloid incapable of being affected chemically in its passage through the gastro-intestinal tract (L. B. Mendel and Saiki), and so any water that it may have absorbed before being swallowed, or may absorb in the gastro-intestinal tract, is retained. In this way the inspissation of the. gastro-intestinal contents (and so the constipation) is prevented. These paragraphs suffice to show that the colloid-chemical theory is adequate to explain the qualitative aspects of water absorption in the complex organism. It remains to show that it is also adequate from a quantitative point of view. This is easily done. The anatomical and physiological conditions exist- ing normally in the body tend to keep the colloids of the gastro- intestinal tract and of the blood and lymph streams passing through it in an unsaturated condition so far as water is concerned, while the reverse conditions hold for any secreting organ such as the kidney. The mouth and esophagus play practically no role in the absorption of water. The stomach, according to von Mering's experiments, also takes but little if any part in the absorption of water. The small and large intestine are the absorptive organs for this substance par excellence. The stomach is richly supphed with arterial blood. The small and large intestine are also generously supplied, but not as generously as the stomach. The separate branches of the mesenteric arteries which go to supply the villi occupy a fairly central position in this structure and break up into a capillary network which lies close under the intestinal epithelium. As clearly evidenced by the dark color of the portal blood, and direct gas analysis, the blood returning from the intestine is intensely venous (poor in oxygen and rich in car- ABSORPTION, SECRETION— COMPLEX ORGANISM 271 bonic acid). The experiments of von Limbeck, Guhber, and Hamburger i show that under the influence of such an increase in carbonic acid concentration as exists normally in venous blood over arterial blood the red and white corpuscles absorb an amount of water which easily amounts to from 5 to over 15 per cent^ of their volume in arterial blood. If we use only the lower of these values and ignore entirely the water-carrying power of the colloids contained in the plasma, a little calculation shows that every hter of blood passing through the intestinal tract is capable of absorbing 17.5 cc. of water, for the corpuscles when moist make up in round numebrs about 35 per cent of the blood. Even these values, which have been chosen as low as possible, easily suffice to account for the absorption of great amounts of water from the gastro-intestinal tract. 4. Historical and Critical Remarks on the Theory of Absorption. Peritoneal and Alimentary Absorption of Dissolved Sub- stances. Let us now consider for a moment the explanations of absorp- tion that have been given by other authors, and select from them not only the elements which we ourselves think to be correct, but point out, with the help of a few examples, how certain experiments which have long stood as the bulwark of " physiolog- ical " or " vitalistic " interpretations of certain life phenomena are easily explained on the colloid basis, and how others long held to support different theories of absorption fall in with the colloid one. ^ H. J. Hamburger: Osmotischer Druck u. lonenlehre, 1, 291, Wiesbaden, (1902); ibid., 1, 404 (1902). ^ These figures are nearly doubled if instead of comparing the sizes of the corpuscles in arterial and in ordinary venous blood the sizes in arterial and passively congested venous blood are compared. In other words, the same circumstances that make the passively congested organ become oedematous make the corpuscles in the blood become " oedematous," and since the (colloid) plasma also " swells," venous blood or " passively congested " blood is, if water is available, richer in water than normal arterial blood. The blood is " hydremic," but this hydremia is not the cause of an oedema; it is an oedema itself, an expression of the same factors which make the more solid tissue " oedematous." 272 (EDEMA AND NEPHRITIS For half a century various authors have thought that filtra- tion plays an important part in the absorption of liquids. Accord- ing to definition, filtration represents the passage of a liquid through a separating membrane of some sort in consequence of differences in hydrostatic pressure. On this basis it has been held that a Uquid is forced from the lumen of the gut or from the peritoneal cavity into the blood because of a pressure within the gut or peritoneal cavity (produced through gas or the action of muscles of various kinds) which exceeds the pres- sure in the blood. Such a belief has been supported by the experiments of Leubuscher and H. J. Hamburger/ who found that with an increase in intra-intestinal or intra-abdominal pressure there resulted an increase in absorption, at least up to a certain point. Without for the moment questioning the correct- ness of the experimental finding itself (a serious error enters into it) we know that such an intra-intestinal or intra-abdominal pressure is not necessary for absorption. E. Waymouth Reid ^ found absorption (of water) to occur from the intestine of the dog, when the pressure within the gut was decidedly lower than that in the mesenteric veins, and Hamburger himself describes experiments in which he observed a ready absorption of water from the peritoneal cavity when the abdomen was opened or when the animal was dead. After what has been said above regarding water absorption as a colloid phenomenon these findings are entirely to be expected. What is needed is an interpretation of Leubuscher and Ham- burger's experiments with changes in pressure. Leubuscheb's result has been explained by saying that through increased intra- intestinal pressure the folds of the intestinal mucosa are smoothed out, and so an increased surface of gut is rendered available for absorption. But this explanation has not been accepted as complete by Hamburger, who found an increased absorption from the gut with every increase in pressure up to a certain point, even after surrounding the intestine with a wire cage which prevented its unfolding. In explanation of Hamburger's result, I would agree with the view that with the first increase ^H. J. Hamburgee: Osmotischer Druck und lonenlehre, 2, 176, Wies- baden (1904). 2E. Waymouth Reid: Philosoph. Trans. Royal Soc, 192, 231 (1900). ABSORPTION, SECRETION— COMPLEX ORGANISM 273 in pressure the flow of blood out of the veins is favored. In consequence of this the blood flow through the gut is favored and so the conditions for absorption. With a further increase in pressure, the blood vessels are compressed, and now the blood flow is diminished, in consequence of which a decrease in absorp- tion is observed, as Hamburger found. I am not even inclined to accept the view of those authors who, while unwilling to look upon filtration as an important factor or as the most important factor in the passage of liquid from one region to another, nevertheless consider that it is of some physiological importance. To my mind, this cannot have a magnitude any greater than, say, the theoretical " solubility " of quartz in water, for the membranes in health through which water is supposed to be forced are built up of emulsion colloids, and the differences in pressure available in the body for filtra- tion through the membranes existing here, have a value approx- imating zero, when compared with the enormous pressures required in the laboratory to force water through the thinnest layers of such emulsion colloids as gelatin. § 2 The question of osmotic pressure (even as modified through the conception that the surface layer of the cells is lipoid in character) in the problem of water absorption from the peritoneal cavity or the gut needs no special discussion — its inadequacy to explain the phenomena of absorption as observed here is admitted on all sides. That it is still adhered to, no doubt depends on the fact that we have had nothing more adequate to substitute for it; and any number of biological workers have been unwilling to believe that a present inability to explain on a purely physico- chemical basis all the phenomena of absorption (or secretion) presages that such will never be forthcoming, and that, in con- sequence, support is lent " physiological " or " vitalistic " conceptions of the process. It seems to me that on the basis of what has been said in these pages and in my previous papers, we are now in a position where we not only may, but must, dis- card the osmotic conception of cell behavior so far as water absorption is concerned. We must also discard it so far as the absorption of dissolved substances is concerned. 274 CEDEMA AND NEPHKITIS As already noted, if cells were surrounded by semi-permeable films, as demanded by the osmotic theory, dissolved substances could neither get into nor out of them. Yet both must be pos- sible, as well as the movement of water into and out of cells, otherwise their life would cease. Dissolved susbtances get into and out of cells by diffusion. The role of this factor has been recognized and discussed as active in the process of absorp- tion since the days of Carl Ludwig. R. Heidenhain succeeded in minimizing its importance in the analysis of the whole prob- lem by showing that an absorbed fluid (or a secretion) usually differs in quantitative composition from the source from which it was derived. On this is based his belief in the selective " physiological " character of absorption (and secretion). There is nothing surprising about these phenomena to us. We expect them, in fact. As has been said, a solution is never absorbed (or secreted) as such. Whenever a solution is seen to be absorbed (or secreted), we are observing the composite of the absorption (or the secretion) of the solvent plus the absorption (or secretion) of each individual substance dissolved in that solvent. When any solution is introduced into the intestine, for example, each one of the dissolved substances diffuses into the wall of the intes- tine until an equilibrium is established in the distribution if each of these substances between the (liquid) phase represented by the solution and the more solid phase represented by the (colloid) intestinal wall. Similarly, every substance present in the intestinal wall tends to diffuse out into the solution to the estab- lishment of an equilibrium. In biological material it has been very generally assumed that the distribution of dissolved substances between two such phases should attain equilibrium when the concentration of any dissolved substance is the same in both. Such an a priori con- clusion is entirely unjustified. We deal in this problem with the distribution of a dissolved substance between water and a colloid, and, as we know from the facts now available on this subject, equilibrium may be reached when the dissolved substance is contained in less, the same, or a higher concentration in the col- loid than in the solution surrounding it.^ Now, while the absorptive membrane is trying to get into 1 Feanz Hofmeister: Arch. f. exp. Path. u. Pharm., 27, 395 (1890) • 28, 210 (1891). ABSORPTION, SEOKETION—COMPLEX ORGANISM 275 equilibrium with the solution to be absorbed on the one side, it is also trying to get into equilibrium with the blood on the other. The whole absorptive system consists of the three phases (the material to be absorbed, the colloid absorbing membrane, the Uquid colloid blood and lymph) already discussed, and the ■problem of the " selective " absorption of the fiissolved substance is the problem of the agencies concerned in establishing an equilibrium between all the various dissolved substances in these three phases. The factors of greatest importance in such a problem are the character of the various colloids concerned, and their physico-chemical state as determined through the presence of acids, alkalies, salts and various non-electrolytes; the nature of the dissolved substance to be absorbed, as its rate of diffusion; the presence or absence of lipoids in the colloid, absorbing membrane and in the blood, etc. In other words, the laws of adsorption, of partition, and of ohemical combination are all at work. To the process of simple diffusion in this matter of absorption (or secretion) become added therefore a series of secondary phenomena that obscure its purity. To illustrate what has been said, let us try to follow the relatively simple process of the absorption of a strong (so-called hypertonic) sodium chlorid solution when this is introduced into the peritoneal cavity, or into the intestine. Both the water and the salt begin immediately to diffuse into the absorbing membrane. As this progresses, the concentration of the sodium chlorid in the absorbing membrane rises. This rise in con- centration so affects the colloids of the absorbing membrane that they stop taking up water, or, if sufficiently strong, an actual secretion of water into the peritoneal cavity or the gut may follow. While this is occurring, an equilibrium is tending to be established between the concentration of the sodium chlorid in the solution undergoing absorption and the sodium chlorid in the absorbing membrane. But this is never attained under normal circumstances, because the salt in the absorbing membrane is at the same time trying to get into equilibrium with the sodium chlorid in the blood. Now, since this is circu- lating, it is evident that the equilibrium is constantly being broken down toward the side of the blood. In consequence of this, more and more salt must move over into the blood (be absorbed) . But, as this occurs, the state of the colloids of the 276 CEDEMA AND NEPHRITIS absorbing membrane again returns to a more " normal " one and so the absorption of water, wbich could not occur before, can again take place. With a dilute (a hypotonic) solution of sodium chlorid the water does not meet with so great a resistance to absorption, and it is, therefore, possible for the dilute salt solution to become more and more concentrated as the water is (the more rapidly of the two) absorbed from it. Even salt solutions isotonic or isosmotic with the blood must be absorbed. Though such a solution cannot be absorbed on the .osmotic basis because no differences in osmotic pressure exist to make the water move, there is no difficulty in interpreting what happens on the colloid basis. Let the colloids of the absorb- ing membrane take a little water from the isotonic solution and salt must quickly follow, for now its concentration is no longer in equilibrium, with that of the sodium chlorid in the absorb- ing colloid membrane. Then more water goes in, and then more salt, until all is absorbed. Or we could start the absorption by having a little salt go in first and then the water, etc., for if the truth be told we do not yet know just what concentration characterizes the " isotonic '' solution, nor shall we until the colloid constitution of living matter has been adequately taken into account. Finally, on the basis of these conceptions of absorption, we experience no difficulty in understanding why any solution remaining for longer periods in the peritoneal cavity or in the intestine (while being " absorbed ") has substances found in the blood or tissues appear in it which it did not originally contain. As dissolved substances diffuse out of a solution undergoing absorption into the absorbing membrane until an equilibrium is established, just so, of course, must the substances contained in the absorbing membrane tend to diffuse into the solution. It has been generally held that this diffusion of salts (and other substances) out of an absorbing membrane into a solution that is being absorbed constitutes " an attempt to establish osmotic equilibrium between the two." As a matter of fact such a conclusion is premature if nothing more. We do not yet know all the factors involved in determining the point of equilibrium in the body, in the distribution of the various dissolved substances between the phases concerned. One thing. ABSORPTION, SECRETION— COMPLEX ORGANISM 277 however, is certain and that is that the final equihbrium is not a simple osmotic equilibrium. • This is clearly enough evidenced not alone by the well-known fact that the physiological behavior of different salts, in this process of absorption, for example, bears no relation to their osmotic concentration, but by the further fact that the distribution of most dissolved substances between a colloid and a solvent is practically never the same in both. The " selective " character of absorption depends upon the fact that the absorption of water in the living organism is a process entirely separate from the absorption of dissolved substances. Of the latter each moves at its own rate, and is influenced in its movement by factors that may not affect others in the same way or to the same degree. A dissolved substance may actually be absorbed while water is being secreted, and vice versa. Thus, to produce catharsis a saline cathartic diffuses into the wall of the gut (is absorbed) while water is being given off (is being secreted); and salts diffuse into the distilled water introduced into the bowel while this is being taken up by the mucous membrane. Such facts would be impossible of explanation if the cells had " osmotic," " lipoid," or similarly constituted " membranes " about them. If what is here written will be kept in mind the " selective " character of absorption ceases to produce astonishment — it would be more strange were it not selective. § 3 The workers in physiology and experimental medicine, who have called attention to the "secretory " and " physiological " activities of absorbing (and secreting) membranes and to the " physiological driving force " ^ situated in them, are deserving of blame or praise depending upon whether they used these words in the despairing attitude of those biological workers who believe that life phenomena will never be interpretable solely in the terms of the physical sciences, or as convenient heads under which to group certain of the phenomena of absorp- tion and secretion which could not be so analyzed at the time they prosecuted their scientific studies. But the necessity of retaining these terms, even in the latter sense, may now largely. 1 " Physiologische Triebkraft " of the Germans. 278 (EDEMA AND NEPHRITIS disappear. The " secretory " activities of absorbing (and secreting) membranes as evidenced through their " selective " absorption and secretion of water and dissolved substances we have already discussed in the preceding paragraphs. Their "physiological " activity retains a meaning only in the sense that the absorbing (and secreting) membranes of multicellular organisms contain living cells in each of which there are occur- ring well-ordered series of chemical and physico-chemical reac- tions which are capable of influencing the colloid constitution of these membranes. To do this is to influence the nature of the phases and the conditions of equilibrium in our ab- sorptive (secretory) systems, and therefore secretion and absorp- tion. But these reactions are not impossible to analyze. The need for the " physiological driving force " also disappears. Such was originally called upon to help explain such a fact as the absorption of a solution, from the intestine for example, when the pressure under which it stands in the lumen of the gut is less than that of the pressure of the blood in the mesenteric veins through which it is being absorbed (E. Waymouth Reid). Such a view regards absorption as a process in which water is forced into the tissues. This is not what happens. It is sucked in, and such a process can occur even when the hydrostatic pressure in the veins happens to be a few millimeters above that in the lumen of the intestine. The pressures produced in the swelling of emulsion colloids are enormous as compared with the highest hydrostatic (arterial blood) pressures ever observed in animals possessed of a circulation. The use of "physiological" poisons to illuminate the " physiological " element in absorption (or secretion) proves nothing. Such poisons simply constitute a direct or indirect means of altering the physico-chemical state of the absorbing (or secreting) structures. And is it not the problem of physiology to state in terms of physics and chemistry just what this " normal " absorption (and secretion) is? When we use a " physiological " poison we have to explain the action of the poison along with what constitutes " normal " absorption (and secretion). Nothing has perhaps so effectively hampered the acceptance of the belief that absorption (and secretion) would ultimately ■prove themselves completely analyzable physico-chemically, and fostered the continuation of the " physiological," " secretory," ABSORPTION, SECRETION— COMPLEX ORGANISM 279 etc., notions of absorption (and secretion) as a series -of experi- ments first described by R. Heidenhain, and more recently repeated in modified form by E. Waymouth E.eid and Otto CoHNHEiM. The most striking of tiiese is the often quoted experimental finding that a dog will absorb his own blood serum. A word regarding experiments of this character may serve to show how they are interpretable on the basis of the colloid theorj^ of water secretion and absorption. In not one of these experiments, except where the possibility of the presence of pro- teolytic ferments is not excluded, is the serum or plasma completely absorbed. The reason why some is absorbed, but never all, is clear from the following: Blood serum and blood plasma are not blood; they are blood minus much of its emulsion colloid content. They are not solutions in which all the water they contain is bound to colloid material as in normal blood, but they contain " free " water over and above that necessary to saturate the colloids remaining in the " serum " or the " plasma." When they are introduced into the intestine they are therefore absorbable but only in so far as they contain free water (and a certain proportion of salts, urea, etc.) The absorption comes to a halt as soon as water has been absorbed down to the point where the remaining water is combined with colloid. In these experiments things are there- fore not the same on both sides of the absorbing membrane. The animal does not absorb serum or plasma as such, much less what these authors seem at times to try to have us believe, namely, something identical with blood itself. The animal absorbs some water and a few dissolved substances, and it does this for the same simple reason that it absorbs under similar circumstances an ordinary " physiological " salt solution. § 4 Of the different factors that are catalogued in our treatises on physiology, and which have at various times and in various ways been looked upon as active in this problem of absorption, only one remains to be discussed — that of imbibition. What Adolph Tick ^ has called molecular imbibition is but another 'Adolph Fick: Medizinische Physik, Dritte Auflage, Braunschweig (1885). 280 (EDEMA AND NEPHRITIS term for what we to-day call the absorption of water by an emul- sion colloid. It is, therefore, of interest that mention is made of imbibition as being important in the general problem of absorp- tion as far back as I88I.1 But the real significance of imbibition as a factor concerned in absorption was only pointed out more recently by H. J. Hamburger.^ This author correctly empha- sized the theoretical importance of his observation that animals absorb various solutions from their peritoneal (and other serous) cavities after death. While we cannot agree with the details of his ideas outlining the way in which the forces of imbibition act normally, a discussion into which it is not necessary to go here, there is no disputing his claim that imbibition plays a r61e. But Hamburger does not regard imbibition as the most sig- nificant factor in absorption, and continues to hold to the ideas of filtration, osmotic pressure and the " mitschleppende Wirkung " of the circulation as also concerned in the process. Nor does he suggest any way by which a fluid absorbed by imbibition is again gotten rid of. In view of all this, it seems to me that we owe a special debt to Franz Hofmeister,^ who, as early as 1891, pointed out that the salts which make (partially) water- soaked gelatin discs give off their water (secrete it) are identical with the so-called saline cathartics, and suggested that the two processes are in essence the same. In spite of the numerous papers on alimentary absorption and secretion, and on the mode of action of the saline cathartics that have appeared since Hof- meistbr's writings, there seems little question that we are destined to return to Hofmeister's conclusions,' and find in them not only an explanation of the mode of action of these cathartic salts, but a model of that which constitutes the essence of absorption and secretion. iW. VON Wittich: Hermann's Handbueh d. Physiol., 6, 2ter Theil, Leipzig (1881). 2 H. J. Hamburger: Osmotischer Druck und lonenlehre, 2, 108 and 164, Wiesbaden (1904). 3 Franz Hopmeister; Arch, f. exp. Path. u. Pharm., 28, 210 (1891). ABSORPTION, SECRETION— COMPLEX ORGANISM 281 III ON SECRETION 1. Introduction The paragraphs on absorption have shown how water and various dissolved substances get from the lumen of the gut (or any other point of absorption, as the skin, peritoneum or other serous cavity) into the absorbing , colloid mucous membrane and through this into the blood and lymph. The blood and lymph are ultimately united when they empty into the large veins near the heart. What is the fate of the absorbed water and the dissolved substances? Clearly, two possibilities present themselves: they may be retained in the various cells, tissues and organs of the body, or they may again be given off (perhaps after having suffered antecedent chemical changes) through the various secretions from the body. Under the first of these headings the body cells come to absorb from the blood the requisite amount of water to maintain their normal water content (normal turgor) ; or, if for any reason the hydration capacity of the (hydro- philic) emulsion colloids in any cell, tissue or organ, or in the organism as a whole has been abnormally heightened, then these absorb more than the usual amount of the proffered water, and swelling to more than normal size give evidence of an oedema in the involved parts. The dissolved substances brought with the water are accepted or rejected by the body cells depending upon whether their content of these is or is not in equilibrium with the concentration of these same substances in the blood. Oxy- gen, food-stuffs, substances like the salts when present beyond their " physiological " concentrations, various poisons or medici- nal agents which by accident or design have gotten into the blood are thus taken up, while carbonic acid and the products of cell metabolism are for the same reason given off. What water remains above the amount necessary to satisfy the hydration capacity of the body colloids as well as the dis- solved substances representing, for example, the products of cell activity, are lost thi'ough the secretions. Why this happens and how is our next problem. As the kidney represents both from a qualitative and a quantitative point of view the great 282 (EDEMA AND NEPHRITIS secretory organ of our bodies, we will limit our discussion largely to it. Our remarks upon it may with little modification be applied to any of the other secreting organs, as the skin, the salivary glands, the stomach, etc. The fol-owing paragraphs do not presume to give a complete analysis of the physiology and path- ology of kidney function; they try, however, to show how the general problem can be broken up into a series of smaller ones. An attempt is made to explain some of these, while others are correlated with problems which still await an answer in physical chemistry. 2. General Remarks on the Structure of a Secreting System in the Complex Organism The kidney function shows in common with all secretory systems (1) a secretion obtained through (2) a secreting mem- brane from (3) a source of some kind. In the case of the kidney these terms are synonymous with urine, kidney parenchyma and blood. In their physico-chemical properties they parallel the three phases discussed above as entering into the construction of any absorbing system. (1) The urine is essentially a watery solution of various electro- lytes and non-electrolytes. At times it may be acid, at other times neutral or even alkaline in reaction. Under normal cir- cumstances, as it escapes from the uriniferous tubules, it contains so little colloid material that for our purposes it is negligible. Albumin, mucin, etc., are, of course, present even in normal urine, though in such small amounts that they escape notice when only our ordinary analytical methods are employed. (2) The secreting membrane through which the urine comes is made up of all the cellular and intercellular elements found between the urine on the one hand and the circulating blood on the other. It is, in other words, the kidney itself. From a histological standpoint this membrane differs somewhat in its different parts. To start with, the membrane consists of a layer of endothelial cells (of the blood capillaries) covered by a layer of the cells of Bowman's capsule, the whole joined together by a certain amount of intercellular substance. While the endothelial cells continue throughout the length of the membrane the addi- tional covering changes, first to the cells of the convoluted tubules. ABSORPTION, SECRETION— COMPLEX ORGANISM 283 then to those of the different parts of the loop of Henle, then to those of the second set of convoluted tubules and finally to the cells of the collecting tubules. The membrane, from a physico-chemical point of view, is colloid in constitution and simi- lar in its general properties to the more soUd protein colloids such as fibrin or gelatin. But in no sense are the different por- tions of the tube made up of exactly the same colloid material, either in a chemical or a physical sense, as clearly indicated by the fact'that dyes, for example, do not stain all portions equally. The membrane, moreover, is alive. As we said in discussing absorption, this means that a series of more or less independent cheniical and physico-chemical reactions are thereby introduced into our general problem of secretion which demand additional care in the analysis of the whole problem. (3) The physico-chemical properties of the blood have been previously described. It represents a liquid colloid menstruum in which float the more solid colloid corpuscles, the whole showing the general physico-chemical reactions characteristic of simple proteins. 3. A Model Illustrating Some Phases of Urinary Secretion Before continuing our main argument it is well to digress here and describe a somewhat crude but quite efficient model of urinary secretion (Fig. 8). Familiarity with it may help to a better understanding of what follows. This model consists of a layer of finely powdered (preferably faintly acid) blood fibrin (6) in the bottom of an ordinary calcium chlorid tube (C), the outlet of which has been plugged with a little cotton (a) to keep the fibrin from falling through. The overflow tube (c) keeps the liquid in (7 at a constant level. The whole is fastened in an upright position into a support. Above it are clamped two large separ- ator funnels {A and B) furnished with stopcocks which permit regulation of outflow. If now a " physiological " salt solution contained in one of the funnels (m/8 or 0.72% NaCl) is allowed to flow into the cal- cium chlorid tube in such a way that a constant level is main- tained, it is seen to pass through the fibrin (which swells some- what) and to escape in drops at the lower end of the tube. The rate at which the salt solution escapes (cc. in units of time) 284 (EDEMA AND NEPHRITIS remains constant for indefinite periods of time if the pressure remains the same. If the level of the solution in the calcium chlorid tube is raised, then the " secretion " occurs more rapidly. When a dilute acid or a sodium chlorid solution containing an acid is substituted for the pure sodium chlorid solution, the rate of out- flow is seen to diminish gradually, and finally, perhaps, to stop entirely. At the same time the fibrin swells and the solution that drips through gives an albu- min ring with nitric acid. If the pure sodium chlorid solution is re- turned to, or enough of this salt, or better, sodi- um citrate, tartrate or sulphate (or any other of the " saline diuretics ") is added to the acid solution, the secretion can be made to recom- mence, first slowly, then more rapidly, and ulti- mately the normal, or even a better flow may be obtained. At the same time the albumin ring disappears from the liquid that passes through. If various non-elec- trolytes (ethyl or methyl Figure ABSORPTION, SECRETION— COMPLEX ORGANISM 285 alcohol, urea, glycerin) are used in place of the salt solutions, either alone or in combination with an acid, the changes in rate of outflow are scarcely noted, if at all. The interpretation of these simple facts offers no particular difficulties. The liquids introduced into the calcium chlorid tube escape below after traversing a capillary bed formed by a (hydrophilic) emulsion colloid. Acids of various kinds which make the fibrin swell, decrease the rate of outflow by decreasing the size of the capillaries. The effect of neutral salts and acids on the sweUing of fibrin explains why such salts as the citrate, sulphate and tartrate of sodium can make a layer of fibrin permeable to water once more, after it has been rendered impermeable by a pure acid. Albumin appears in the filtrate when it does because the fibrin is (pseudo-) soluble in acid solu- tions. It becomes less in amount or disappears entirely when enough of different salts is added, because these reduce the " solubility " in acids. The question now arises whether this model of secretion has anything in common with the physiological and the pathological secretion of urine. I believe it has, though not in as coarse a form as the rough analogy between the model and certain phases of urinary secretion might at first suggest. The model here described was, as a matter of fact, constructed to give tangible evidence to conceptions of urinary secretion which familiarity with the well-known facts of kidney physiology and my own experiments had led me to formulate in my mind. Just how far I think the model simulates conditions observed in experiments on the kidneys will appear later. We will take up now a series of experimental findings on the secretion of urine which it seems to me can be interpreted, in the light of our knowledge regarding colloids, in a different and simpler way than is generally done. Our discussion will deal separately with the subjects of the secretion of water by the kidneys, and the secretion of substances dissolved in the water. Here too, many writers in physiology and pathology to this day look upon the two as parallel processes. As a matter of fact, they constitute separate problems, and should be dealt with separately. 286 (EDEMA AND NEPHRITIS 4. The Output of Water by the Kidney We are not surprised to find that the secretion of urine ceases (practically) during absolute starvation. If the colloids of the body as a whole are holding on to all the available water, if, in other words, an amount more than necessary to saturate them is not present, then none is left over to be secreted. Only as the tissues undergo gradual consumption during the process of starvation or their colloids suffer changes which decrease their capacity for holding water is any liberated. On the other hand, if a non-thirsting organism (as I will, for short, call one whose colloids are saturated with water) consumes a quantity of water, an amount of urine is excreted (skin and lung ignored), after a variable length of time, which is equivalent to the amount of water that was drunk. (Not to do so means the development of an oedema.) It does not matter how this water was con- sumed. It may simply have been swallowed or have been experimentally introduced into the gastro-intestinal tract, or it may have been injected into the peritoneal cavity, under the skin, or directly into the blood. By a process of colloid absorption we ultimately reduce all these to one, namely to the presence of water in the blood. What becomes of the water after it has gotten into the blood, as into the venous blood returning from the alimentary tract, is our next problem. In its transit through the lungs the venous blood coming from the ahmentary tract loses the carbonic acid responsible for the increased hydration capacity of its colloids. As soon as this has happened the blood contains more water than the blood colloids are capable of holding, and so this separates off as urine or as some other secretion. ^ In order to obtain a secretion from the kidney {or any other gland) conditions the reverse of those which favor absorption must be established. Thus, highly venous blood favors absorption, but secretion can occur only when a gland is supplied with arterialized blood (blood low in carbonic and other acids and high in oxygen) . Let us now consider some experiments which prove the truth of these contentions. 1 1 have often been asked why the secretion does not occur into the lungs (where the carbonic acid escapes). The truth is that just as much water is lost daily (by evaporation) through the lungs as through the kidneys. ABSORPTION, SECRETION— COMPLEX ORGANISM 287 Since the presence of free water in the blood is in my opinion the sine qua non for urinary secretion, let us first see what is the effect on secretion of introducing a certain amount of this into the circulation. Since distilled water is destructive to the red blood corpuscles we will inject instead a salt solution. To accomplish such injection without injury to the animal we make use of the apparatus shown in Fig. 90. The graduated cylin- der c is closed with a soft rubber stopper provided with two glass tubes bent as shown in the diagram and supplied with the rubber tubes a and h ending respectively in the injection needle n and the pressure bulb d. When d is compressed, any fluid contained in c is forced into the delivery tube. After the air is driven out of this the needle n is inserted into the ear vein of a rabbit and held in place by two small artery clamps. By pressure upon d the fluid may now be injected directly into the animal's cir- culation at any rate desired. The animal is comfortably tied into an animal holder and the urine is collected through a small soft rubber catheter in- serted into the bladder. Un- less poisonous substances are injected, the whole procedure does not injure an anunal in the sUghtest, so that it may (with suitable periods of rest between) be used over and over again. No operations are necessary, the animal suffers no pain and therefore there is no need for anesthetics. These prolific sources of error are hence eliminated. In Fig. 91 are shown the effects on urinary output in a rabbit of injecting intravenously the same amount (125.9 cc.) of dif- ferently concentrated sodium chlorid solutions.^ All the curves 1 James J. Hogan and Martin H. Fischer : Kolloidohem. Beihefte, 3, 304 (1912). Figure 288 CEDEMA AND NEPHRITIS Hours 1 FlGUBE 91. in this figure were obtained from the same animal. The lowermost curve a serves as a control and indicates normal urinary secretion. The curves are constructed by plotting time on the horizontal and the amount of urine secreted in fifteen minute intervals on the vertical. Curve b shows the effect of injecting at the uniform rate of 10 cc. every five minutes m/8 (0.729 per cent) sodium chlorid. As is clearly evident such in- jection soon increases the amount of urine secreted. Curve c shows the effect of injecting an equivalent amount of a sodium chlorid solution alleged to be more nearly "isosmotic " or "iso- tonic " with the tissue fluids of a rabbit, namely a 0.92 per cent solution. As is readily apparent, in spite of the fact that exactly the same amount of water was injected and at exactly the same rate the urinary out- put is still further increased. We also notice that the effect of this injection evi- dences itself sooner upon the secretion. Curves d and e show the effects of injecting m/4 (1.459 per cent) and ABSORPTION, SECRETION— COMPLEX ORGANISM 289 m/2 (2.909 per cent) sodium chlorid solution. Again, in spite of the fact that the same amount of water has been in- jected and at the same rate, the loss of water from the body occurs the more rapidly and is the greater the higher the con- centration of the injected salt. In these later experiments little or none of the injected water is retained in the body and if enough salt is injected with the water, the animal loses more water than was injected.^ How are we to interpret these simple facts? As previously emphasized, the whole animal, including his blood and lymph, represents a system of (hydrophilic) emulsion colloids which are saturated with water. This colloid system cannot take up any more water or give off any except as it first suffers chemical or physico-chemical changes which either increase or decrease the capacity of these colloids for holding water. If the system com- posing the body is saturated with water then it cannot, of course, take up any more, and so if " free "water in the form of a " phy- siological " (m/8 or 0.729 per cent) salt solution is injected, this cannot be retained, but must escape as urine (or some other secretion). But why does a stronger salt solution bring about an earlier increase in urinary output and a greater one? It is ordinarily said that this occurs because the salt " stimulates " the kidney in some mysterious way. The explanation is really simpler. The salts decrease the capacity of the body colloids (protein colloids) for water, and this the more the higher the concentra- tion of the salt. The higher, therefore, the concentration of salt in our injection mixture the higher must it be in the blood, and in consequence (after diffusion) in all the tissues of the body. By injecting a strong salt solution we therefore not only inject a certain amount of free water as before, but we make the tissues give off water. This " free " water is then added to that which we injected and it is the sum of the two which appears as urine. The administration of salt increases the water output through the kidney primarily, not because of any effect upon the kidney but because of an effect upon the body as a whole. Incidentally, we observe that contrary to much clinical teaching, sodium chlorid 'These facts are con&med by the experiments of H. Roger and Gahnier: Arch, de M6d. exp., Mai (1913); La Presse M^dicale, 885 (1913) who, however, jnake no attempt to interpret their findings. 290 (EDEMA AND NEPHRITIS administration does not lead to a retention of water in the body and thus to an cedema, but just the reverse. This is in harmony with our previous observations on simple colloids. Curves a, b, c, d and e of Fig. 91 have been constructed from the observations contained in Experiments 12, 13, 14, 15 and 16. Experiment 12. Normal Urinary Secretion. — Belgian male rabbit. Weight 2495 grams. Fed on mixed standard diet. Tied into animal holder and catheterized. No anesthesia. Time. Urine in cc. Remarks. 11 15 Tied down, catheterized. 11.30 — 11.45 1.6 Clear, yellow, alkaline, no albumin. 12.00 6.4 Same. 12.15 4.8 Same. 12.30 0.9 Same. 12.45 1.7 Same. 1.00 1.3 Same. Animal released in good condition. Total urine in the period of orle and three-quarter hours, 16.7 cc. ExPEEiMENT 13. Injection Fluid: m/8 (0.72 per cent) NaCl. — Belgian male rabbit. Weight 2495 grams. Kept on standard mixed diet of clover hay, oats, corn and greens. 124.9 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. It was estimated that this amount was equivalent to the total volume of blood of the animal. No anesthetic. L ' Time. Urine in cc. Remarks. 2.15 21.5 Tied down, catheterized, injection begun, yellow, neutral, no albumin. Urine, clear 2.30 0.3 Same. 2.45 1.0 Clear, yellow, faintly alkaline, no albumin. 3.00 1.9 Same. 3.15 7.7 Same. Injection stopped. 3.30 14.0 Clear as water, faintly alkaline, no albumin. 3.45 14.0 Same. 4.00 14.0 Same. 4.15 16.0 S'ame. 4.30 11.2 Same. 4.45 4.8 Same. Animal released in good condition. Total urine in two and one-half hour-period since beginning of injec- tion: 84.9 cc. Urine secreted in the first two hours: 68.9 cc. ABSORPTION, SECRETION— COMPLEX ORGANISM 291 Experiment li.— Injection Fluid: 0.92 per cent NaCl.— Belgian male rabbit. Weight, 2495 grams. Kept on standard mixed diet. 124.9 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every 5 minutes. This amount was estimated as equivalent to the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks, 9.30 2.5 Tied down, catheterized, injection begun. Urine clear, yellow, no albumin. 9.45 7.5 Clear, yellow, alkaline, on albumin. 10.00 24.0 Clear aa water, yellow, faintly alkaline, no albumin. 10.15 22.0 Same, 10.30 21.0 Clear as water, neutral, no albumin. Injection ended. 10.45 5,0 Clear as water, yellow, faintly alkaline, no albumin. 11.00 5.2 Cloudy, pale yellow, faintly alkaline, no albumin. 11.15 11.1 Same. 11,30 9.5 Yellow, alkaline, no albumin. 11.45 8.5 Same, 12.00 8.2 Same, Animal released in good condition. Total urine in two and one-half hour-period since beginning of injec- tion; 123.3 cc. Urine secreted in the first two hours: 105.3 cc. Experiment 15. Injection Fluid: m/4 (1.45 per cent) NaCl. — Belgian male rabbit. Weight 2495 grams. Kept on standard mixed diet. 124.9 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. This amount was estimated as equiv- alent to the total volume of blood of the animal. No anesthetic. Time, Urine in cc. Remarks, 9 45 3,0 Cloudy, yellow, alkaline, no albumin. 10.00 19,5 Clearing, alkaline, no albumin. 10,15 36.0 Clear as water, faintly alkaline, no albumin. 10,30 41.0 Same, 10.45 34.5 Same, Injection ended. 11,00 18.5 Same, 11,15 8,5 Same, Animal released in good condition. Drinks water at once. Total urine in the one and one-half-hour period since beginning of injection: 159.0 cc. Experiment 16. Injection Fluid: m/2 (2.8 per cent) NaCL— Belgian male rabbit. Weigh 2495 grams. Kept on standard mixed diet. 124.9 cc. of the above solution are injected into ear vein at uniform 292 (EDEMA AND NEPHRITIS rate of 10 cc. every five minutes. This amount was estimated as equivalent to the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks. 10.45 Tied down, catlieterized, injection begun. Cloudy, yellow, alkaline, no albumin. 4.0 11.00 64.0 Clearing, alkaline, no albumin. 11.15 84.0 Clear as water, neutral, no albumin. 11.30 48.0 Same. 11.45 28.0 Same. Injection ended. 12.00 33.0 Same. 12.15 3.5 Same. 12.30 22.0 Same. 12.45 6.0 Same. Animal released in good condition. Drinks water at once. Total urine in two-hour period since beginning of injection: 288.5 cc. The correctness of these views can at once be tested by sub- stituting for the simple salt solutions used above, one in which the water is not " free," but united to a colloid. The natural liquid which has these properties is, if our reasoning is correct, blood itself. We should therefore expect that the injection of no amount of blood would yield any increased flow of urine. That it does not is a familiar fact since the experiments of E. PoNFicK 1 and R. Magnus,^ only its interpretation has until now been missing. Curves a, b, c and d of Fig. 92 and the Experiments 17, 18, 19 and 20, from which they were constructed, show that the injection of a solution in which all the water is united to a colloid leads to no in- crease in urinary output. In these experiments we did not inject whole blood, but blood serum from freshly drawn horse blood obtained under aseptic conditions and permitted to coagulate in an ice box. Experiment 17. Injection Fluid: The Serum of Horse Blood. — White Belgian rabbit. Weight 1692 grams. Kept on standard mixed diet. 42.4 cc. of the serum are injected into ear vein at uniform rate of IE. Ponfick: Virchow's Archiv, 62, 277 (1875). 2 11. Magnus: Arch. f. exp. Path. u. Pharm., 46, 210 (1901). FiGUBE 92. ABSORPTION, SECRETION— COMPLEX ORGANISM 293 10 cc. every 5 minutes. This amount was estimated as equivalent to one-half the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks. 10.45 Tied down, catheterized, injection begun. 4.6 Thick, yellow, alkaline, no albumin. 11.00 1.2 Alkaline, no albumin. 11.07}^ Injection ended. 11.15 3.0 Same. 11.30 9.0 Clear, alkaline, traces ot albumin. 11.45 3.0 Alkaline, albumin, no casts. 12.00 0.8 Same. 12.15 0.6 Alkaline, less albumin, no casta. 12.30 1.2 Same. 12.45 0.7 Alkaline, faint trace of albumin, no casts. 1.00 0.4 Strongly alkaline, trace of albumin. 1.15 0.8 Same. 1.30 0.7 Same. 1.45 0.5 Same. 2.15 0.5 Same. 2.45 0.6 Same. Animal released in good condition. Total urine in four-hour period since beginning of injection: 23.0 cc. Urine secreted in first two hours: 19.9 cc. Experiment 18. Injection Fluid: The Serum of Horse Blood. — White and Belgian male rabbit. Weight 1585 grams. Kept on standard mixed diet. 79.1 cc. of the serum are injected into ear vein at uniform rate of 10 cc. every five minutes. This amount was estimated as equivialent to the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks. 11 25 Tied down, catheterized, injection begun. 2.5 Yellow, alkaline, no albumin. 11.40 11.55 12 02 1.0 0.6 Same. Cloudy, alkaline, no albumin. 12.10 12.25 1.3 5.5 Cloudy, alkaline, trace of albumin. Cloudy, faintly alkaline, albumin, no casts. 12.40 12.55 l.IO 3.5 3.5 3.5 Same. Slightly cloudy, alkaline, albumin. Alkaline, trace of albumin. 1.25 1.55 2.25 2.55 3.5 3.5 4.2 0.6 Same. Cloudy, alkaline, red blood corpuscles, no casts. (Presence of blood due to traumatic bleeding in the bladder?) . Trace of albumin, red blood corpuscles, no casta. Bloody, alkaline, albumin, red blood corpuscles, no casts. Animal released in good condition. Total urine in the three and one-half hour period since beginning of injection: 31.1 cc. Urine secreted in the first two hours: 22.8 cc. 294 (EDEMA AND NEPHRITIS ExPEBiMENT 19. Injection Fluid: The Serum of Horse Blood. — Himalaya rabbit. Weight 1564 grams. Kept on standard mixed diet. 117.3 cc. of the serum are injected into the ear vein at uniform rate of 10 cc. every five minutes. This amount was estimated as equiv- alent to one and one-half times the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks. 3.30 1.0 Thick, yellow alkaline, no albumin. 3.45 0.3 Same. 4.00 2.6 Tliick, yellow, strongly alkaline, trace of albumin. 4.15 2.6 Thick, yellow, strongly alkaUne, albumin. 4.30 3.0 Alkaline, albumin. Injection ended. 4.45 4.7 AlkaUne, more albumin, some traumatic blood. 5.15 4.2 Alkaline, more albumin, red blood corpuscles. 5.30 0.3 Albumin. 5.45 0.2 Albumin. Animal released in good condition. Total urine in the two hour period after beginning of injection: 17.7 cc. Experiment 20. Injection Fluid: The Serum of Horse Blood. — White male rabbit. Weight 1371 grams. Kept on standard mixed diet. 110 cc. of the serum are injected into ear vein at uniform rate of 10 cc. every five minutes. This amount was estimated as equivalent to one and two thirds times the total volume of blood of the animal. No anesthetic. Time. Urine in cc. Remarks. 4.45 Tied down, catheterized, injection begun. Clear, yellow, acid, no albumin. 5.00 4.2 5.15 0.5 Clear, no albumin. 5.30 0.4 Clear, faint trace of albumin. Injection ended, animal in good condition. 5.45 0.4 Clear, trace of albumin. 6.00 0.3 Thick, bloody, more albumin. 6.02 Total urine since beginning of injection: 2.4 cc. Autopsy: The heart and blood vessels are filled with blood, peritoneal and pleural cavities are empty. The As we proceed we shall find further experimental evidence for the truth of these views. The injected blood serum remains in the blood vessels. To the important physiological and thera- peutic consequences of this we shall also return later. ABSORPTION, SECRETION— COMPLEX ORGANISM 295 5. On the Colloid-chemical Action of the Diuretic Salts.^ the Saline Diuretics Produce Diuresis How Under ordinary circumstances the water output from the kidneys occurs so uninterruptedly and within such " normal " limits that we take it for granted. But in physiology, in pharmacology and particularly in the practical medicine of every day, when for any reason we observe a diminution in urinary out- put, then a discussion of " diuretics " develops, and of the means of increas- ing the observed urinary output. In the end certain " saline diuretics " may be prescribed which we know from both physiological and clinical experience to be capable of increasing the urinary output. How such diu- retics act has been much debated. As the following shows, the saline diu- retics are nothing but those salts which without being markedly -poisonous are the most -powerful dehydrants of the body colloids. They owe their action primarily not to any effect upon the kidney, but to an effect upon the body as a whole. By diffusing into the tissues of the body they liberate water from them, and their diuretic activity is but an expression of the amount of water they are thus able to liberate. The proof for this is easily brought. We need but make use of the pro- cedure previously employed and inject intravenously into animals equimolar amounts of different salts. When definite volumes (175 cc.) of equimolar solutions are thus injected at a uniform rate into rabbits they ' Martin H. Fischer and Anne Sykes: Science, 37, 845 (1913); Kolloid- Zeitschr., 13, 112 (1913). Figure 93. 296 (EDEMA AND NEPHRITIS lead to an increased output of urine which parallels the order in which they dehydrate protein colloids. For purposes of control Curve x of Fig. 93 and Experiment 21 are introduced. This shows the normal urinary output in a rabbit kept on our standard laboratory diet of hay, oats, corn and greens when simply tied comfortably into an animal holder and catheterized. The remaining curves with their corresponding experiments need no further explanation, for the experimental conditions were the same in all cases. The difference lies in the various salts injected. After a number of trial experiments we adopted the effects of injection of m/4 sodium chlorid solution as a standard for comparison. Curve a of Fig. 93 shows the urinary output of a rabbit when injected intravenously with such a solution. If a part of the sodium chlorid is replaced by an equimolar amount of magnesium chlorid, strontium chlorid or calcium chlorid, the urinary output is markedly increased. In place of the 147.4 cc. of urine secreted in the control experiment with sodium chlorid, we now obtained 179.3 cc, 185.5 cc. and 224.2 cc, respectively. The bivalent metals act as more powerful diuretics than the monovalent, and the order in which they produce diuresis is the same as the order in which they dehydrate simple (protein) colloids in test-tube experiments. This is shown by the curves of Fig. 93 as well as Experiments 22, 23, 24 and 25, from which they are constructed. Experiment 21. Normal Urinary Secretion. — ^White male rabbit F. Weight 2639 grams. Kept on standard mixed diet. Tied into animal holder and catheterized. No anesthetic. Time. Urine in cc. Remarks. 10 30 10.45 11.00 11.15 0.6 0.8 0.2 Thick:, yellow, alkaline, no albumin, no sugar, no casta. Same. Same. 11.30 11.45 12.00 12.15 12.30 3.0 1.5 2.0 1.5 2.0 Same. Same. Same. Same. Same. Animal released in good condition. Total urine in two-hour period, 10.0 cc. ABSORPTION, SECRETION— COMPLEX ORGANISM 297 Experiment 22. Injection Fluid: m/4 NaCl. — White male rabbit B. Weight 2129 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 10.30 Tied down, catheterized, injection begun. 10.0 Alkaline, no albumin, no casts. 10.45 0.4 Same. 11.00 9.0 Same. 11.15 11.0 Same. 11.30 19.5 Same. 11.45 32.0 Clear as water, alkaline, no albumin, no casts. 11.57J^ Injection ended. Same. 12.00 41.0 12.15 22.5 Same. 12.30 12.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 147.4 cc. Experiment 23. Injection Fluid: 180 cc. m/4 NaCl-|-20 cc. m/4 MgClj. — ^White male rabbit B. Weight 2146 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11.15 Tied down, catheterized, injection begun. 0.5 Yellow, faintly alkaline, no albumin, no casts. 11.30 2,0 Clearing, alkaline, no albumin, no casts. 11.45 13.5' Slightly cloudy, no casta. 12.00 19.5 Same. 12.15 36.5 Same. 12.30 36.5 Slightly cloudy, alkaline, sugar present, no casts. 12.45 34.0 Clear, neutral, no albumin, sugar present, no casts. 1.00 19.0 Same. 1.15 18.3 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 179.3 cc. Experiment 24. Injection Fluid: 180 cc. m/4 NaCl-|-20 cc. m/4 SrCl2.— White male rabbit B. Weight 2170 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. 298 (EDEMA AND NEPHRITIS Time. Urine in oo. Remarks. 2.45 Tied down, catheterized, injection begun. 5.0 Yellow, faintly alkaline, no albumin, no eugar, no casts. 3.00 2.2 Same. 3.15 25.0 Same. 3.30 33.0 Slightly cloudy, faintly alkaline, no albumin, sugar pres- ent, no casts. 3.45 34.5 Same. 4.00 31.5 Same. 4.12H 4.15 37.0 Same. 4,30 14.0 Same. 4.45 8.4 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection: 185.5 cc. Experiment 25. Injection Fluid: 180 cc. m/4 NaCl+20 cc. m/4 CaCla. — ^White male rabbit J. Weight 2214 grams. Kept on standard mixed diet. 176 cc. of the above solution are .injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 3.15 Tied down, catheterized, injection begun. Tliick, cloudy, alkaline, no albumin, no casts. Same. Clearing, neutral, no albumin, no casts. Same. Clear as water, neutral, no albumin, no casts. Same. Injection ended. Same. Same. Same. 3.30 3.45 4.00 4.15 4.30 4.42M 33.0 2.0 18,0 38.0 66.0 51.0 4.45 6.00 5.15 38,0 18.0 3.2 Animal released in good condition. Total urine in two-hour period since beginning of injection, 224.2 cc. Feces, 16 grams. The diuretic action of salts with different acid radicals is shown in Figs. 94 and 95. In both figures the sodium chlorid curve', constructed from Experiment 26, is introduced as a control. Tf the sodium chlorid is entirely or partly replaced by another sodium salt, we find the effect on urinary secretion to be as follows: ABSORPTION, SECRETION— COMPLEX ORGANISM 299 Salt. Urine in cc. Sodium Chlorid 127.5 Sodium Nitrate 129.4 Sodium Bromid 181.3 Sodium Acetate 182 Di-Sodium Hydrogen Phosphate 184.2 Sodium Di-Hydrogen Phosphate 204 . 4 Sodium lodid 237.5 Soium Sulphate 247 Hours 1 FlQUKE 94. Hours 1 2 Figure 95. 300 (EDEMA AND NEPHRITIS With the exception of the nitrate and the iodid, the diuretic action of the acid radicals parallels completely their dehydrating effect upon (protein) colloids. In the amount and concentration employed the nitrate shows a markedly poisonous effect. This is possibly the reason why it stands lower in the series than we should expect. The iodid, on the other hand, stands unexpectedly high. But as we employed a higher concentration of the iodid than of the acetate, the phosphate or the sulphate, the use of a greater absolute amount of the iodid more than compensated for its less powerful action. The curves of Figs. 94 and 95 are constructed from Experi- ments 26, 27, 28, 29, 30, 31, 32 and 33. Experiment 26. Injection Fluid: m/4 NaCl. White male rabbit J. Weight 2169 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 12.00 Tied down, catheterized, injection begun. 5.0 Cloudy, alkaline, no albumin, no casts. 12.15 1.0 Same. 12.30 7.0 Clearing, otherwise the same. 12.45 10.0 Same. 1.00 17.0 Same. 1.15 29.0 Same. 1.27H Injection ended. 1.30 38.5 Same. 1.45 12.0 Same. 2.00 13.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 127.5 cc. Experiment 27. Injection Fluid: m/4 NaNOa. White and Belgian female rabbit. Weight 1393 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 3.00 Tied down, catheterized, injection begun. Yellow, alkaline, no albumin, no sugar, no casts. 0.5 3.15 1.4 Same. 3.30 9.6 Same. 3.45 23.0 Clear as water, otherwise the same. 4.00 36.0 Same. 4.15 37.0 Clear, neutral, trace of albumin, sugar present, red blood corpuscles, no casts. 4.27H 4.30 16.0 Same. 4.45 5.4 Same. 5.00 1.0 Clear, alkaline, no albumin, sugar present, no casts. ABSORPTION, SECRETION— COMPLEX ORGANISM 301 Animal released. Total urine in two-hour period since beginning of injection, 129.4 cc. For some hours after its release the animal seems exhausted and takes no food. Next morning is again in good condition. Experiment 28. Injection Fluid: m/4 Sodium Bromid. White and Belgian female rabbit. Weight 1393 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11.00 Tied down, catheterized, injection begun. Clear, yellow, alkaline. 1.0 11.15 2.8 Yellow, alkaline, traces of albumin, no sugar, many blood cells, many casts. 11.30 20.5 Same. 11.45 23.0 Clear as water, alkaline, faint trace of albumin, no sugar, occasional casts. 12.00 38.0 Clear, alkaline, very faint trace of albumin, sugar present, no casts. 12.15 34.0 Same. 12.271^ Injection ended. 12.30 36.0 Same. 12.45 17.0 Same. 1.00 10.0 Same. Animal released in sleepy condition. Total urine in two-hour period siace beginning of injection, 181.3 cc. Experiment 29. Injection Fluid: 100 cc. m/4 NaCl+lOO cc. m/4 Sodium Acetate. White male rabbit J. Weight 2419 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 2.45 Tied down, catheterized, injection begun. No urine. 3.00 3.15 3.30 3.45 4.00 4.12H 33.0 31.0 34.0 34.0 Clear, alkaline, no albumin, no sugar, no casts. Same. Same. Clear, alkaline, no albumin, trace of sugar, no casts. Injection ended. Same. Same. Same. 4.15 4.30 4.45 36.0 8.0 6.0 Animal released in good condition. Total urine in two-hour period since beginning of injection, 182 cc. 302 CEDEMA AND NEPHRITIS Experiment 30. Injection Fluid: 180 cc. m/4 NaCl+20 cc. m/4 NajHPOi. Yellow and white male rabbit. Weight 1444 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11.45 Tied down, catheterized, injection begun. Tliick, yellow, alkaline, no albumin, no sugar, no casts. 5.5 12.00 1.5 Clearing, otherwise the same. 12.15 5.2 Same. 12.30 27.0 Same. 12.45 41.0 Same. 1.00 42.0 Clear as water, faintly alkaline, no albumin, trace of sugar, no casts. 1 121^ 1.15 37.5 Same. 1.30 20.0 Same. 1.45 10.0 Same. Animal released. Found dead in cage next morning. Total urine in two hour-period since beginning of injection, 184.2 cc. Experiment 31. Injection Fluid: 180 cc. m/4 NaCl-|-20 cc. m/4 NaH2P04. Black male rabbit C. Weight 2277 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11.30 Tied down, catheterized, injection begun. Yellowish-brown, alkaline, no albumin, no sugar, no 8.6 casts. 11.45 1.8 Same. 12.00 12.5 Clearing, otherwise the same. 12.15 33.5 Same. 12.30 52.0 Same. 12.45 60.0 Clear as water, faintly alkaline, no albumin, trace of sugar, no casts. 12.57K 1.00 36.0 Same. 1.15 7.0 Same. 1.30 1.6 Same. Animal teleased in good condition. Total urine in two-hour period since beginning of injection, 204.4 cc. Experiment 32. Injection Fluid: m/4 Nal. Black male rabbit C. Weight 2539 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. ABSORPTION, SECRETION— COMPLEX ORGANISM 303 Time. Urine in cc. Remarks. 3.15 Tied down, catheterized, injection begun. Tliick, amber, alkaline, no albumin, no sugar, no casts. 4.0 3.30 1.0 Same. 3.45 15.0 Same. 4.00 43.0 Clear, neutral, no albumin, trace of sugar, no casts. 4.15 40.0 Same. 4.30 42.5 Same. 4.42J.^ Injection ended. 4.45 48.0 Same. 5.00 30.0 Same. 5.15 18.0 Same. Animal released. Total urine in two-hour period since beginning of injection, 237.5 cc. Animal found dead in cage three days later. Autopsy: Decomposition advanced. Hemorrhagic spots observed in kidney. ExpEEiMENT 33. Injection Fluid. 100 cc. m/4 NaCl+100 cc. m/4 Na^SO,. White male rabbit P. Weight 2670 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 10.30 Tied down, catheterized, injection begun. 10.45 11.00 11.15 11.30 11.45 11.57}^ 12.00 12.15 12.30 4.3 2.4 33.6 38.0 43.0 47.0 Clear, amber, acid, no albumin, no sugar, no casts. Same. Clear as water, neutral, no albumin, no sugar, no casts. Same. Same. Same. Injection ended. Clear, no albumin, trace of sugar, no casts. Same. Same. 58.0 16.0 9.0 Animal released in good condition. Total urine in the two-hour period since beginning of injection, 247 cc. Salts made up of a powerfully dehydrating base with a power- fully dehydrating acid have, of course, the most effect upon (protein) colloids. Magnesium sulphate is an example of such a combination. Therefore, if our theory is true, we should expect this salt to produce a greater diuresis than any yet described. That this is actually the case is shown by the curve in Fig. 96 as well as the corresponding Experiment 34. With injection of magnesium sulphate we obtained a urinary output of 300 cc. 304 (EDEMA AND NEPHRITIS Experiment 34. Injection Fluid: 180 cc. in/4 NaCl+20 cc. m/4 MgS04. White male labbit J. Weight 2407 grams. Kept on standard mixed diet. 175 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time._ Urine in cc. Remarks. 10.45 Tied down, catheterized, injection begun. 45.0 Cloudy, yellow, alkaline, no albumin, no sugar, no casts. 11.00 16.0 Same. 11.15 44.0 Almost as clear as water, alkaline, no albumin, trace of sugar, no casts. 11.30 45.0 Clear as water, faintly alkaline, no albumin, much sugar, no casts. 11.45 46.5 Same. 12.00 51.5 Same. 12.12M Injection ended* 12.15 59.0 Same. 12.30 21.0 Same. 12.45 17.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 300 cc. The analogies between these various experimental facts on the secretion of urine, and the model previously described and constructed with a view to elucidating some of these facts need no special comment. In our model we have eliminated the colloids of the blood stream, and a colloid reservoir of water corresponding with the water-saturated body colloids by work- ing with simple salt solutions only. As our considerations have shown, the colloids of the blood and of the tissues generally play a part in urinary secretion in so far as they furnish storage places for water which is liberated under the conditions employed in the described experiments on diuresis. Our model, therefore, illustrates particularly the changes that occur in the kidney (which represents physico-chemically only a much folded col- loid membrane). In our model the "secretion" of "urine" is induced through a hydrostatic pressure which forces a liquid through the capillary bed formed by the powdered fibrin. When an acid is brought in contact with the fibrin it swells, closes the capillary pores, and in proportion to the amount of this closure a formerly effective hydrostatic pressure becomes less and less capable, or, finally, even entirely incapable of forcing any liquid through this bed. We can counteract the effect of this acid by various salts, wherein again we find the saline diuretics to be more effective than certain other common salts. Such ABSORPTION, SECRETION— COMPLEX ORGANISM 305 a salt as sodium sulphate, of course, makes the fibrin shrink more than an equally concentrated sodium chlorid, and so the former, by increasing the size of the capillary pores, more than the latter favors an increased " secre- tion " in our model. Matters in the living kidney are, of course, not so simple as in this model. To mere changes in hydrostatic (blood) pres- sure, per se, I think but little signifi- cance can be attached. Enormous pressures (not such as we encounter in the body) are needed to filter liquids through even very thin colloid membranes. In the changes in the surface tension of liquids we deal with more adequate forces.^ Nor would I be understood as of the opinion that the colloid membrane consti- tuting the kidney is grossly capillary in character as is our powdered fibrin. But in an analogous behavior of the so-called " microcapillary " character of the colloid membrane separating the urine from blood I feel quite confident that the solution of our problem lies. 6. On the Colloid-chemistry of Sugar Diuresis 2 While the various non-electrolytes as compared with the electrolytes do not produce a great dehydration o Hours i of protein colloids, some of them exhibit considerable acti/ity in this 1 While I do not feel that the ideas of I. Thaube on this subject are en- tirely satisfactory in detail, his general conceptions of the r61e of surface tension in secretion strike me as most suggestive and valuable. 2 Martin H. Fischer and Anne Stkes: Science, 38, 486 (1913); Kolloid- Zeitschritt, 14, 223 (1914). 306 (EDEMA AND NEPHRITIS 70 - / / Hours 1 FiGTJBE 97. regard, notably the sugars. The ex- periments which follow reveal their similar dehydrating action upon the whole "living" animal. They there- fore do away with the behef of various authors that the dehydrating effects of non-electrolytes on tissues and or- ganisms as a whole, furnish support for the " osmotic " theory of water absorption. At the same time the absurd view is again met that although the colloid-chemical theory explains water absorption in dead tissues its laws do not hold for living animals; for it need scarcely be said that our rabbits were alive. It will now he shown that the diuretic action of the sugars parallels their dehy- drating effect on protein colloids, and that like the diuretic salts previously discussed, the sugars owe their action primarily not to any effect upon the kidney, but to an effect upon all the tissues of the hody generally. With any sugar the degree of diuresis in- creases with every increase in the concentration. When we compare the degree of increase in urinary output with that in concentration we find, roughly, that doubhng the concentra- tion of sugar more than doubles the urinary output. This is just the re- verse of what happens with salt solu- tions where the lower concentrations produce relatively greater effects than the higher. This difference in the be- havior of sugars and of salts upon rab- bits parallels their effects upon simple protein colloids. At the same concen- tration the three sugars produce differ- ABSOKPTION, SECKETION— COMPLEX ORGANISM 307 ent degrees of diuresis just as they bring about different degrees of dehydration in gelatin or fibrin, dextrose and levulose stand- ing very close together, while saccharose is far more powerful. Figs. 97, 98, and 99 present in graphic form the results obtained from Experiments 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45 and 46. Experiment 35. Injection Fluid: m/4 dextrose. Gray rabbit R. Weight 2500 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 2.15 Tied down, catheterized, injection begun. Yellow, allcaline, no albumin, no casta. 15.0 2.30 A few dropa Same. 2.45 4.5 Same. 3.00 4.5 Same. 3.15 4.0 Clear, neutral, no albumin, no casta. 3.30 10.0 Same. Injection ended. Animal released in good condition. Total urine in one and one-quarter-hour period since beginning of injection, 23 cc. Experiment 36. Injection Fluid: m/2 dextrose. White rabbit T. Weight 1600 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11 15 Tied down, catheterized, injection begun. Yellow, alkaline, no albumin, no casts. 0.5 11.30 3.0 Same. 11.45 12.0 Clear as water, alkaline, no albumin, no casts. 12.00 19.0 Same. 12.15 21.0 Same. 12.30 24.0 Same. Injection ended. 12,45 17.0 Same. 1.00 9.0 Same. 1.15 6.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 111 cc. 308 (EDEMA AND NEPHRITIS 70 - 50 40 30 Levulose \ Hours 1 Figure 98. 80 - 60 50 40 30 20 10 \ Saccharose Hours 1 FiGUBB 99. ABSORPTION, SECRETION— COMPLEX ORGANISM 309 Experiment 37. Injection Fluid: 3/4 m dextrose. Gray rabbit R. Weight 2500 grams. Kept on standard mixed diet. 150 cc. of the above solution. are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic Time. Urine in cc. Remarlcs. 2.30 Tied down, catheterized, injection begun. 30.5 Yellow, alltaline, no albumin, no casta. 2.45 3.0 Clearing, allcaline, no albumin, no casts. 3.00 10.0 Clear as water, alltialine, no albumin, no casts. 3.15 42.0 Same. 3.30 46.0 Same. 3.45 57.0 ffame. Injection ended. 4.00 33.0 Same. 4.15 3.5 Same. 4.30 2.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 230.5 cc. Experiment 38. Injection Fluid: m/1 dextrose. Belgian hare. Weight 1534 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11 30 Tied down, catheterized, injection begun. No urine. 11.45 25.0 Clear, neutral, no albumin, no casta. 12.00 78.0 Same. 12.15 62.0 Same. 12.30 45.0 Same. 12.45 50.0 Neutral, traumatic red blood corpuaclea.no casts. Injection ended. Animal released shivering and with teeth chattering. 1.10 Dies. Autopsy: Everything negative. Cavities empty; tissues dry; petechial hemorrhages in membrane of urethra and bladder. Kidney, pelvis and ureter uninjured. Total urine in one and one-quarter hour period since beginning of injection, 262 cc. Experiment 39. Injection Fluid: m/4 levulose. Yellow rabbit Smutty. Weight 2250 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. 310 (EDEMA AND NEPHRITIS Time. Urine in cc. Remarks. 2.30 Tied down, catheterized, injection begun. 2.45 3.00 3.5 Cloudy, alkaline, no casta. 3.15 5.5 Same. 3.30 5.5 Same. 3.45 10.0 Same. Injection ended. 4.00 9.0 Same. 4.15 4.0 Same. 4.30 5.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 41.5 cc. ExPEBiMENT 40. Injection Fluid: m/2 levulose. White rabbit B. Weight 1700 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No ane.'sthetic. Time. Urine in cc. Remarks. 11.30 Tied down, catheterized, injection begun. 9.0 Yellow, alkaline, no albumin, no casts. 11.45 5.0 Clear as water, otherwise the same. 12.00 20.0 Same. 12.15 30.0 Same. 12.30 31.0 Same. 12.45 35.0 Same. Injection ended. 1.00 15.0 Same. I.IS 7.0 Same. 1.30 4.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 147 cc. Experiment 41. Injection Fluid: 3/4 m levulose. Brown and white rabbit. Weight 1750 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 11.45 Tied down, catheterized, injection begun. Yellow, alkaline, no albumin, no casts. Clear, alkaline, no albumin, no casts. Same. Same. Same. Same. Injection ended. Same. Same. Same. 12.00 12.15 12.30 12.45 1.00 1.15 1.30 1.45 12.0 22.0 40.0 42.0 50.0 45.0 17.0 10.0 7.0 Animal released. Found dead in cage next morning. Total urine in two-hour period since beginning of injection, 233 cc. ABSORPTION, SECRETION— COMPLEX ORGANISM 311 Experiment 42. Injection Fluid: m/1 levulose. Black rabbit V. Weight 1530 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 3 15 Tied down, catheterized, injection begun. Yellow, alkaline, no albumin, no casts. 2.0 3.30 38.0 Clear, neutral, no albumin, no casts. 3.45 72.0 Same. 4.00 62.0 Same. 4.15 64.0 Same. 4.30 47.0 Same. Injection ended. 4.45 17.0 Same. S.OO 12.0 Same. 5.15 2.0 Same. Animal released. Dies in a, short time. Total urine in two-hour period since beginning of injection, 308 cc. Experiment 43. Injection Fluid: m/4 saccharose. Yellow rabbit Smutty. Weight 2250 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 2 45 Tied down, catheterized, injection begun. 6.0 Yellow, alkaline, no albumin, no casts. 3.00 3.0 Same. 3.15 22.0 Clear, neutral, no albumin, no casts. 3.30 22.0 Same. 3.45 23.0 Same. 4.00 25.0 Same. Injection ended. 4.15 15.0 Same. 4.30 11.0 Same. 4.45 9.0 Same. Animal released in good condition. Total urine in two-hour period since beginning of injection, 130 cc. Experiment 44. Injection Fluid: m/2 saccharose. Belgian rab- bit H. Weight 1800 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. 312 OEDEMA AND NEPHRITIS Time. Urine in cc. Remarks. 2 45 Tied down, catheterized, injection begun. 10.0 Yellow alkaline, no albumin, no easts. 3.00 23.5 Clear as water, neutral, no albumin, no casts. 3.15 73.5 Same. 3.30 61.0 Same. 3.45 59.0 Same. 4.00 41.0 Same. Injection ended. 4.15 24.0 Same. 4.30 10.0 Same. 4.45 3.0 Same. Animal released limp and shaking. Next morning alive and well. Five days later alive and well. Total urine in two-hour period since beginning of injection, 295 cc. Experiment 45. Injection Fluid: 3/4 m saccharose. Yellow rabbit Sammy. Weight 2250 grams. Kept on standard mixed diet. 150 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 12.00 Tied down, catheterized, injection begun. No urine. 12.15 10.5 Clear, neutral, no albumin, no casts. 12.30 63.0 Same. 12.45 74.0 Same. 1.00 87.5 Same. 1.15 38.0 Same. Injection ended. 1.30 36.0 Same. 1.45 18.0 Same. 2.00 16.0 Same. Animal released. Found dead in cage next morning. Total urine in two-hour period since beginning of injection, 343 cc. Experiment 46. Injection Fluid: m/1 saccharose. Black rabbit U. .Weight 1500 grams. Kept on standard mixed diet. 90 cc. of the above solution are injected into ear vein at uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarks. 3.15 3.30 3.45 4.00 4.05 32.5 75.0 56.5 No urine. Clear, neutral, no albumin, no casts. Same. Same. ABSOKPTION, SECRETION— COMPLEX ORGANISM 313 When we compare the diuretic action of the sugars with that of the salts we note the same interesting differences as when these two classes of substances are compared in their effect upon the dehydration of protein colloids. Thus, when the same amount of differently concentrated salt solutions (sodium chlorid, for example), are injected intravenously ' a relatively greater effect is produced by the weaker solutions than by the more concentrated, while just the reverse is the case when the sugars are used. This is clearly apparent in the following table :^ NaCl. Dextrose. Levulose. Saccharose. m/8 68.9 cc. m/4 159 cc. m/4 23 cc. m/2 111 cc. 3/4 m 230.5 cc. m/4 40 cc. m/2 147 cc. 3/4 m 231 cc. m/4 130 cc. m/2 295 CO. 3/4 m 343 CO. m/2 288.5 cc. The table also shows how tremendously saccharose dehy- drates. Easy as it is to understand on a colloid basis these differences between the sugars or between the sugars and the electrolytes, equally impossible is it to interpret them or any " osmotic " ground. With these experiments we think that the colloid-chemical theory of water absorption and secretion by protoplasm has answered the last objections of those who oppose its claims to be considered the dominant if not the only factor in this problem. We believe that on the basis of the dehydrating effect of dextrose upon the tissues can be understood the dryness of the diabetic's tissues and his thirst, as well as the increased urinary output observed resulting from the excessive consumption of water in response to this thirst. A therapy which decreases the amount of circulating sugar in the diabetic organism (with- out increasing an " acidosis ") decreases thirst, water consump- tion and urinary output. ' See page 287; also the first edition of "CEdema " and James J. Hogan and Martin H. Fischer: KoUoidchem. Beihefte. Martin H. Fischer and Anne Sykes: Science, 37, 845 (1913); KoUoidZeitsohr., 13, 112. 2 It should be remembered that owing to dissociation m/8 NaCl solution is " osmotically " almost equivalent to m/4 solution of a non-electrolyte. 314 (EDEMA AND NEPHRITIS 7. The Kidney in Secreting Water Does Work. Discussion of Some General Conditions Influencing Water Output by the Kidney. Diuretics of the Second Order In the evidence thus far presented, which has shown that an output of water by the kidney is possible only as free water is brought to it and in proportion to the amount of water thus brought, we have tacitly assumed that the kidney is always capable of secreting the water when thus offered it. But in secreting the water, the kidney does not play a passive role. In order to transport the water from the blood out into the uriniferous tubules the kidney does work. In order to get a normal urinary output the kidney must not only have free water at its disposal, but must also be able to do this secretory work. We have now to discuss the conditions under which this secretory work may be done, in other words, this second factor necessary for an out- put of water from the kidney. Evidence that the kidney actually does work in secreting water is brought from several sides. There is the older well- established fact that the venous blood returning from an active kidney as well as the urine coming from it have a higher tem- perature than the arterial blood entering it. Second, we have the findings of Barcroft and T. G. Brodib ^ which show that an actively secreting kidney uses more oxygen and gives off more carbonic acid than a resting one, and that the amount of oxygen thus consumed and the amount of carbonic acid thus produced rises hand in hand with the amount of water secreted. Thirdly, E. Heilner^ found that when the urinary output in starving dogs and rabbits is increased by forcing water the carbonic acid elimination is increased. To use a homely simile, fuel is consumed in order to get the energy expressed in a separation of water from the kidney. As in any machine which makes possible such energy transformations many totally different causes may interfere with its smooth running, so also in the kidney many different and apparently disconnected agencies may serve to interfere with the normal energy trans- formations which permit a kidney to transport the free water ' Barcroft and T. G. Bhodie: Jour. Physiol., 32, 18 (1904); ibid., 33 52 (1905). 2E. Heilner: Zeitschr. f. Biol., 49, 373 (1908). ABSORPTION, SECRETION— COMPLEX ORGANISM 315 offered it in the blood over into the uriniferous tubules. ' Let us for a moment consider some of these. As the quantitative and qualitative output from a machine is dependent upon the quantity and nature of the materials fed into it, so also the secretion of urine by the kidney is dependent in a striking way upon the circulation. Not only can the " nor- mal " secretion of urine be increased through changes in the circulation, but it can still more strikingly be decreased. From histological studies, and, on the whole, very hypothetical reason- ings, W. Bowman (1842) first laid stress on the importance of the pressure under which the blood flows through the kidneys as a factor in determining the secretion of urine. This pressure idea was further developed and given an experimental basis by Carl Ltjdwig (1884) and his pupils. Through their work considerable evidence was advanced to show that changes in blood pressure, no matter how induced, are always followed by changes in the amount of urine secreted, and, on the whole, in this sense, that an increase in blood pressure is accompanied by an increased urinary secretion, while a decrease in blood pres- sure is followed by an opposite result. This long-accepted belief met a serious setback in the critical studies of R. Heidenhain.^ He showed very clearly that the parallelism between blood pres- sure and urinary secretion is by no means absolute. Not only does interference with the outflow of venous blood from the kidney — a condition associated with an increase rather than any decrease in blood pressure — lead to a fall in the amount of urine secreted easily equal to the fall encountered after inter- ference with the arterial influx of blood, but various diuretics which do not alter the blood pressure are known to bring about a decided increase in urinary secretion. (Some of the saline diuretics belong in this group, the behavior of which we have already discussed.) Heidenhain also showed that the experi- mental facts available in his day are best harmonized by saying that the velocity with which the blood passes through the kidney determines the amount of urinary secretion. But further than this he does not go with any mechanical or, to put it more generally, physico-chemical conception of urinary secretion. 'R. Heidenhain: Hermann's Handbuch d. Physiologie, 5, 309, Leipzig (1883). 316 (EDEMA AND NEPHRITIS In place of the teaching of Ludwig that a secretion of urine is primarily dependent upon a blood pressure, or Heidenhain's belief that the velocity with which the blood passes through the kidney is of primary importance, the experiments and cUnical observations at hand on this subject are best interpreted by saying that the normal urinary secretion is absolutely dependent upon an adequate oxygen supply to the cells constituting the parenchyma of the kidney. Any interference with this oxygen supply leads to a decrease in urinary secretion even to the point of absolute and permanent stoppage. Through a particularly favorable oxygen supply to the kidneys the secretion of urine may be increased above that ordinarily considered " normal." This interpretation meets with no experimental objections. Nature has seen to it that the kidneys shall not lack facilities for a plentiful supply of oxygenated blood by endowing them with strikingly large renal arteries. Any considerable inter- ference with the oxygen supply to the kidney is followed by a drop in urinary secretion. It does not matter how such an inter- ference is brought about. It may be brought about through a change in the action of the heart itself, such as a decrease either in the number or the force of the heart's contractions or both (vagus stimulation, myocarditis, valvular heart disease, dilata- tion). Or the deficiency in oxygen supply to the kidneys may be brought about through hemorrhage or through stimulation of vasomotor nerves whose effect tends, in the aggregate, to decrease the amount of oxygenated blood passing through the kidneys. Most effectively can the oxygen supply to the kidneys be diminished to any degree or be cut off entirely through com- pression of the renal artery from without (experimental ligation, clamping, tumor) or occlusion from within (arteriosclerosis, experimental or clinical embolism). The same result is accom- plished if the outflow of blood through the renal veins is suf- ficiently impeded (experimental hgation, tumor, passive con- gestion due to heart disease). An adequate oxygen supply to the kidney, on the other hand, favors the secretion of urine. This is evidenced by the fact that the removal of the various conditions outlined above (provided they have not acted too long) is followed by a reestablishment of the urinary secretion to normal. When special efforts are made to increase the oxygen supply to the kidneys, as by Ugating ABSORPTION, SECRETION— COMPLEX ORGANISM 317 several of the larger arteries that pass off the aorta, or by stimulat- ing vasomotor nerves which tend to increase the quantity of oxygenated blood passing through the kidneys, a secretion of urine in excess of that considered normal may be obtained. On the other hand, the most liberal supply of poorly oxygenated blood to the kidneys even without any other disturbance in the circulation (such as variations in blood pressure) is incapable of maintaining a normal secretion of urine even for a little while. On the basis of these facts we shall now be able to discuss and to understand the mode of action of certain drugs, exclusive of the saHne diuretics, which are capable of increasing the out- put of water through the kidney. This second type of diuretic owes its action primarily to its power of favoring the oxygen supply to the kidney. It is readily appreciated that in liberating water from a tis- sue its degree of swelling is reduced. Pressure upon the blood vessels lying within an organ is thereby removed and a better blood flow through the organ favored. Secondarily, therefore, a diuretic salt also brings about a better blood supply to all the organs of the body, including the kidney. On the other hand, the diuretic drugs of the second order in favoring a better oxygen supply to the tissues of the body favor the removal of the acid products of normal and abnormal metabolism, and so incidentally they further diuresis by furnishing " free " water. We are also famihar with drugs which can decrease the out- put of water from the kidney. They act in a way the opposite of the diuretics of the second order. We can best begin our discus- sion with them. It is a familiar fact that after the administration of morphin or atropin or of chloroform, ether or alcohol, in any considerable amounts, there is always a fall in urinary secretion that may at times amount to complete suppression. It is possible for these substances to lead to such suppression through action upon the kidneys alone. Under ordinary circumstances such a purely local action is, however, not to be anticipated. In fact, it is perfectly possible for a temporary suppression of urine to follow, say, a general anesthetic, without any changes in the kidneys themselves. The administration of all these anesthetics and of certain alkaloids is accompanied by such a state of lack of oxygen in the tissues of the body generally, as we have before described for 318 (EDEMA AND NEPHRITIS isolated organs. In consequence of this, the capacity of the colloids of all the tissues of the body (including the blood and the lymph) for holding water is increased above that considered normal. After administration of any of these drugs the body gen- erally, therefore, is holding on to its water with special avidity, so that none is left over to be free in the blood and so be excreted through the kidneys. This condition of the tissues after an anesthetic or a dose of morphin, for example, is evidenced not only by the lack of urinary secretion, but by the thirst complained of by the patient. As the patient gets over his anesthetic his urinary secretion not only comes up, but his thirst disappears, even though no water has been given. To get the described results it will be remembered that con- siderable amounts of these various drugs have to be administered. Such doses lead to a state of lack of oxygen in the tissues. Small doses of ether, alcohol, etc., increase the urinary output. In try- ing to say how this effect is brought about we have to remember the favorable conditions for secretion that are induced in the kidneys when these are given plenty of oxygen (while at the same time their carbonic acid is being rapidly carried away). Such conditions are brought about through the increased frequency and force of the heart beat, the more rapid breathing and the vasodilatation that are induced by small doses of these drugs. A large part of the diuretic action cf caffein and its various deriva- tives, as well as of digitalis, can also be understood on this basis. The drugs which make for an increased oxygen supply and a favored carbonic acid removal from the kidneys do the same for the body tissues generally. A decreased capacity of all the body colloids for holding water is, therefore, a natural result under such circumstances, in consequence of which water is liberated into the blood. This water then becomes available for urine. A dose of caffein or digitalis, therefore, not only puts the kidneys into a condition which favors the secretion of water by them, but at the same time aids in furnishing them water through an indirect effect upon the body colloids generally. These facts are of the greatest importance when we approach the practical problems of medicine, and we shall, therefore, return to discuss them further when we take up nephritis.^ ' See page 537. ABSORPTION, SECRETION— COMPLEX ORGANISM 319 8. Historical and Critical Remarks on Urinary Secretion We may apply a test to the reasoning of the preceding para- graphs by considering a few of the large number of valuable experimental studies available on the secretion of water by the kidney in the light of the ideas developed above. While the laboratory facts adduced by different authors harmonize well with each other their interpretation has been warmly debated. Let us begin by considering the results that have been obtained when various salt solutions have been injected directly into the blood. When a (" physiological ") 0.9 per cent sodium chlorid solution is injected into a rabbit, a secretion of urine is obtained which quickly equals the amount of salt solution infused. If a somewhat stronger salt solution is employed, more urine is secreted than is infused, and this difference between amount injected and amount excreted becomes the greater the higher the concentration of the sodium chlorid in the injection fluid. If the injections are very large, or are carried on for a long time, the absolute differences' between amount infused and amount secreted remain, but the relative become less and less apparent. ^ If no time limits are set upon the experiment, the end results become somewhat complicated (though not confusing), owing to the fact that the animal develops the symptoms and signs of a general oedema. This oedema is due to the oxygen want from which the animal suffers whenever these great injections of salt solution are continued sufficiently long. Isosmotic solutions of the chlorids, bromids and iodids of sodium or potassium bring about approximately the same excre- tion of urine.^ When, however, equally concentrated solutions of the various saline diuretics (phosphate, sulphate, tartrate or citrate of sodium) are injected, a much greater secretion of urine is obtained.^ How are these various experimental findings to be interpreted? 'See Martin H. Fischer: University of California Publications, Physi- ology, 1, 107 (1904). ' VON Limbeck: Archiv f. exp. Path. u. Pharm., 25, 89 (1888). » Magnus: Archiv f. exp. Path. u. Pharm., 44, 68 and 396 (1900); Ueber Diurese, Heidelberg (1900). Torald Sollmann: Arch. f. exp. Path. u. Pharm., 46, 13 (1901). B. Haake and K. Spiro: Hofmeister's Beitrage, 2, 149 (1902). 320 (EDEMA AND NEPHRITIS Let us first call attention to the important experimental error that is introduced into any of these experiments if an anesthetic is used. When enough is used to produce anesthesia, a lack of ox- ygen in the tissues and a retention of some of the Hquid infused may be expected to follow. The effect of an infusion is divis- ible into two parts: first, the effect of the water injected; second, the effect of the salt. Other things being equal, we may expect the water to behave, so far as diuresis is concerned, just as this be- haves when water only is injected. The salt injected has an effect upon the kidney and also upon the colloids of the body gen- erally, including those of the blood and lymph. As the chief salt of the body fluids is sodium chlorid, we are not surprised to find that a sodium chlorid solution " isosmotic " with the blood, when injected intravenously, yields in a short time an amount of urine about equal in volume to that of the salt solu- tion injected. If, however, a sodium chlorid solution having a concentration above that of the blood is used, an increased secretion of urine is obtained. This is because the salt acts not only directly upon the colloids of the blood and makes them liberate some of their water, but diffuses into the tissues of the body and makes the colloids here also give up a part of their water. This water is then " free," and can be secreted as urine. The salt also acts upon the colloids of the kidney, making the cells of this organ shrink. This shrinkage of the kidney cells necessarily means a decrease in the pressure exerted upon the blood vessels passing through the kidney, and so a better blood flow through this organ is also favored. The higher the concentration of the injected salt the more water must the body tissues yield up for diuresis (and the better must also be the blood supply to the kidney). We have no difficulty in understanding why isosmotic solutions of different salts are not equally effective in producing a diuresis. We have become familiar with the unequal effects of different salts on the absorption and secretion of water by colloids. Just as the sulphate, tartrate, phosphate and citrate of sodium are more effective in making fibrin or gelatin give up their water than the chlorid, bromid and iodid of this same metal, so are the former group expected to make the body colloids yield up a greater amount of water for diuresis than the latter. A similar difference of effect is to be expected upon the colloids constituting ABSORPTION, SECEETION— COMPLEX ORGANISM 321 the kidney. The first-named group must tend to make this " shrink " more than the second. Let us see how well our theory fares if we apply it to the care- fully worked-out ' experiments of Ernst Frey.^ This author does not, of course, interpret his experiments as I have taken the liberty of doing for him, but on the more generally accepted basis of alterations in the kidney and changes in blood pressure. Frey finds that when water is given a rabbit by mouth or rectum, or is injected intraperitoneally or into the small intestine, an increased amount of urine is secreted by the kidneys. I would say that this is because the tissues of the rabbit are satu- rated with water and so none of it is retained. If in place of water, a sodium chlorid solution is injected, the same or even a greater diuresis is obtained. This diuresis is the greater the higher the concentration of the salt solution injected (the amount of fluid injected being the same), just as in our own experiments already described. The diuresis following the introduction of water does not occur if any anesthetic is administered (morphin, chloral, ether, ure- thane). This is evidently because the anesthetics all produce a state of lack of oxygen, so that the tissues have an increased capacity for holding water and so do not secrete that which has been absorbed from the alimentary tract or peritoneum. Let Frey's finding be noted that these anesthetics do not interfere with the absorption of water from the gastro-intestinal tract. We are not surprised in the face of our explanation to note that Frey found this retention of water to occur just the same whether he had previously bled the animal or had cut the nerves to the kidneys, or changed the posture of the animal. Not even when he gave phloridzin or salicylic acid in an attempt to " stimulate " the kidneys did he get a urinary flow. According to our ideas of urinary secretion such results are entirely to be expected. None of these procedures affect the hydration capacity of the colloids of the tissues except as some increase it. The continuance of an absorption of water from the gastro- intestinal tract while none is being secreted through the kidneys is easily explained by the increased hydration capacity of the body colloids induced through the effects of the anesthetics. 1 Ernst Fret: Pfluger's Archiv, 120, 66 to 136 (3 papers) (1907). 322 (EDEMA AND NEPHRITIS 9. Transition from the Physiological to the Pathological in Kidney Function One does not pass from the physiological to the pathological in a jump, but insensibly. Thus, the " normal " absorption of water by a cell is ordinarily regarded as subject to well-defined variations, and yet as we approach the physiological extremes of high normal turgor we are likely to be halted by the information that we have already wandered into the regions of the patholog- ical and are face to face with an " oedema." The same holds true of all other functions. When in physiology we speak of a diminished urinary output we have really gotten into a region which others will call pathology. These paragraphs will show how the physiology and pathology of function as observed in the kidney fade inseparably into each other. We noted above how a series of most dissimilar disturbances in the circulation to the kidneys are in effect all the same in that they lead to a decreased output of water from the kidney. We have to say now what is the change wrought in the kidneys through the lack of oxygen which is produced in common by all of them. This is a question that we have argued many times before. Not only do we have an abnormal accumulation of carbonic acid in any organ when the blood flow through it is cut down, but through the interference with the oxygen supply to the part we expect an abnormal accumulation and production of various other acids in the affected tissues. Since the tissues contain various (hydrophilic) protein colloids we may expect these to swell if only a source of water is present. It is not surprising, therefore, that oncometric measurements have shown that every interference with the normal blood supply to the kidneys is followed by an enlargement of the organ, independently of any increase in size that may be due to mere filling of the vessels with blood} No further comment is necessary to show how these observations and our previously detailed experiments on the oedema of passively congested parenchymatous organs dovetail.^ The lack of oxygen induced in the kidneys through circulatory disturbances makes itself felt in the end in the oxidation chem- 1 See Gottlieb and Magnus: Arch. f. exp. Path. u. Pharm., 45, 223 (1901). The earher contradictory results of Starling are open to question. E. H. Starling: Jour, of Physiology, 24, 317 (1899). 2 See page 225. ABSORPTION, SECRETION— COMPLEX ORGANISM 323 istry of the kidney cells themselves. Now various chemical means are at our disposal by which we can interfere with the oxidations that occur normally in the kidney parenchyma without in any way altering the circulation of the kidney itself. We need mention only the effect of uranyl nitrate and various other metallic poisons, amyl nitrite, the cyanids and in lesser degree the various anesthetics, such as morphin, chloroform and ether. Every member of this list, which with varying degrees of ease is known to lead to albuminuria, hematuria, partial or complete suppression of urine, enlargement of the kidney and various " degenerative " changes in this organ considered char- acteristic of " nephritis," is known to interfere with the normal oxidations occurring in living tissues. The bearing that these remarks have on a large number of nephritides, encountered clinically, is apparent not only from the fact that almost every one of the poisons here mentioned has been known to lead to nephritis in man, but by the additional fact that the " toxic " nephritides that appear in the course of various acute infections also belong to this group. When the oxygen supply to a kidney is cut off sufficiently, or that which passes into it cannot be used properly, albumin, and at times blood, appears in the urine. Such an albuminuria and hematuria may be made to subside if the condition leading to the lack of oxygen is removed after not too long a time. When now, we add to these facts of changes in circulation, increase in the size of the kidneys, progressive diminution in urinary secretion to the point of absolute stoppage, albuminuria, and hematuria, the further fact that on section the kidney parenchyma appears swollen, grayish, and with kidney markings obscured, we have no difficulty in recognizing that we are dealing with a series of changes that characterize the ordinary acute parenchy- matous nephritis. To this question we return in detail later, but it is brought up here to emphasize the fact that the changes occurring in these experimentally induced nephritides are in part analyzable, and so help not only toward a theoretical under- standing of what is observed in clinical cases of acute nephritis, but in so doing give promise of being of practical worth. 324 CEDEMA AND NEPHRITIS 10. The Secretion of Dissolved Substances We shall enter into the problem of the secretion of dissolved substances only sufficiently to point out the illuminating touch given it by the physical chemistry of the colloids. Great pessim- ism still reigns regarding our ultimate ability to explain, on a purely physico-chemical basis, all the phenomena of secretion. That such a view is not justified must appear from even the brief remarks that follow. What has been most difficult to explain in secretion has been its selective character; in other words, the ability of the kidney, for example, to separate from the blood a liquid which has a totally different quantitative and qualitative composition. Qualitative differences are for the most part explainable through chemical changes that occur in the secretory cells themselves, whereby substances are produced (such as mucin for example) which do not appear in the blood at all. In other respects a secretion differs only in quantitative composition from the blood. This may go to the point of having almost entirely absent from a secretion certain constituents of the blood, as, for example, albumin from the urine. For the most part, however, the secretion contains some substances in higher, others in lower, concentration than the blood. To limit ourselves again to the urine, we need by way of illustration only recall that, under ordinary circumstances, the urine contains less chlorids than the blood, and more sul- phates and urea. How are such differences to be explained? To begin with, it is well to call to mind that a secretion of dissolved substances is possible only so long as water is furnished the living organism. A secretion of water is necessary before we can hope to have any secretion of dissolved substances. This is a physiological truth that is utilized daily by the intelligent physician when he orders the drinking of large amounts of water to aid the organism in ridding itself of any poison, as the toxin of an infectious disease, for example. How the secretion of water by the kidney may be made a continuous affair we have learned from our previous discussion. How it must make for a con- tinuous secretion of dissolved substances is apparent from what follows. Let us recall here our division of the urinary secretory system into its three parts: the blood, the secreting membrane, and the ABSORPTION, SECRETION— COMPLEX ORGANISM 325 urine, and our brief characterization of the first as a liquid colloid in which various crystalloids are dissolved, the second as a solid colloid also containing various crystalloids, and the third as a watery solution of various crystalloids (practically) free from colloids. Thus far our discussion has shown that under the conditions normally existing in the body no water can be intro- duced into the blood without getting the secretion of an equal amount as urine. And what is secreted as urine is water, and only secondarily do substances come to be dissolved in it, so that it assumes a chemical composition which permits it to be characterized as urine. Let us see now what must happen if some soluble ^ (or pseudo- soluble) substance is introduced into the blood. To simplify the problem and not make our discussion unnecessarily long, let us think of the blood as one homogeneous system, and the urinary membrane as another. Under such circumstances one of three possibilities presents itself from a physico-chemical standpoint. The dissolved substance may distribute itself uni- formly throughout the blood and the urinary membrane, or it may be present in either a greater or a less concentration in the urinary membrane than in the blood. Just what will happen is dependent upon the nature of the dissolved substance and the physical and chemical composition of the blood and the urinary membrane at the time. Of greatest importance are such facts as the presence and absence of lipoids, the character of the colloids concerned, and the state of these colloids as determined by the presence of acids, alkalies, salts, or various non-electrolytes. In other words, the laws of partition again come into play. These differences in the distribution of a dis- solved substance between the blood and the urinary membrane are rendered strikingly apparent when dyes are used as the dissolved substances. But this distribution of a dissolved substance between the blood and the urinary membrane represents in the end only a static affair, and the secretion of dissolved substances in the urine is a dynamic one. It requires no special comment to see now why only through the continuous secretion of water from the kidney can a continuous separation of dissolved substance from the urinary membrane (secretion) be rendered possible. The presence ^ The word soluble is used in these paragraphs in its broadest sense, so as to include even the pseudo-soluble (colloid) substances. 326 CEDEMA AND NEPHRITIS of water in Bowman's capsule and in the uriniferous tubules intro- duces the third phase into our secretory system and breaks down continuously the equilibrium that is trying to become established between the dissolved substances in the blood and the dissolved sub- stances in the urinary membrane. ■ ' ' The attempt to establish an equilibrium between the dis- solved substances in the urinary membrane and the dissolved substances in the urine (originally only water) as it passes down the uriniferous tubules makes for a diffusion of dissolved sub- stances out of the urinary membrane, and so all the time that water is being secreted by the kidney, tends to destroy the equi- librium, which is trying to become established between the dis- solved substances in the blood and the dissolved substances in the urinary membrane. When now we recall the physico- chemical fact that when any dissolved substance is offered simul- taneously a liquid colloid, a solid colloid, and water (as is the case in the kidney) , an unequal distribution of the dissolved substance between the three phases is the rule, then we will have no dif- ficulty in understanding why a difference in quantitative com- position between the blood, kidney tissue, and urine, so far as dissolved substances are concerned, is also the rule. Wherefore a "selective" secretion is to be expected rather than to be wondered at. Further than this we cannot pursue this subject at this time. In passing I would only like to point out that the fruits of colloid chemistry help us to understand even the most radical differences that exist between secretions and their source. None is perhaps more striking than the strongly acid reaction of the urine or the gastric juice against the practically neutral reaction of its source, the blood. But even these can be accounted for through the selective absorption by the colloids of the urinary membrane of the sodium-di-hydrogen phosphate, and by the colloids of the gastric mucosa of the hydrochloric acid of the blood. Such concentration of an acid by colloids from ve,ry dilute solutions of acid salts or acids has been proved directly by Goppelsroeder. Our considerations also indicate how, corresponding with differences in the colloid constitution of the different parts of the urinary tubule, these may show qualitative and quantitative differences in the way in which they secrete the various constit- uents of the blood. Physiologists have long beUeved that such differences in function exist. ABSOEPTION, SECRETION— COMPLEX ORGANISM 327 It makes no difference, of course, as to where we consider the water of the urine to be secreted. If this be the glomeruh, as generally held (but not as yet experimentally proved), then we can imagine the water to leach out the various urinary con- stituents from the secreting membrane as it passes down the uriniferous tubule on its way to the pelvis of the kidney. If water is secreted by several or all portions of the uriniferous tubule, the problem remains, from our point of view, essentially the same. Our theory also permits of the reabsorption of water, or of dissolved substances, or of both from the fluid passing down the uriniferous tubules as postulated by some observers. It cannot, of course, as yet be accepted that such a reabsorption does occur physiologically. That a reabsorption can occur is undoubtedly correct, but the experiments made to furnish evidence for such a belief unquestionably interfere with the normal function of the kidney. 11. Concluding Remarks on Absorption and Secretion. Lymph Formation. Vasomotor and Secretory Nerves The variable capacity of the body colloids for holding water also helps us to understand some of the phenomena of lymph formation. It seems to me that the formation of lymph is in many points entirely analogous to the secretion of urine and gov- erned by similar laws. The " secreting membrane " in this case is found in the cells and the intercellular substances that separate the blood capillaries from the lymph capillaries.^ It is, of course, clear that these cells and their intercellular substances con- stitute the bulk of the body tissues. Anything that makes these cells with their intercellular substances yield up water increases lymph flow. ' Let us first call attention to the fact that an increased arterial circulation to a part increases lymph flow. A classic experiment in this line is the observation that an increased lymph flow from the neck is obtained when the salivary glands are active (supplied with much arterial blood). Under such circumstances ' Recent histological and physiological studies indicate clearly that the lymph circulates through a series of closed tubes as does the blood. The old assumption that a direct communication exists between the two (through holes) i.s certainly not correct. 328 (EDEMA AND NEPHRITIS various tissues in the neck are rapidly freed of their carbonic (and other) acids. This decreases the capacity of their colloids for water, and so they give it up, in part to the blood, in part to the lymph. All salt solutions, when injected into the blood in sufficiently concentrated solutions, increase lymph flow. When sodium chlorid, sodium bromid, etc., are employed, these have to be injected in (osmotically) stronger solutions than when sodium sulphate, sodium phosphate, etc., are used. The same thing hap- pens in experimental diuresis. These experiments on the forma- tion of lymph are easily explained by saying that the salts diffuse into the tissues and make them give up their water which then passes in part into the blood, but in part a.gain, into the lymphatics. A similar explanation can be given of the " lymphogogue " action of various sugars. Physostigmin and pilocarpin increase lymph flow; atropin and morphin decrease it. In the doses ordinarily used, the former make in toto for an increased supply of oxygen and the more rapid removal of carbonic acid from the cells, the latter for a decreased one. While the former means a decrease in the capacity of the tissue colloids to hold water, the latter means an increase; these in turn mean a giving up of fluid to the lymph in the first case, and none available for such a purpose in the second. A word may not be amiss regarding the useful purpose served by the vaso-motor mechanism in this whole problem of absorp- tion and secretion. Changes of both a quantitative and a quahtative character must of course follow the changes conse- quent upon any variation in the caliber of the blood vessels sup- plying a part. With blood of a given composition, it is evident that with vaso-dilatation more will flow through a part, and so the opportunities for absorption or secretion whether of water or dissolved substances be increased. But such quan- titative changes in the blood flow through a part affect at the same time the chemical and physico-chemical character of the cells in that part, and so a series of quahtative changes in the character of the absorption or the secretion may be added to the quantitative ones already noted. It is these facts that we have to bear in mind when we attempt the analysis of the various phenomena that characterize absorption and secretion as observed, for example, in a mammal. ABSORPTION, SECRETION— COMPLEX ORGANISM 329 The organs that are predominantly secreting organs (kid- ney, sahvary glands, stomach, pancreas) are all supplied with large arteries, and when these glands are active, their arteries are dilated. The supply of highly arterilized blood which makes possible the secretion of gastric juice (as it makes possible the secretion of urine) makes it impossible at the same time for this organ to act as an absorptive organ. And experimentally we know the stomach to act indifferently well in this direction as far as water absorption is concerned. Other substances can, of course, be absorbed from the stomach (alcohol, salts) and be secreted into it (various salts) independently of any absorption of water. Failure to absorb water only means, of course, that the stomach wall, and the (arterial) blood coursing through it is saturated with water — the three phases of the system are in equilibrium so far as their water content is concerned. In so far as any dissolved substance is not distributed in such a way through the three systems as to be in equilibrium, it must move (be absorbed) into the stomach wall and the blood, or out of these (be secreted) into the gastric contents until the equilibrium is established. When the rich supply of arterial blood to a secreting organ fails, no secretion occurs, as can be seen particularly well in the kidneys, the salivary glands, etc., when their blood supply is cut down either through experimental constriction of the arteries supplying them, or when the vaso-constrictor nerves are stimu- lated. It is true that under certain conditions no secretion may occur from a gland even when an abundant arterial flow is fur- nished the secreting cells, but this is only possible if the normal chemistry of the cells constituting the secreting membrane is first disturbed, as after poisoning with atropin, which so inter- feres with the oxidation chemistry of the cells that they are put in a state of lack of oxygen in spite of all that is flowing by them. We can also understand the meaning of some of the morpho- logical changes observed in the cells of any secreting organ so situated as to have alternate periods of rest and activity. While the process differs somewhat in different cells, it may be stated in general that the cells become larger during rest, and smaller during activity. The interpretation of this simple fact as gen- erally given is very complicated. Need we say more than that they absorb water (become cedematous) when arterial blood is 330 (EDEMA AND NEPHRITIS scarce and they cannot get rid of their carbonic acid easily; and that they secrete water, that is, shrink, when the carbonic and other acids that are produced in cells when oxygen is scarce, are removed through a better arterial blood supply? With the sweUing of the cells during a period of rest there is an accumu- lation of granules in the cells. Most extravagant interpretations have been made of their physiological significance. Need they be anything more than protein (including mucin) precipitates occurring in the bodies of the cells because in the period of gland- ular rest the reaction of the cell protoplasm tends to move toward the acid side? When the granules disappear during glandular activity it simply means a reversal of the process — they go back into solution as the reaction moves back toward the neutral point or the alkaline side. The changes observed during rest and activity of the salivary glands, pancreas, etc., therefore become similar to the changes of " cloudy swelling," ^ observed in the liver or kidney in various pathological states (including interferences with the arterial blood supply to the cells making up these organs). Need we also to continue our belief in " secretory " nerves? I think not. We do not know a single secretory nerve effect in the complex organism which is not preceded by a vasomotor (vasodilatation) effect, and the increased secretion is easily explained through the increased oxygen supply furnished the gland by this means. The secretory nerves are, in other words, identical with the vasomotor nerves. There may be vasodilatation without secretion as when defectively oxygenated blood is fur- nished, or the gland cells themselves are rendered incapable of using the proffered oxygen, but there is no secretion without a large arterial blood supply which is furnished some glands con- stantly while it is furnished others temporarily through vaso- dilatation. After what has been said it is evident that no great differences exist between the essential nature of absorption and of secretion. Secretion is only the mirror of absorption. This truth seems simple enough, and yet it cannot be said that it has received any special attention from the workers in experimental medicine 1 For a discussion of the nature and cause of cloudy swelling, see page 455, or Martin H. Fischer: Kolloid-Zeitschr., 8, 159 (1911); ibid., 8, 201 (1911). ABSORPTION, SECRETION— COMPLEX ORGANISM 331 or physiology. And yet it ought to, for absorption and secretion in a complex animal bear a reciprocal relation to each other. It is because this fact has been ignored that much of our present- day confusion exists in this field. An adult organism in order to continue ahve has to maintain a certain constancy of physico-chemical composition. It follows that if it absorbs anything it must secrete this again within a reasonable time thereafter. It is in this " reasonable time " and the conditions that are at the bottom of the fact that this " reasonable time " has to intervene between the absorption and the secretion of any substance that makes us lose the con- nection between the two, even when we deal with the absorption and secretion of substances (water, certain salts) which are not chemically changed in the body. When these facts are borne in mind, the surprise expressed by some authors that atropin or morphin, which decrease various secretions, do not similarly decrease absorption from the gut or the peritoneal cavity dis- appears. Hardly! These substances favor the formation and accumulation of acids in the tissues of the body, wherefore, no secretion. We should, rather, discover an increased absorption of water after use of these drugs, which, in fact, we do. Other anesthetics act Uke morphin, and other drugs like atropin. When we use such agents in our experiments we have to remember what they do, and not ignore them when we come to interpret our findings. Operations, animal boards and physiological apparatus produce collectively effects similar to drugs, so these too must not be ignored. It is for this reason, as I stated above, that all these procedures must be reduced to a minimum if we would complete our analysis of just what constitutes the physi- ology and the pathology of absorption and secretion. The analysis of the problems of absorption and secretion could already be carried with entire safety beyond the limits outlined here and in my previous papers, which have had as their chief aim the mere establishment of the thesis that the colloids and their physical state determine both the quantitative and the qualitative character of the absorption and secretion of water and dissolved substances by protoplasm. This will be done elsewhere. In passing, however, attention must be called to the excellent service that will be rendered the further analysis of the problem by the theories of the colloid state which 332 CEDEMA AND NEPHRITIS are becoming progressively more clearly defined. Especially helpful to the biological worker must become the conclusions of Wolfgang Pauli ^ and his co-workers,^ more particularly Hans Handovsky ^ and Karl Schorr,* as well as those of R. C. ToLMAN,^ whose theoretical deductions regarding the colloid state seem broader and less capable of adverse critical attack than any yet proposed. The theoretical elucidation of the absorption and secretion of dissolved substances will necessitate adequate use of Wolf- gang Ostwald's ® work. Ostwald has shown that the math- ematical formulas of adsorption are applicable to the process of absorption (intoxication) as shown in certain fresh-water animals {Gammarus) when they are placed in solutions of various kinds. It is evident that these animals swimming about in a solution are no differently situated than a group of cells, say, in the mucous membrane of the intestine, which are bathed by such a solution. But Ostwald has developed the biological significance of what represents in a sense the mirror image of the adsorption formula, namely, the washing- out formula. This may be used to express mathematically the " toxic effect " of distilled water upon these animals — an effect brought about by the diffusion out into the distilled water of the salts contained in the animal. It is evident that the leaching out of dissolved substances from the kidney by the pure water origi- nally secreted from the organ constitutes the parallel of this " toxic effect " of the distilled water on Gammarus. Ostwald has further shown that the effect of a solution having but one salt dissolved in it is the composite of the adsorption effect of that salt plus the washing out effect of all the other salts contained in the animal but absent from the solution that is being experimentally employed. This phenomenon has its analogue in the experi- mental absorption of any pure solution from the intestinal tract 1 Wolfgang Pauli: KoUoid-Zeitsohrift, 7, 213 (1910). ^Wolfgang Pauli and Hans Handovsky: Biochem. Zeitsohr., 18, 340 (1909). ' Hans Handovsky: KoUoid-Zeitschrift, 1, 183 and 267 (1910), where references to earlier papers will be found. * Karl Schorr: (Cited by Pauli and Handovsky.) " R. C. Tolman: Jour. Am. Chem. Soo. 35, 307 (1913); 35, 317 (1913). 'Wolfgang Ostwald: Pfiuger's Arohiv, 120, 19 (1907); Kolloid-Zeit- schrift, 2, 108 and 138 (1907). Wolfgang Ostwald and A. Dernoschek KoUoid-Zeitschrift, 6 (1910). ABSORPTION, SECRETION— COMPLEX ORGANISM 333 of a mammal, for example, in which, as was noted above, there is a " secretion " of dissolved substances from the intestinal wall into the gut, while the dissolved substance originally intro- duced is being " absorbed." IV MAINTENANCE OF THE CIRCULATING FLUIDS IN THE BODY 1. Why the Blood Remains in the Blood Vessels We have up to this point been chiefly interested in the mechan- ism by which we manage to get a secretion from the blood, and by way of illustration have discussed with particular intensity the mechanism by which more or less water may be obtained from a kidney. We have seen how a kidney will secrete only as water is brought to it, and this in proportion to the amount of " free " water furnished. Let us now change our viewpoint, and instead of asking how we may get more urine from the blood, ask why all the blood is not poured out as urine (or some other secretion), in other words, why does the blood remain in the blood vessels? ^ This is biologically just as important, and medically and surgically just as practical a question as that of the ways and means by which a urinary secretion is maintained and increased or decreased. The maintenance of a normal circulation is absolutely neces- sary in the complex organism. It provides the individual cells with the materials necessary for their life, while at the same time it carries away the poisonous substances produced by them, which, if allowed to accumulate, threaten their existence. What is said in these paragraphs holds for both the blood and the lymph, but unless otherwise stated, we shall limit ourselves to the problem of the circulation of the blood. For the main- tenance of a circulation we need a properly working pump and a suitable circulating fluid. The question of what constitutes a suitable circulating fluid may be discussed from two points of view, from a chemical and 1 See the first edition of "CEdema" and James J. Hogan and Martin Fischer: KoUoidchem. Beihefte, 3, 385 (1912). 334 (EDEMA AND NEPHRITIS from a physical. The chemical side will be touched upon but incidentally. Our chief interest lies in certain physical aspects of the problem. Since our experimental studies were instigated by considera- tion of some problems in practical medicine, our discussion may begin with these. That part of the problem of maintaining a normal circulation which has to do with the existence of a suitable circulating fluid is usually taken pretty much for granted. We shall see later that this is dangerous. The coarser disturbances which may affect a circulating fluid like the blood are, however, familiar to every clinician and are striking enough. We need but call to mind the consequences of hemorrhage. If by accident or otherwise one of the larger vessels is opened in man or a labora- tory animal, we see following each other in the course of a very few minutes all those alarming symptoms which culminate in death. When now we try to say why this occurs it is quickly brought home to us that the most serious mischief done by the hemorrhage does not reside in a great loss of red blood corpuscles or in a loss of certain of the chemical constituents found in the blood, say the hemoglobin or certain salts, but in a diminution in the volume of the circulating blood. The proof for such a conclusion is easily brought, for to protect or save an animal from the effects of hemorrhage it is not necessary to transfuse blood; but trans- fusion with water containing various salts (so-called " physi- ological salt solution," " Ringer solution " or " Locke solu- tion ") may do. In fact, the dangers incident to transfusion of whole blood have made medical men in practice depend more and more upon salt solutions of various kinds and less upon the transfusion of blood itself. But even though salt solutions of various kinds work excel- lently, they do this only for a limited time. In other words, it is too often noted that while a physiological salt solution or a RiNGEE solution produces immediately briUiant results, this effect wears off in an hour or two so that the individual who has been roused from the threatening effects of a great hemorrhage begins to sink again, and even though we repeat our injection, the improvement in patient or animal is again only temporary. It is easy to see why this happens. The injected salt solution does not remain in the blood vessels. Proof of this is at hand. ABSORPTION, SECRETION— COMPLEX ORGANISM 335 Not only does the blood pressure attained after the injection gradually fall, hut the injected fluid leaves the body as urine or is taken up by the tissues (an oedema develops) or both. If the water could be retained in the blood vessels we should get more lasting results from the- injection of a salt solution. We are now in the heart of the problem and theme of these para- graphs. As seen in the experiments on the intravenous injection of blood and blood serum already described above, ^ the blood remains in the blood vessels (and the lymph in the lymph vessels) because the water is all held as hydration water combined with the colloids of the blood (and lymph) , and in this form cannot escape as a secretion. To meet the objection that the blood remains in the blood vessels because of some specific property and not simply because all its water is held in combination with the colloids of the blood, we insert Fig. 100 and Experiments 47 and 48 where water is again injected intravenously, but this time in combination with a colloid foreign to the blood, namely, gelatin. Experiment 47, which describes the intravenous injection of a pure 2 per cent gelatin solution, is inserted merely for purposes of control. Such a solution yields no rise in urinary output (see Curve B of Fig. 100), but objections may be raised to this experiment. As the proto- col shows, the animal de- ^r velops hemoglobinuria and albuminuria, casts appear in the urine and the urinary secretion is very low, in other words, it develops a "neph- Hours ritis." It could justly be Figure 100. charged, therefore, that such an animal secretes no water simply because it is nephritic. One of the causes of this nephritis resides in the acid properties of the gelatin used, but the whole picture cannot be thus explained. The pure gelatin solution produces all the signs and symptoms following injection of an equal amount of distilled water. The injection of this causes no rise in urinary output, but only be- cause the destructive action of the water on the blood with its consequent interference with the normal oxidation chemistry 1 See page 292. 336 (EDEMA AND NEPHRITIS of the body more than offsets the diuretic action of the water alone. There are therefore at least two factors responsible for the poisonous effects of a pure gelatin solution, its acid content and the retention in its hydration water of those properties of distilled water destructive to the blood. This conclusion is of importance not only from the standpoint of practical medicine, but also from that of the theory of the colloid state and the physiological character of hydration water. We can avoid the practical and theoretical objections to the experiment just described by adding a salt to the gelatin solu- tion as. in Experiment 48. Then there is no hemolysis, no hemo- globinuria, no albuminuria, no casts, and the urinary output is normal (see Curve a in Fig. 100). This experiment proves that water in combination with a colloid (gelatin) remains in the blood vessels.^ ExPEEiMBNT 47. Injection Fluid: 2 per cent gelatin solution. Belgian male rabbit. Weight 1448 grams. Kept on standard mixed diet. 72.4 cc. of the above solution, an amount estimated as equivalent to the total blood volume of the animal, are injected into an ear vein at the uniform rate of 10 co. every five minutes. No anesthetic. Time. Urine in cc. Remarlta. 2.15 Tied down, catheterized, injection begun. 7.0 ' Clear, yellow, no albumin. 2.30 0.5 Clear, yellow. 2.45 0.4 Slightly bloody, albumin. 2.55 Injection ended. 3.00 0.4 Somewhat bloody. 3. IS 0.4 Bloody, many yellowish casts. 3.30 3.45 0.2 Same. 4.00 One drop The color of port wine, many yellow casts. 4.15 0.5 Same. 4.30 1.4 Same. 4.45 O.S Same. Animal released. Total urine in two and one-half-hour period since beginning of injection, 4.6 co. 1 Roger and Garnier (Soc. de Biol., Mai 4 (1912); C. R. Soc. de Biol., Mai 5 (1912)) have also found an increased urinary output after intravenous injection of Locke's solution, but none when gelatin was added to this. They conclude that solutions " isoviscid" with the blood do not act diuretioally but attempt no explanation of the fact. ABSOEPTION, SECKETION— COMPLEX ORGANISM 337 Animal found dead in cage next morning. Autopsy : Peritoneal and pleural cavities contain a little bloody fluid. Heart is fiUed with blood. Kidneys, grayish and swollen. Experiment 48. Injection Fluid: 2 per cent gelatin in m/8 NaCl. White male rabbit. Weight 1249 grams. Kept on standard mixed diet. 62.5 cc. of the above solution, an amount equivalent to the total blood volume of the animal, are injected into an ear vein at the uniform rate of 10 cc. every five minutes. No anesthetic. Time. Urine in cc. Remarka. 10.45 Tied down, catheterized, injection begun. Cloudy, yellow, alkaline, no albumin. 0.2 11.00 0.2 Same. 11.15 1.5 Cloudy, yellow, alkaline, no albumin, no blood. Injection ended. 11.30 1.7 Same. 11.45 3.2 Same. 12.00 9.6 Same. 12.15 3.5 Same. 12.30 2.5 Same. 12.45 2.5 Clear, no albumin, no blood. 1.00 2.5 Same. Animal released in good condition. Total urine in the two and one-quarter-hour period since beginning of injection, 19 cc. If the explanation of the foregoing protocols is correct we should be able so to arrange our experiments that water will be retained in the blood vessels of an animal or secreted, depend- ing upon whether we introduce it in combination with a colloid or as " free " water. Such proof must, moreover, be adducible in one and the same animal. Fig. 101 shows the results when the injection of horse serum (water in combination with a colloid) is followed by one of salt solution ("free" water). Curves a, b, and c correspond to Experiments 49, 50, and 51, respectively. It is strikingly apparent that as long as we inject blood serum there is no increase in urinary output, whereas it rises enormously as soon as the salt solution is started. Experiment 49. Injection Fluids: Horse-blood serum followed by m/2 (2.9%) NaCl. White rabbit. Weight 1417 grams. Kept on standard mixed diet. 90 cc. of the serum, an amount equivalent to about 1^ times the total blood volume of the animal, are injected into an ear vein at the uni- 338 CEDEMA AND NEPHRITIS form rate of 10 cc. every five minutes, are then injected in the same way. 60 cc. m/2 NaCI solution Time. Urine in cc. Remarks. 10 00 0.3 Clear, dark amber, no albumin. 10.15 0.3 No albumin. 10.30 7.5 Neutral, faint trace of albumin. 10.45 S.O Clear, neutral, albumin, one cast. Injection of serum ended and injection of NaCl solu- tion begun. 11.00 18.0 Last part clear as water, faint trace of albumin, one cast. 11.15 78.0 Clear as water, neutral to litmus, faint trace of albumin. Injection of NaCl solution ended. Animal released in good condition; eats and drinks at once. Next morning the animal is alive and well, .i Total urine in the forty-five-minute period of the serum injection, 12.8 cc. In the next forty-five-minute period of the salt injection, 134.0 cc. Experiment 50. Injection Fluids: Horse-blood serum, followed by m/2 (2.9%) NaCl. Black male rabbit. Weight 1795 grams. Kept on standard mixed diet. 90 cc. of the serum, an amount equivalent to the total blood volume of the animal, are injected into an ear vein at the uniform, rate of 10 cc. every five minutes. 90 cc. of the NaCl solution are then injected in the same way. Time. Urine in cc. Remarks. 11.00 Tied down, catheterized, injection begun. Clear amber, alkaline, no albumin. 0.5 11.30 1.8 Same. 11.43 3.8 Clear amber, alkaline, faint trace of albumin, three casts. Injection of serum ended and injection of NaCl begun. 12.00 32.0 Last part clear as water, alkaline, faint trace of albumin, no casts, no red blood corpuscles. 12.15 67.5 Clear as water, neutral, no albumin. 12.30 75.0 Same. Injection of NaCl ended. 12.45 22.5 Same. 1.00 8.0 Same. Animal released in good condition. Next morning is alive and well. Total urine in the forty-five-minute period of the serum injection, 5.5 cc. In the forty-five-minute period of the salt injection, 174.5 cc. ABSORPTION, SECRETION— COMPLEX ORGANISM 339 Experiment 51. Injection Fluids: Horse-blood serum, followed by m/2 (2.9%) NaCl. Mixed Himalaya rabbit. Weight 1593 grams. Kept on standard mixed diet. 25 grams of blood are taken from the carotid artery, an amount equivalent to about one-third of the total blood volume of the animal. 96 cc. of serum are then injected at the uniform rate of 10 cc. every five minutes into an ear vein. This amount is about 1| times the blood volume of the animal. 60 cc. of the NaCl solution are then injected in the same way. Time. Urine in cc. Remarlo. 1 35 Tied down. Drawing of the 25 grama of blood from the carotid artery begun. Bleeding ended and injection of aerum begun. 36 cc. are injected in the first five minutes, then at the uniform rate of 10 cc. every five minutes. Clearing, trace of albumin. Slightly red, albumin, no casts, no red blood corpuscles. 1.55 2.10 2.25 2 30 0.3 2.3 3.0 2.40 2.55 3 00 7.5 61.0 Last part clear, faint trace of albumin. Clear, trace of albumin. 3.10 6.30 Same. Animal released; found dead in cage next morning. Total urine in the thirty-five-minute period of the serum injection, 5.3. In the following forty-minute period of the salt injection, 131.5 cc. In Experiments 49, 50, and 51 the injection of horse serum was followed by an albuminuria. This albuminuria is attribut- able to the large injections made, in consequence of which the circulatory system becomes overfilled. This leads to circulatory disturbances which affect the kidneys (and other organs) result- ing in the appearance of casts and albumin in the urine. The dyspnea caused by such large injections of serum is plain evidence of the general disturbance. Interestingly enough it can be made to disappear in a few minutes by injecting a strong salt solution, for this dehydrates the blood colloids and as the freed water escapes through the kidneys the volume of fluid in the blood vessels sinks nearer the normal.^ 1 C. H. Neilson: Jour. Am. Med. Assoc. (1913), has made a careful study of the effects of various salines on the high blood pressure shown by different types of patients. The salines decrease the blood pressure. It is an interest- ing fact that they do this in the order in which they dehydrate colloids. The salines dehydrate the liquid blood, of course, as they do the rest of the body and as the volume of circulating fluid in proportion to the capacity of the blood vessels for holding it falls, the blood pressure must sink. 340 CEDEMA AND NEPHRITIS TO- 50- 40 30 20 10 bia Hours 1 Figure 101. ro 60 50 40 30 20 10 Ls=z Hours 1 2 Figure 102. ABSORPTION, SECRETION— COMPLEX ORGANISM 341 Curves a and b of Fig. 102 are based on Experiments 25 and 53; they show results similar to those of Fig. 101, except that in these experi- ments the water was held in combination with gela- tin instead of the colloids of blood. In Experiment 52 we used a pure gelatin so- lution, in Experiment 53 a gelatin solution in a weak salt solution. What we said above regarding the rela- tive toxicity of such solu- ■ tions is true here, and the analogous effects on the ani- mal show plainly in the protocols. In Experiment 54 the water was held in combi- nation with the colloid ca- sein. The casein solution was prepared by neutraliz- ing the acid casein with a weak sodium hydroxid solu- tion and adding enough so- dium chlorid to counteract the poisonous effects incident to the use of the casein pure. The effects of a pure casein solution are the same as those of a pure gelatin so- lution. Fig. 103 based on this experiment needs no ex- planation. Figure 103. Experiment 52. Injection Fluids: 2 per cent gelatin solution fol- lowed by m/2 (2.9%) NaCl. Belgian male rabbit. Weight 1822 grams. Kept on standard mixed diet. 91 cc. of the gelatin solution, an amount equivalent to the total blood volume of the animal, are injected into an ear vein at the uniform rate of 10 cc. every five minutes. 91 cc. of the NaCl solution are then injected in the same way. 342 (EDEMA AND NEPHRITIS Time. Urine in cc. Remarks. 9 50 Tied down, catheterized. 2.8 Yellow, alkaline, no albumin. 10.15 0.4 Injection of gelatin solution begun. 10.30 0.6 Cloudy, yellow, alkaline, no albumin. 10.45 0.3 Same. 11.00 0.3 Same. Injection of gelatin solution ended, injection of NaCl solution begun. 11.15 5.0 Bloody, alkaline, albumin, filled with long, coarse-gramed casts. 11.30 76.0 Pale red, faintly alkaline, trace of albumin, no casts, (hemoglobinuria) . 11.45 6.9 Same. (Rapid breathing.) Injection of NaCl ended. 12.00 4.3 Hemoglobinuria. 12 15 Animal released, breathing short and rapid, head held high. 2 45 Autopsy. — In peritoneal anci pleural cavities some bloody fluid containing no red lolood corpuscles. Total urine in forty-five minute period of gelatin injection, 1.2 cc. In the forty-five minute period of the salt solution injection, 87.9 cc. Experiment 53. Injection Fluids: 2 per cent gelatin solution in m/8 NaCl, followed by m/2 NaCl solution. Yellow male rabbit. Weight 2083 grams. Kept on standard mixed diet. 104 cc. of the gelatin solution, an amount equivalent to the total blood volume of the animal, are injected into ear vein at uniform rate 10 cc. every five minutes. 104 cc. of the NaCl solution are then injected in the same way. Time. Urine in cc. Remarks. 11.00 2.3 Dark amber, alkaline, no albumin. 11.15 1.5 Same. 11.30 0.8 Dark amber, alkaline, faint trace of albumin. 11.45 0.4 Same. 11.50 12.00 3.0 Clearing, alkaline, faint trace of albumin, no casts, no blood. Injection of NaCl begun. 12.15 44.0 Clear as water, faintly alkaline, no albumin. 12.30 72.0 Same. Injection of NaCl ended. 1.00 56.5 Same. Animal released in good condition. Next morning alive and well. Total urine in forty-five minute period of gelatin injection, 2.9 cc. In the forty-five minute period of the salt solution injection, 172.5 cc. ABSORPTION, SECRETION— COMPLEX ORGANISM 343 ExPEBiMENT 54. Injection Fluids: Casein solution made by dissolv- ing 8 grams casein and 1.4 grams NaCl in 200 cc. n/30 NaOH followed by m/2 (2.9%) NaCl. Belgian male rabbit. Weight 2098 grams. Kept on standard mixed diet. No anesthetic. 104.9 grams of the casein solution are first injected into an ear vein at the uniform rate of 10 cc. every five minutes. This is followed by 104.9 cc. of the NaCl solution injected in the same way. Time. Urine in cc. Remarka. 2.15 Tied down, catheterized, injection begun. Amber, acid, no albumin. 0.4 2.30 2.4S 0,2 Acid, no albumin. 3.00 0.2 Amber, acid. On addition of acetic acid a white precipitate is tlirown down wliich disappears on the addition of more acetic acid. If this precipitate is filtered off, the filtrate gives no albumin reaction on the addition of concentrated nitric acid. The acetic acid precipitate disappears on heating, but on cooling is again formed. No casts, no red blood corpuscles. 3.05 Injection of casein solution ended. 3.15 One drop Injection of NaCl begun. 3.30 36.0 Last part clear as water, faintly acid. Albumin reaction as at 3 o'clock. 3.45 60.0 Clear as water, neutral. Albumin reaction as described. 4.00 35.0 Same. 4.15 4.4 Same. 4.30 0.2 Same. 5.00 Animal released and returned to cage. 5.30 Animal dies. Mere inspection of the animals during the injection of a colloid solution or a salt solution shows that in the first case the injected fluid is retained in the blood vessels, while in the second it is not. If we palpate the superficial blood vessels (veins and arteries of the skin and ear, the carotid and femoral arteries) we note that during the injection of a hydrophilic colloid solution they become gradually fuller and remain so. As the blood vessels become more distended the amplitude of the pulse lessens, and this effect persists. Even when the non- poisonous blood serum is injected the animal's breathing becomes disturbed before long, and if the injection is continued the animal dies from mere overdistention of its blood vessels with its result- ing mechanical disturbances in the circulation of the blood. The injection of a salt solution is not attended by such con- sequences. The distention of the blood vessels is not so marked and the effect is not lasting. There is no disturbance in breath- ing and if a dyspnea was previously induced by distention of 344 CEDEMA AND NEPHRITIS the blood vessels with a colloid solution it improves. Or if we have first brought on a (presumably) fatal hemorrhage and then saved the animal by an injection of serum, we may kill it subsequently by injecting a concentrated salt solution (which again diminishes the volume of the blood in the blood vessels, as in Experiment 51). Autopsy of the animal also shows that only the colloid solu- tion remains in the blood vessels. When an animal has been injected with salt solution (especially if concentrated) the amount of blood in the heart and large blood vessels is normal or even below normal. On cutting through organs like the liver or kidney they bleed but little, or are quite dry. But if we have injected a colloid solution, the heart and large vessels are filled with blood and on cutting through one of the parenchy- matous organs it wells out. Still other facts show that the colloid solution remains in the blood vessels. It simply cannot get out. Its escape from the blood vessels, which represent a closed system of tubes, is analo- gous to its disappearance from the serous cavities or the gastro- intestinal tract. Water bound to a colloid (blood, lymph, gela- tin solution, agar-agar, native albumin) cannot be absorbed as such by the peritoneum, or the mucosa of the gastro-intestinal tract, as previously emphasized, and the same holds for the blood in the blood vessels and the lymph in the lymph vessels. Finally — and to many physiologists and clinicians this will seem most convincing — direct measurements of blood pressure show that colloid solutions remaiii in the blood vessels while salt solutions do not. The rise in blood pressure after the injec- tion of a 0.9 per cent sodium chlorid solution into a non-anes- thetized rabbit is only temporary — in from five to thirty minutes the pressure sinks once more to its normal level. At the same time the urinary output rises. The same is true in patients. The intravenous injection of two or three liters of properly prepared salt solution does not change the blood pressure at all. If the same amount of water is injected in the form of blood serum, or a gelatin solution, the blood pressure rises from 10 to 25 mm. of mercury and remains there. At the same time there occurs no increase in urinary output. Such findings are especially marked in the abnormally low blood pressures fol- lowing hemorrhage. While salt solutions effect perhaps a tern- ABSORPTION, SECRETION— COMPLEX ORGANISM 345 porary rise in pressure (and for the time being a general improve- ment in the symptoms consequent upon the hemorrhage) a permanent rise results from an injection of blood serum. 2. On the Treatment of Shock The foregoing experiments were made in order to formulate more clearly the principles that must govern us in the treatment of those various pathological conditions that are characterized by an abnormally low blood pressure. A low blood pressure may have many causes; if it becomes especially low it is fatal to both man and beast. Why in a given condition the blood pressure is low is answered in very different ways by different authors. But on the question of therapy all authors agree that the maintenance of life becomes possible only if we succeed in raising the blood pressure and keeping it raised until the patient has overcome the condition that led to the low blood pressure. Such a low blood pressure may result from any one or any combination of the following causes : 1. A decrease in the force or number of the heart beats. 2. A diminution in the volume of the circulating blood. 3. An increase in the capacity of the blood vessels to hold fluid. Or to illustrate this in ordinary chnical ■ terms, a low blood pressure may result from a weakened heart muscle (myocarditis) ; from a hemorrhage, or a loss of the watery constituents of the circulating blood (oligemia); from a vasodilatation due either to a loss of tone in the blood vessel walls themselves or to impair- ment of the nervous (so-called vasomotor exhaustion) or chemical mechanisms (loss of the active principle of the suprarenal bodies from the blood) which are in part or wholly responsible for the maintenance of this tone. In these paragraphs we shall disregard the question of failure of the heart itself as responsible for a pathologically low blood pressure and discuss merely the principles involved in our ordi- nary therapeutic attempts at restoring a low blood pressure to normal by introducing fluids intravenously. The simplest problem of low pressure is presented by the ordinary cases of severe hemorrhage. We have learned why 346 (EDEMA AND NEPHRITIS injection of a " physiological " salt solution into such patients is so often disappointing. Only by injecting a suitable colloid solution can we expect to bring up the pressure and have it stay up, for only such remains in the blood vessels. What kind of a col- loid solution can we use and which is best fitted for the purpose? After what has been said it is self-evident that the best transfusion fluid is whole blood. But the obvious difficulties and dangers attendant upon the carrying out of a man to man blood transfusion limits its usefulness. The liquid of next choice for perfusion is that which most nearly approaches blood, namely, blood serum. Human blood serum is, however, difficult to obtain. The possibility of getting a human colloid solution that will stay in the blood vessels resides in the use of hydrocele fluid and of the ascitic accumulations from cases of heart disease. Such fluids can be drawn when opportunity offers into sterile containers and the serum separated from the fibrin clot by the methods employed in collecting horse serum. The use of the serum from human milk is also to be counted in here. Certain dangers are, of course, attendant upon the use of any of this human material, but upon these not overly much emphasis is to be laid, for perfusion is not at present employed in cases that are not desperately ill. It would be ideal if we could obtain a pure colloid solution for intravenous injection from other than human sources. But the number of available substances is very small and their use always connected with some danger. The only blood derivative seems to be horse serum. The danger incident to the intravenous injection of even large amounts of this appears small in com- parison with the certainty of death in cases where we are inclined to resort to such transfusion. During the past year and a half James J. Hogan ^ has made good use of properly prepared gelatin solutions in cases of hemorrhagic, surgical and toxemic shock. The first to aid him by trjdng gelatin solutions in surgical patients in whom death seemed the only prospect was B. F. Alden. Gelatin solutions intended for intravenous use must be pre- pared only from the purest gelatin. The ordinary gelatins are likely to contain much acid and the products of protein decom- position, which when injected intravenously are highly poisonous. 'James J. Hogan: Personal Communication. ABSORPTION, SECRETION— COMPLEX ORGANISM 347 All preparations of gelatin should be tested for these substances. In any case they should be thoroughly washed, and if conveniences for so doing are available, in running, steriUzed, distilled water. An amount of moist gelatin the equivalent of 25 grams of the dry material is then placed in 1000 cc. of freshly distilled water containing 10 grams of sodium chlorid and 2 grams of sodium carbonate crystals (Na2CO3-10H2O.) The whole is then auto- claved for an hour at 120° C. The gelatin solution must be prepared exactly as here described, otherwise trouble will be encountered from the fact that, as ordinarily done, the heating necessary to steriUze the gelatin properly decomposes it and so destroys the very properties for which it is used. The following cases taken from James J. Hogan's series may serve as illustrations of what has been said. Case I. — (Dr. B. F. Aldbn, San Francisco.) A. F., a 39-year old Italian laborer, had always been in good health. On March 3, 1913, he was brought from the country into the French Hospital, the victim of a severe accident which had occurred the day before. A tree limb had fallen across his right leg, crushing the upper end of his tibia, and the head and lower third of the shaft of the femur. All structures beneath the skin, with the exception of the ^emitendinosis tendon, had been completely severed. There existed also a compound frac- ture which, however, did not bleed, as all vascular communication had been severed above. The left tibia and fibula were fractured. Hemorrhage had been excessive, as practically no attempt had been made to control it. The patient was in a state of profound shock. He was almost exsan- guinated and showed a small, rapid, soft, radial pulse with cold extremities, excessive thirst, and shallow and rapid respiration. Blood was oozing from the lacerated wound about the knee. Strychnin, normal salt solution subcutaneously, gelatin subcutaneously, and the usual hemo- static measures were all used without improving the general state of the patient. Under one-half grain tropococain anesthesia, admin- istered intraspinally, the integument, which with a single tendon alone united the crushed leg to the patient, was severed. The clotted blood and crushed bone were removed and the large vessels in the stump rapidly ligated; the wound was dressed and the patient rushed to bed where his head was placed low and surface heat applied. The pulse could now no longer be obtained at the wrist, and as death seemed imminent it was felt that an intravenous injection of a gelatin solution was justified. 600 cc. were given. As there was no radial pulse at the beginning of the injection an accurate blood-pressure reading could not be made. The needle of a Tycos instrument oscillated at 90 mm. While the 348 CEDEMA AND NEPHRITIS injection was being made the radial pulse gradually returned and an hour later the blood pressure measured 125 mm. During the same time the pulse rate dropped from 130 to 88. The pressure was main- tained, rising slightly from day to day, until on the tenth day after the infusion it measured 145, while at the same time the pulse gradually dropped to 78. The pressure now gradually declined to 125 mm. and the pulse to 72. The patient made an uneventful recovery. Case II.— (De. D. N. Richaeds, San Francisco.) Mrs. E. S. R., entered St. Francis Hospital July 2, 1913, at 5.30 p.m. at full term and in labor. Regular pains contained until 4.15 a.m. when ether was given and she was delivered of a female child. The placenta was expelled at 5.10. This was followed by profuse bleeding, which continued until the patient was practically exsanguinated. At 6 a.m. the radial pulse could scarcely be felt. Her pulse was 140, her blood pressure (verified by Drs. Richaeds and Feasee) 58 mm. As her condi- tion was considered alarming, an intravenous injection of 750 cc. of gelatin solution was given. The blood pressure rose to 118 mm. within an hour, her pulse fell to 100, and her general condition improved greatly. For the ■ following four days the pressure varied between 105 and 110 mm., the pulse between 85 and 100. The patient made a good recovery, leaving the hospital July 17. Case III. — (De. B. F. Alden, San Francisco.) J. P. G., a 46-year old Frenchman, was operated upon by an associate on Februaiy 7, 1913, for a ruptured appendix. A suppurative peritonitis with two fecal fistulas, one into the cecum, the other into a loop of the ileum resulted. On April 15 the openings in the gut were repaired with CoN- nel's interrupted sutures of silk covered with continuous Lembeet catgut sutures. No excessive bleeding was noticed at the time of the opera- tion, but one of the Connel sutures must have lacerated a large vessel from which an active hemorrhage into the lumen of the gut resulted. On April 16 a lowered blood pressure with symptoms indicative of hemorrhage were observed. The outer wound was opened and the intestinal site explored, but as no serious bleeding was found the gut was not disturbed. The wound was repacked with gauze. On the morning of AprU 18 a large bowel movement of coagulated blood was observed. In the meantime the patient had declined, steadily. 10 cc. of horse serum were injected subcutaneously. A second blood stool containing a large amount of fresher blood occurred, when another 10 cc. of horse serum and 300 cc. of physiological salt solution were given. The patient continued to decline, exhibiting all the alarming symptoms of active hemorrhage. At 3.30 p.m. on April l8 the patient was reported dying. He was pulseless at the wrist. Oscillations in a Tycos instru- ment were observed at 80 mum. An immediate intravenous injection of 500 cc. of gelatin solution was ordered. Half an hour later the patient's pulse had dropped from 156 beats to 138 and had gained in vol- ume, strength and regularity. The blood pressure rose to 110 mm. and by the following morning it was 140 mm. This pressure persisted while ABSORPTION, SECBETION— COMPLEX ORGANISM 349 the pulse rate within three days gradually fell to 74. An uninterrupted convalescence followed and at the present time this patient is well. Case IV.— (Drs. W. B. Coffey and C. A. Walker, San Francisco.) Mr. B. entered St. Francis Hospital July 26, 1913, having been ill for four days. The patient showed marked evidences of general intoxica- tion. Examination revealed tenderness in the region of the appendix. An immediate operation was performed which revealed a ruptured appendix with marked, circumscribed peritonitis. The patient showed great shock after the operation. A 0.9 per cent sodium chlorid solution was dripped into the rectum throughout the day following the opera- tion. The patient did not rally. His pulse became progressively weaker and rose to 130, while his blood pressure fell, until at 9.30 p.m. it meas- ured 86 mm. As death seemed the only outcome, gelatin solution was given intravenously. This was started at 9.50, the blood pressure at this time measuring 85 mm., the pulse 128. The following records show the rise in blood pressure during the injection: After 150 cc. it was 96 mm. " 250 " " 110 " " 400 " " 128 " " 500 " " 132 " As this pressure was deemed sufficiently high the injection was stopped. The patient's general condition improved at once, his radial pulse filling out and the rate dropping to 90. The following morning his blood pressure was 119 mm., his pulse 90. That afternoon the pressure rose to 128 mm. while the pulse dropped to 85. The pressure remained while the pulse gradually returned to normal. The patient made an uninterrupted recovery. Case V. — (Dh. B. F. Alden, San Francisco.) M. C, a 35-year old native of France, because of an empyema of the left thorax following a lobar pneumonia, was transferred to the surgical service from the medical side May 20, 1913, for a thoracotomy. Before he was brought to the operating room the patient showed the weak, thready pulse, rapid respiration, extreme pallor, drawn face and cold and clammy extremities, ears and neck of profound shock. When placed on the operating table the patient seemed in extremis and was practicaUy pulseless. Because of this no proper blood pressure reading could be made. The anesthetist advised against the use of an anesthetic. As it was agreed that the patient would die, 500 cc. of gelatin solution were injected into the median basific vein. The incision was made without anesthetic as the patient was unconscious. Within ten minutes after beginning the per- fusion a perceptible radial pulse was noted and this gradually improved in quality. A section of the eighth rib was now removed and drainage of the pleural cavity effected. The patient recovered consciousness while on the operating table and his blood pressure rose to 118 mm. It contained to mount until it reached 154 mm. the following day. At 350 (EDEMA AND NEPHKITI8 the same time the pulse rate fell from 135 to 106. Two days later the pressure sank to 106 mm. while the pulse mounted to 136. As the drainage was found to be inefficient, by reason of adhesions in the pleura, the patient was reoperated under nitrous oxid and oxygen anesthesia. This was done without difficulty. The pressure then again rose to 126 mm. and remained there while the pulse fell back to 100. The patient's general condition gradually improved and after a prolonged convales- cence he left the hospital. It can, of course, be foreseen that what may actually be accomplished through gelatin transfusion in any case of low blood pressure depends upon the intensity, persistence and nature of the factors responsible for it. Viewed in this light, it is not remarkable that Hogan's experience thus far has seemed to indi- cate that cases of shock consequent upon simple hemorrhage are most easily relievable. Even those which showed extreme degrees of anemia rallied. Shock consequent upon injury or a surgical operation in the ordinary " clean " case seems to be controllable almost as easily. Least hopeful are the septic cases where the low blood pressure is in large measure dependent upon a heart suffering from the general intoxication. This is illustrated in Case VI : Case VI. — (Drs. W. B. Coffey and C. A. Walker, San Francisco.) [Mr. C, a 32-year old laborer, was brought into St. Francis Hospital at 1 P.M. July 25, 1913, with the diagnosis of a ruptured typhoid ulcer. He had had typhoid since July 1. It was felt that the patient could not live through an operation. The blood pressure was 65 mm. and as it was believed that the patient would certainly die, an intravenous infusion of gelatin solution was started at 2 p.m. After 300 cc. had been injected the pressure rose to 90 mm.; after 400 cc. to 96 mm.; after 500 cc. to 99 mm. At this point the abdomen was opened under local anesthesia and found filled with fluid, fecal in character. The opening in the small intestine was quickly located and closed. The gelatin perfusion was continued during the operation, the blood pressure rising steadily even during operation as follows: after 600 cc. had been injected it was 106 mm.; after 700 cc. it was 108 mm.; after 850 cc. it was 108 mm. This higher pressure continued until 6.10 p.m. when he showed signs of a rapidly falling pressure and died. In closing these paragraphs we would like to have it clearly understood that intravenous injections of proper colloid solutions are not at once to be accepted as panaceas for shock. In the clinical case every effort must be made to discover and meet the factors responsible for a weakened heart action, a lack of ABSORPTION, SECRETION— COMPLEX ORGANISM 351 tone in the blood vessels themselves (not as important as usually believed and then only terminally) and a diminution in the volume of the circulating blood. The monumental studies of Yandell Henderson have taught that the last named is of great importance and determined, in good part at least, by an abstraction of water from the blood by the colloids of the tissues, which in conditions leading to shock develop an increased avidity for it. We have learned how such avidity may under normal circumstances and in oedema be decreased with alkaU, salts and sugar, and so it is not surprising that this fact may be used to advantage both in the prophylaxis and in the treatment of shock. PART FIVE THE COLLOID-CHEMICAL THEORY OF WATER AB- SORPTION AND SOME PROBLEMS IN BIOLOGY, PHYSIOLOGY AND PATHOLOGY. PART FIVE THE COLLOID-CHEMICAL THEORY OF WATER AB- SORPTION AND SOME PROBLEMS IN BIOLOGY, PHYSIOLOGY AND PATHOLOGY. TURGOR, PLASMOLYSIS AND PLASMOPTYSIS In the earlier pages of this volume, when we were first placing our medically interesting problem of oedema, experiments were described which not only make this a problem of the cells, but it was pointed out that oedema really represents only one extreme of a series of phenomena common to all cells, vegetable as well as animal. To a brief consideration of these which are found grouped under the general headings of turgor, plasmolysis and plasmoptysis we shall now turn. By turgor the plant physiologists understand the normal rigidity of the plant cell as determined by a normal or physio- logical water content. When by any means the protoplasm of the cell is made to shrink away from the morphological (cellu- lose) cell wall, the cell is said to be plasmolyzed. When, on the other hand, the protoplasm is made to swell so that the cell wall is ruptured, plasmoptysis is said to have resulted. The animal physiologists have not used these terms in such a strict sense. In the use of the term turgor they agree with the plant physiolo- gists. The term plasmoptysis they do not generally employ at all, and under the heading of plasmolysis they not only consider all the more marked variations in the size of cells both in the way of a decrease or an increase, but also certain phenomena which 355 356 OEDEMA AND NEPHRITIS have become associated with such variations in size, as, for example, loss of coloring matter by the red blood corpuscles (hemolysis). These distinctions in terms must be borne in mind if confusion is to be avoided. To prevent ambiguity in the following paragraphs we will in each case first define our terms. The reason why the phenomena of turgor, plasmolysis and plasmoptysis are brought up in this volume is because discussion of their essential nature has not as yet been brought to a satis- factory conclusion. For this reason the following paragraphs which bring a unifying explanation for many of the apparently disconnected and contradictory experimental facts bearing on the problem are not out of order. Again will we find ample evidence of the important role played by the colloids and thus see an application made to problems considered essentially physiological of certain principles which we have previously discussed under headings considered characteristically patho- logical. II ON THE ABSORPTION OF WATER BY SPERMATOZOA, EPI- THELIAL CELLS AND WHITE BLOOD CORPUSCLES In the attempt to establish the vahdity of the laws of osmotic pressure for certain physiological and pathological manifestations of water absorption, biologists have been particularly eager to work with material which on experiment was found to approximate most closely the behavior demanded by theory. It is for this reason that certain plant cells and the red blood corpuscles have been the subject of more exhaustive study so far as their behavior toward water absorption is concerned than any other cells. The reason why just these cells should have approximated obedience to the laws of osmotic pressure more perfectly than most others that have been studied may appear later. But even these chosen cells show such great exceptions to the behavior demanded by theory that it is impossible to escape the experimentally well- grounded conclusion that most, if not all, cells do not follow the laws of osmotic pressure. The attempts that have been made to harmonize the observed behavior of various cells with that demanded on the theory that cells represent osmotic systems are ingenious, but we can scarcely believe sufficiently supported BIOLOGICAL APPLICATIONS 357 by experiment to be convincing. For the most part the explana- tions given are complicated, which constitutes in itself a threaten- ing feature when the explanation of any natural phenomenon is hazarded. What strikes one as particularly encouraging about the colloid idea of water absorption is its simpUcity, and the breadth of water absorption phenomena to which it may be appUed without apparent experimental or theoretical objection. In a preceding part of this book we tried to show how the absorption of water by the cells of muscle, the eye, the central nervous system, the kidney and the liver is essentially a function of their colloid state. What was said regarding these cells is also true regarding spermatozoa, white blood corpuscles and the epithelial cells of the bronchi, intestine, bladder and esophagus. We need not enter into the detailed experimental findings on this subject which may be found in H. J. Hamburger's ^ e-icellent work.. We again encounter no difficulty in explain- ing the experimentally observed facts when we call to mind the effect of acids, alkalies, salts, and these in mixture upon the swelling of (hydrophilic) protein colloids. All the cells mentioned swell if placed in distilled water. This fact, which has always been interpreted as due to differences in osmotic pressure, is really to be explained by remembering that under the conditions prevaihng in these experiments the cells produce acids which increase the capacity of their colloids for holding water. A second factor is found in the diffusion of at least some salts out of the cell, for the higher the concentration of the neutral salts in a colloid the less does it swell.^ ' H. J. Hamburgbe: Osmotischer Druck und lonenlehre, 3, 2 to 33; ibid., 52; 2, 400 to 432, Wiesbaden (1904). ^ The question of the antagonism between acids and neutral salts has given rise to meaningless priority claims." In biological material it was first discovered by H. J. HAMBnRGER (Arch. f. (Anat. u.) Physiol., 513 (1892); 153 (1893); Zeitschr. f. Biol., 28, 405 (1891); 35, 252, 280 (1897) ), when he noted that it required a higher concentration of salt to prevent swelling and lysis of various body cells when acid was present than when such was not the case. He also noted that sulphates were more powerful in this regard than chlorides. Hamburger explained his findings in the terms of osmotic pressure saying, in essence, that the' acid acted by increasing the intracel- lular osmotic pressure and that an increased salt concentration was therefore needed about the cell to counteract it. Other writers haye since Ham- btjrgbe's work claimed both the discovery of the fact and the explanation. So far as I know I was the first to observe the general antagonism between acids (and alkalies) and neutral salts on the hydration capacity of protein 358 (EDEMA AND NEPHRITIS The direct swelling effect of acids, including carbonic, is readily understood. Acids always bring about the greatest amount of swelling in colloids, and they are found to do this also in this biological material. The effect of alkahes is variable. Sufficiently dilute alkalies inhibit the swelling of spermatozoa in water (through the combined effect of neutralization of the acid formed in the spermatozoa and the production of salts) and of epithehal cells and white blood corpuscles suspended in water, salt solution, sugar solution or serum. The alkali neutrahzes the progressive production of acid in these cells, as this occurs under the conditions of the experiments (for example, separation from an adequate oxygen supply, as when the epithelial cells are scraped off a mucous membrane). In some concentrations and in some cells a greater swelling is produced by alkali than by any other chemical except an acid. The less evidence there is of a production of acids in a cell or a tissue used for such experi- ments as we are describing, the greater the power of alkalies to make them swell. This is because when much acid is produced the alkali is largely neutralized so that in the end we really observe the cells swelling in a low concentration of alkali with much salt (formed through neutralization); while when little, or no acid is produced the cells are swelling in alkali with a little salt (that normally found in the cell). All the cells mentioned in this paragraph swell less in any salt solution than in distilled water. With every increase in the concentration of the salt there comes a progressive decrease in the amount of the swelling. At a certain concentration the cells maintain for a variable length of time what is considered their " normal " volume. If the concentration is increased beyond this they shrink. In this brief description are exemplified all that is contained in the terms plasmoptysis, turgor and plasmolysis as understood by the plant physiologists. Impossilbe as it is to understand these phenomena on the basis of osmotic pressure, equally easy is it to see in them a perfect parallel of (hydrophilic) protein colloids swelling in a dilute acid or alkali in the presence of variable amounts of different salts. colloids (Am. Jour. Physiol., 20, 330 (1907) ), and to use this finding not only in the interpretation of biological observations like those of Hamburger but in the explanation of my own water absorption experiments. (Am. Jour. Physiol., 20, 330 (1907); Pfliiger's Arch., 124, 69 (1908)). BIOLOGICAL APPLICATIONS 359 The experimental observations on changes in cell volume upon which the just detailed conclusions are based were made by Hamburger in 1887, though they were not published until 1904, because the results did not fit in with the conception of the hving cell as an osmotic system which Hamburger, hke the plant physiologists, H. De Vribs and W. Pfeffer, before him, was most interested in seeing established experimentally. The r61e of the colloids in accounting for the exceptional behavior of these various cells was, however, considered by Hamburger. Unfortunately he believed the latter a mere adjunct ^ to the bio- logical importance of osmotic pressure, and not, as seems more correct, of primary importance — of such importance, in fact, that it not only relegates the r61e of osmotic pressure to a secondary place, but in most instances, if not all, questions its entire bio- logical significance so far as water absorption is concerned. In a much more positive way has Wolfgang Pauli^ declared the swelling of white blood corpuscles in dilute acids and alkalies to be analogous to the swelling of colloids under similar conditions. ni ON THE INTERPRETATION OF SOME EXPERIMENTS ON WATER ABSORPTION BY MUSCLE It is well to return at this point to a consideration of certain experiments carried out by Jacques Loeb and E. Overton on the absorption of water by muscle. While the experimental results of the two authors agree very well, their explanations of them are very different. As the views of neither have found general acceptance on account of the serious objections that can be raised against them, I would Hke to call attention to the har- monizing explanation that can be given of the observed facts on the basis of the colloid idea of water absorption as already discussed in a previous section ^ dealing with the absorption of water by muscle. 1 " Die an der wasseranziehenden Kraft des Zellinhalts wenig betheiligten CoUoidtheilchen," Hamburger, Osmotisoher Dmck und lonenlehre, 3, 4., Wiesbaden (1904). 2 Wolfgang Pauli: Ergebnisse der Physiologic, 6, 126 and 127 (1907;. 3 See page 109. 360 CEDEMA AND NEPHRITIS If a frog's muscle is dropped into distilled water it suffers a progressive increase in weight. This phenomenon is usually- interpreted as a response to immersion in a solution of too low an osmotic pressure, so that water is absorbed by the cell contents. I maintain that this is not correct, for were it, all our muscles ought to swell whenever we consume a quantity of fresh or dis- tilled water, and a frog living in a fresh-water pond ought to do Ukewise. But this does not occur. Clearly the muscle swells only because removed from the body. The difference between the muscle inside and outside of the body is this: Outside of the body the muscle develops acid, and in this and its effects upon the muscle colloids I would find the cause for the increased absorption, in distilled water. Added to this is the effect of the diffusion of salts out of the muscle, for the higher the concentration of salts in a (hydrophilic) pro- tein colloid the less does that colloid swell in a dilute acid. Quite contrary to the generally accepted belief, a loss of the osmotically active electrolytes of a tissue may, therefore, distinctly favor the absorption of w^ter. We will do well to consider this when- ever we try to define wherein lies the " poisonous " effect of distilled water. That the extirpated muscle develops acid must be borne in mind when we try to interpret the effects of acids, alkalies and salts upon it. To put a muscle into a' dilute acid instead of into distilled water is simply to add the effects of the external acid to that produced spontaneously by the muscle. The effect of putting a muscle into an alkali must depend .upon the con- centration of the acid formed spontaneously in the muscle and the concentration of the added alkali. Depending upon whether the latter partially, entirely or more than entirely neutralizes the acid formed in the muscle we get as a final result the muscle swelling in a dilute acid plus certain salts, in a neutral solution of certain salts, or in an alkaline solution plus certain salts. As the amount of acid formed in a muscle is quite variable, and as in consequence the possibility arises of many differently concen- trated mixtures of acid, salt and alkali, we have no difficulty in accounting for the large variation in results obtained when extirpated muscles are placed in dilute alkahes. Interesting are the effects obtained when muscles are placed in solutions of various electrolytes or non-electrolytes. Let it BIOLOGICAL APPLICATIONS 361 again be recalled that the extirpated muscle develops acid and that in consequence its colloids are really absorbing water in a medium containing acid as well as the salts or non-electrolytes. To consider first the electrolytes. Overton expresses surprise that while a 0.6 per cent sodium chlorid solution is in " osmotic " equilibrium with the red blood corpuscles of the frog, the muscle of the same frog demands a 0.7 per cent solution to keep it from swelling. The explanation is found in this: The muscle pro- duces acid rapidly (within minutes to hours), while the red blood corpuscles do so only very slowly (requiring several hours to days). To counteract the earlier accumulation of acid in the muscle requires more neutral salt. The sodium chlorid solution that is customarily spoken of as a " physiological," " isosmotic " or " isotonic " salt solution for use with frogs' muscle is, therefore, one that is sufficiently concentrated to just prevent the swelling of the muscle through the production of acid that takes place within it. When now the " isotonicit^'^ " of different salts is determined it does not surprise us to find that this is not identical with their " isosmoticity," for the physiological coefficient is not identical with the physical one. On the osmotic conception of water absorption physically "isosmotic " solutions ought to be physiologically " isotonic." Yet experimentally this is not found to be the case. On the colloid basis of water absorption this result, of course, does not surprise us, for physically isosmotic solutions of different salts are not equally effective in reducing the amount of water absorbed by a (protein) emulsion colloid swelling in the presence of a dilute acid. With every increase in the concentration of the salt solution we expect on the colloid basis a decrease in the amount the muscle swells. Experiment shows this to be the case. As we pass from the " hypotonic " solutions to those considered " isotonic " the muscle swells progressively less. If enough salt is added, the muscle not only does not swell, but shrinks to less than the volume of the freshly extirpated muscle. This marks the pro- gression from the " isotonic " solutions to the " hypertonic." To explain these facts on the osmotic basis, Overton assumes the individual muscle cells to be impermeable to the salt. In the colloid theory the cells may be freely permeable, which, as a matter of fact, we know physiologically they must ye, other- wise it would be impossible to affect the behavior of muscle as 362 CEDEMA AND NEPHRITIS markedly as we can experimentally through various electro- lytes. Let us now turn to the non-electrolytes. Overton concludes that the muscle cells are permeable to practically all these. This conclusion, drawn from the fact that a long series of chemical compounds permit muscle to swell just as though they were not present, is undoubtedly correct, though it is not explained by saying that an osmotic membrane exists about the muscle cells which excludes salts while it is permeable to these non-electrolytes. The extirpated muscle again absorbs water because it develops acid when taken out of the body, and non-electrolytes in con- trast to the electrolytes are, in the concentrations employed, practically without effect in antagonizing the action of the acid. These conclusions may be illustrated by citing two of Overton's experiments. (a) A sartorius muscle which has not changed in weight after some hours in a 0.7 per cent NaCl solution undergoes no change in weight if placed in a solution of 0.7 per cent NaCl contain- ing 5 per cent methyl alcohol, in spite of the fact that the osmotic pressure of this mixture is equal to a 5.2 per cent NaCl solution. Overton explains these facts by saying that in a solution of 0.7 per cent NaCl, the osmotic pressure within and without the cells is the same, and that while the osmotic, pressure of the second solution is vastly higher than that of the contents of the muscle cell, it cannot become effective and withdraw water from the cell, because the methyl alcohol enters almost instantly into the muscle fibers. The correct explanation to my mind is this: The sodium chlorid solution has a concentration just sufficient to counteract the effect of the acid formed in the excised muscle, and so maintains the colloids of the tissues in a condition in which their capacity for holding water suffers no great change in the hours devoted to the experiment. As the non-electrolji;es are practically without effect upon this, an addition of 5 per cent methyl alcohol to the pure sodium chlorid solution does not alter this absorption of water by the muscle. (6) A sartorius muscle which is placed in a solution of 0.5 per cent NaCl plus 3 per cent methyl alcohol — a solution which has approximately the osmotic pressure of a 3.6 per cent NaCl solution — gains in weight just as though it had been placed in a pure 0.5 per cent (a somewhat hjrpotonic) NaCl solution. If BIOLOGICAL APPLICATIONS 363 removed to a 0.7 per cent NaCl solution, the original weight is regained. Our explanation of these facts reads as follows: The muscle gains in the NaCl-methyl alcohol mixture because the concentra- tion of the NaCl is too low to keep the colloids of the tissues from swelling in consequence of the acid produced in the muscle after removal from the body, and so it absorbs water. The presence of the methyl alcohol is without effect because the non-electro- lytes are practically without effect on the swelling of colloids in the presence of an acid. When the muscle is removed to the 0.7 per cent NaCl solution a concentration is encountered which counteracts the effect of the acid more completely, and since the taking up and giving off of water by colloids represent in large measure reversible processes, the muscle gives up some of its absorbed water and so assumes its original weight. It is a simple matter, therefore, to account for the available experimental facts on the absorption of water by muscle on the colloid basis. Not only are the facts which it has been difficult to harmonize with the osmotic conception of water absorption explained in this way, but all the phenomena which we have been most willing to accept as osmotic may well represent only a fraction of that greater series of phenomena which we have designated as colloid. The question of whether the laws of osmotic pressure are at all applicable to the biochemistry of water absorption is therefore raised in the special case of muscle as it was previously raised in the case of spermatozoa, isolated epithelial cells, and white blood corpuscles. That the laws of osmotic pressure, even as rendered more generally appUcable to biological material through Overton's special assumptions, are incapable of accounting for all the ob- served biological phenomena, is admitted by this author himself, and in seeking an explanation of various aberrant phenomena he too considers the role of the tissue colloids. He refers, as did Pfeffbr before him, to the part played by the imbibition water of the cells (Quellungswasser), and at one point, correctly to my mind, declares the swelling of muscle in dilute acids to be identical with the swell- ing of fibrin in dilute acids. But upon this colloid absorption he does not lay much weight, as is evident even in his latest writings.^ ^ See, for example, his article in Nagel's Handbuch der Physiologie, 2, 2te Halfte, 744 to 896 (1907). 364 -CEDEMA AND NEPHRITIS It must be clearly understood that this questioning of the r6Ie of osmotic pressure in biological material so far as water absorp- tion is concerned does not question the importance in the general problem of the diffusion of dissolved substances. This is an entirely separate problem. The advantage of the colloid concep- tion of water absorption is that it permits of the diffusion of dissolved substances into regions where on the osmotic concep- tion they could never get. As already pointed out, neither do my views affect or tend to minimize in the slightest the great biological significance of the law of partition as worked out by Hans Meyer and E. Overton in their experimental studies on the cell lipoids. IV ON THE NATURE OF HEMOLYSIS ^ The important place which the teachings of colloid-chemistry find in the analysis of such a problem as that of hemolysis is indicated in these paragraphs. The most important ways now known by which hemolysis (an escape of hemoglobin from the red blood corpuscles) may be brought about may be summed up thus: (a) Through the addition of water to the blood, or through immersion of the red blood corpuscles in any salt solution hav- ing a concentration below a certain value (as ordinarily stated, below the osmotic concentration of the plasma). (6) Through the addition of alkalies. (c) Through the addition of acids. (d) Through the addition of urea and certain other simple chemicals such as alcohol, acetone, and most ammonium salts. Most ammonium salts allow hemolysis to occur even when present in concentrations at which other salts do not permit it. (e) Through putrefaction of the blood. (/) Through electricity, but only under circumstances which allow of the formation of acids and alkalies in the solutions con- taining the corpuscles. This paragraph, therefore, constitutes only a subheading of b and c. (g) Through heating the blood. 1 Martin H. Fischer: Kolloid-Zeitschr., 5, 146 (1909). BIOLOGICAL APPLICAIIONS 365 (h) Through the addition of complex chemical substances, such as saponin, sapotoxin, bile derivatives, and snake venom. With these we must class the specific thermolabile hemolysins. While hemolysis is easily produced by any of the methods outlined, the following difference is to be observed between the various methods. When a specific hemolysin, or a poison capable of acting at a very low concentration, is added to the blood, the hemoglobin escapes from the corpuscle, but the corpuscle undergoes no change in size. With few exceptions this is not the case in any of the other solutions — in all of them the red blood corpuscles increase in size when the concentration at which hemolysis occurs is reached. Especially marked is this increase in size in solutions of acids and alkalies in which he- molysis occurs very rapidly, and in which swelling is most pro- nounced. It is not strange, therefore, that a causal connection has been sought between this escape of hemoglobin and the swelling of the corpuscle. In nearly all of the illustrations given, the two processes go hand in hand — it is generally stated that as soon as the red blood corpuscle swells it gives up its hemoglobin. It is not surprising in consequence that to several observers it should have seemed as though the only thing necessary for a complete understanding of the physical (not biological) half of this problem of hemolysis was a physico-chemical concep- tion of the process of swelHng. Close study revealed that the salt solutions which just pre- vent the hemolysis of red blood' corpuscles all have very nearly the same osmotic pressure, and so we find the theory advanced that the red blood corpuscles are surrounded by a semipermeable film, and that they swell or shrink, give up their hemoglobin or do not do so, depending upon whether the surrounding solu- tion has a lower osmotic concentration than the corpuscular contents or the reverse. This conception is a purely mechanical one — as soon as the osmotic concentration of the fluid without the cell is below that of the cell contents, water passes into the cell, which in consequence swells. When this swelling has become sufficiently great, the corpuscle is rent asunder and the hemo- globin escapes. As the number of experimental observations on the behavior of the red blood corpuscles has increased, more and more facts 366 OEDEMA AND NEPHRITIS have come to light which show that the laws of osmotic pressure, so far as this sweUing is concerned, have only a most limited application to the problem of hemolysis, if any.^ It is needless to recite all the objections here. By way of illustration, it is enough to mention that isosmotic solutions of all salts and non-electrolytes do not at a certain concentration prevent hemolysis (ammonium salts, for example); that the amount of swelling of red blood corpuscles in isosmotic solutions of different salts and non-electrolytes is not the same; and that with the same salt the calculated decrease or increase in the volume of the corpuscle is not strictly proportional to the increase or decrease in the osmotic concentration of the surround- ing medium. Certain of these objections have been met, at least in part, through Overton and Meyer's studies on the lipoids. But even with the modifications introduced by these and other workers many of the phenomena observed, notably the action of acids or alkalies, cannot be at all explained on the osmotic basis. If we call to mind once more the effect of various external conditions on the swelling of protein colloids, a ready explanation is obtained of most of the phenomena observed in hemolysis so far as the changes in the size of the corpuscles is concerned. Red blood corpuscles (or more correctly put, their stromas) swell most in solutions of acids or alkalies. This is also true of protein. The presence of various salts diminishes the amount that red blood corpuscles swell. The same is true of protein. Doubling the osmotic concentration of the salt does not halve the volume of the protein — the volume remains greater than half the original. Red blood corpuscles behave similarly. When isosmotic solu- tions are compared, red blood corpuscles are found to swell more in some than in others. We found the same to be true of protein. All these analogies seem to me to indicate that the changes in the volume of the red blood corpuscles are dependent primarily upon changes in their colloids.^ Those external condi- ^ See for example the careful studies of Hans Koeppe, Pfliiger's Arch. (1900). 2 Wolfgang Pauli, Ergebnisse der Physiologie, 6, 127 (1907), seems first to have considered it possible that the absorption of water by colloids and the absorption of water by red blood corpuscles are analogous. He does not discuss the matter of loss of color. More recently Julius Kiss (Das period- ische System der Elemente und Giftwirkung, Vienna, 1909, only accessible as BIOLOGICAL APPLICATIONS 367 tions which increase the capacity of the colloids for holding water cause red blood corpuscles to swell, and those which do the reverse, cause them to shrink. It will be noticed that I have limited myself thus far to a discussion of the swelling and the shrinking of the red blood corpuscle, and have connected these processes in no synonymous way with the escape of hemoglobin from the stroma. This is because / consider changes in the volume of the red blood corpuscles and the loss of hemoglobin by the stroma separate processes, which while they may often be associated, have really nothing to do with each other. This conclusion is based upon the following facts: When we attempt to construct a red blood corpuscle mentally, these points are of interest: The red blood corpuscle is essentially a mixture of several colloids. Of first interest is the protein body which is ordinarily said to constitute the stroma and which, from the way it becomes gelatinous in agglutination experiments, has been described as " fibrin-like " in character. Every one of its physico-chemical reactions betray it to be a (hydrophilic) emulsion colloid. Mixed with this stroma are the two lipoids, lecithin and cholesterin. According to R. Hobeh, the former of these particularly shows some of the pronounced reactions of the (hydrophilic) emulsion colloids (as witness its so-called "myelin reaction")- These two fat-like bodies have, however, a property not possessed by the protein portion of the corpuscle — they are good solvents for ether, chloroform, alcohol, and the remaining lipoid-soluble substances. A fourth important con- stituent of the red blood corpuscle is the hemoglobin. This, too, is colloid, even though most of the hemoglobins can be obtained with varying ease in crystalline form. A class differ- ence, however, exists between hemoglobin and the other colloids enumerated as contained in the red blood corpuscle. Hemo- globin is not a hydrophilic, but a hydrophobic colloid; it is not an emulsion colloid (emulsoid) as the protein constituents of the red blood corpuscle, or lecithin, but a suspension colloid (suspen- soid). What now is the nature of the combination between the various emulsion colloids (the stroma with the lechthin, and cholesterin mixed in) and this colored suspension colloid? Hemoglobin a book review) comes to the same conclusion. Kiss, however, seems to con- sider swelling and loss of hemoglobin parallel processes. See above. 368 (EDEMA AND NEPHRITIS cannot simply be dissolved in the red blood corpuscle, for the amount present is entirely too high. Neither can it be held in the corpuscle because this is covered by a semipermeable mem- brane. For reasons touched upon in discussing the biological significance of the analogy between the swelling of (hydrophilic) emulsion colloids and of protoplasm such a conception of the living cell is impossible. Nor is this membrane notion tenable in the modified form according to which it is made up of lipoids. The hemoglobin is contained in no such oil-covered sac, to flow out whenever this is dissolved or punctured. The hemoglobin must be combined in some more or less fixed way with the rest of the cor- puscle. The lack of evidence to show that this combination between stroma and hemoglobin is a chemical one, and the fact that an enor- mous amount of hemoglobin is held by a very small amount of stroma ^ leads me to assume that the combination between the hemoglobin and the rest of the corpuscle represents an adsorption phenomenon? To test this hypothesis I constructed a system which simulated red blood corpuscles and tried the effect of different external con- ditions upon it. I used powdered fibrin stained deeply red with neutral carmine. This combination was chosen in order to obtain a (hydrophihc) emulsion colloid (fibrin) united with a (hydro- phobic) suspension colloid (carmine).^ As many, if not most, of the dyeing processes represent just such combinations between colloids, it would, of course, be an easy matter to choose other hydrophilic colloids, and other dyes, and depending upon their general and their specific properties, obtain results similar to and different from those which I obtained with my carmine-stained fibrin. I found, for example, that fibrin stained with hematoxylin behaves much like that stained with carmine. The fibrin is stained by placing it in a beaker and covering it with a carmine solution. It is interesting to see how the fibrin absorbs enormous amounts of the dye. One has to add fresh ' Red blood corpuscles contain from 35 to 45 per cent solids, of which from 80 to almost 95 per cent (in man) consists of hemoglobin. ^ See page 168. As generally held, this adsorption is a purely physical combination dependent upon the enormous surface presented by the adsorbing material. T. Brailsford Robertson has recently criticised this view and insisted that the combination might be chemical. It does not matter, so far as our argument is concerned, how this discussion is finally settled. ' My carmine solution was readily precipitated by salts, and was analyzable under the ultra-microscope. BIOLOGICAL APPLICATIONS 369 dye time after time to replace that which the fibrin has taken up from the supernatant liquid. The retention and loss of color by this carmine-stained fibrin is very similar to, and occurs under the same conditions as, the retention and loss of hemoglobin by the red blood corpuscle. This is readily apparent on comparing the following paragraphs with the similarly lettered ones given earlier in this section: (a) If the red-stained fibrin is placed in water, the water slowly becomes red. In a solution of sodium chlorid, or in the chlorids, bromids, acetates and sulphates of sodium, potassium or lithium, this loss of color does not occur until after two or more days, when the supernatant liquid may become faintly pink. (6 and c) If a little of any acid or alkali is added to the colored fibrin, whether suspended in distilled water or in a solution of sodium chlorid, the loss of color occurs promptly. While the colored and powdered fibrin when suspended in salt solution has an opaque appearance, the bright transparency of a blood that has been laked is suggested after the carmine has come out. Upon standing for a little while the fibrin flakes sink to the bottom of the test-tube, so that the clear, transparent, red solution collects above the swollen uncolored " shadows " of the fibrin particles. (d) Urea at any concentration brings about a prompt loss of color by the carmine-stained fibrin. Ethyl and methyl alcohol or glycerin act similarly, but not so powerfully. Solutions of ammonium salts also allow the stained fibrin to lose color in a way that the other salts do not.^ (e) I allowed some carmine-stained fibrin to putrefy in an uncovered dish. As the putrefaction progressed the supernatant liquid became more and- more red. (/) The effect of electricity was not studied. (g) Gently heating some carmine-stained fibrin brings about a prompt loss of color. (h) The effect of such substances as saponin, snake venom, etc., has not yet been studied. The way in which red blood corpuscles lose their hemoglobin is not unlike the manner in which carmine-stained fibrin loses 1 It is, of course, to be foreseen that were the carmine dissolved or adsorbed in a lipoid, the effect of the ethyl and methyl alcohols would be much more marked, and so imitate the phenomena observed in hemolysis yet more perfectly. 370 OEDEMA AND NEPHRITIS its red color in a dilute alkali. As is well known, red blood cor- puscles when subjected to a hemolytic agent do not lose their coloring matter suddenly, but progressively. When ordinary blood is mixed with water the hemoglobin ring above the sedi- mented corpuscles slowly grows a deeper and deeper red. The same occurs with colored fibrin. In this simple fact is found a serious argument against any of the generally accepted mechan- ical conceptions of hemolysis which only postulate ruptured membranes and the escape of the hemoglobin contained within these membranes. Were the idea correct the escape of hemo- globin would always have to occur more or less suddenly, while we know it to be, as a matter of fact, a progressive affair. Just as it has been found that an escape of hemoglobin and a change in the size of the red blood corpuscle (the stroma), while frequently associated, do not quantitatively parallel each other, so also can carmine-stained fibrin be made to lose or retain its red color entirely independently of the amount of change in the volume of the fibrin particles. The red fibrin swells enormously and promptly loses its color in a dilute alkali. The higher the concentration of the alkali the more rapidly and completely does the fibrin lose its color, yet so far as the swelling of the fibrin is concerned an optimal point is reached beyond which every increase in the concentration of the alkali only makes for a dimin- ished absorption of water. Again, if a little ammonium chlorid is first added to the alkali, the loss of color is (practically) unaf- fected, and yet the fibrin swells but little. In other words, the swelling of the colored fibrin follows the laws which Gertrude Moore and I have previously laid down; the loss of color entirely different ones. It seems to me that this analogy between the loss of hemo- globin by the red blood corpuscles and the loss of color by carmine- stained fibrin is more than accidental, and lends no mean support to the contention that the combination between hemoglobin and stroma is an adsorption phenomenon. If this view is accepted, then henceforth we will have to look for an interpretation of the phenomena of hemolysis into a different chapter of physical chemistry than that into which we have been accustomed to look. Instead of fibrin and carmine, as already pointed out, any other hydrophilic colloid united with any other of the ordinary (colloid) dyes might just as well have been chosen. The majority BIOLOGICAL APPLICATIONS 371 of these dyeing processes represent adsorption phenomena. We have also learned how these adsorptions may be increased, decreased or prevented altogether, as witness our use of the most varied mordants, precipitants, fixants and bleaches. Many of the methods thus employed (as the use of salts, acids, bases, formaldehyd, colloids of various kinds, heat, electricity, etc.) have a parallel in the ways and means by which the combination between hemoglobin and stroma may be increased or decreased. ^ The relationships between the different colloids in the case of the red blood corpuscles are, of course, much more comphcated than in the case of carmine-colored fibrin. In place of only two colloids, we have in the red blood corpuscle at least four, and this makes for an infinitely more complicated system. Not only may the adsorption characteristics of the individuals of a group of colloids toward any one other (hemoglobin in this case) be different, but they may mutually affect each other and so alter each other's adsorption characteristics. Lecithin and cholesterin, for example, have properties which allow them not only to share in, or modify the ordinary adsorption phenomena, as exhibited by the protein constituents of the red blood corpuscles, but because of their lipoid character they may not only absorb sub- stances which the rest of the corpuscle cannot take up, but they may be affected by means which do not affect the rest of the blood corpuscle. Just in so far as these hpoids affect the relationship of hemoglobin to the protein constituents, or of hemoglobin to themselves, any substance capable of affecting the lipoids (chloro- form, ether, acetone, etc.), must be able to influence the whole problem of the relation of the hemoglobin to the rest of the blood corpuscle, and so the problem of hemolysis. 1 Oscar Berqhausen showed me in Paul G. Woolley's laboratory in the University of Cincinnati, an excellent illustration of this. The hemolysis of human red blood corpuscles by carbonic acid can be markedly inhibited or prevented entirely by the addition of various salts. This explains why adrninistration of properly selected salts (calcium salts and alkaline salts) tends to inhibit or stop the periodic destructions of the red blood corpuscles in paroxysmal hemoglobinuria, while the presence of acids (as after exposure to cold, compression of the blood supply, etc.) precipitates such attacks. 372 (EDEMA AND NEPHRITIS V ON GROWTH AND SOME GROWTH PHENOMENA The turgor of plant and animal cells is generally recognized as of such fundamental importance in growth and some of its associated phenomena, that the following remarks, which are merely intended to show how important a role the colloids may play in the whole problem, are perhaps not out of order. Let us first consider the question of growth in general. As the term has been given various meanings by different authors during the past half century, it is well that we begin with a defi- nition. Least open to objection is that of T. H. Huxley, who speaks of growth as " increase in size." C. B. Davenport defines it more, precisely when he regards it as "increase in volume." Objections to all other definitions arise from the fact that in them are too often included those changes which are better considered under the caption " differentiation." These changes, while they may serve as a necessary introduction to, accompaniment of, or consequence of growth, have really nothing to do with the process itself. Driesch's distinction between a " passive " growth due simply to the taking up of water and an " active " growth due to assimilation is excellent, though, as Davenport has pointed out, the term "passive " is poorly chosen, for the taldng up of water is by no means a passive process, and that part of growth in which water is absorbed usually gives far more pal- pable external evidence of its existence than that included under Driesch's heading of " active " growth.^ An objection that we might raise against Hxixley and Daven- port's definition arises from the fact that not every increase in the volume of a cell, a tissue or an individual necessarily repre- sents what is ordinarily regarded as growth. The development of an tt'dema in the extremities of an individual, the temporary swelling of a muscle after exercise, the imbibition of water by certain cells of the sensitive plant when touched, would all, according to Htocley and Davenport, have to be regarded as " growth." In actual practice we would, of course, have little difficulty in distinguishing between true growth and the phe- iSee C. B. Davenport: Experimental Morphology, 281 to 375, New York (1908). BIOLOGICAL APPLICATIONS 373 nomena cited. What seems interesting to me is that in the end, when the physical analysis of these various physiological and pathological processes is complete, I believe we will find that what makes these processes overlap in definition makes them overlap in nature also. The question to which anyone discussing the general problem of growth (increase in volume) is most desirous of getting an answer is this: What is the source of the energy for growth? That the energy set free is at times exceedingly great is clearly enough indicated by the every-day evidence of the enormous pressures exerted by the growing tips of roots and stems, and the direct measurements that have been made of the pressures exerted by woods, pulps and seeds when soaked in water. The greatest osmotic pressures that may be conceived in cells (assuming them, for example, to contain saturated solutions of substances of very low molecular weight) cannot account for more than a fraction of the observed pressures. The pressures exerted by swelling colloids constitute an adequate source. We need only to say how under the conditions in nature these pressures are rendered effective. Let it first be called to mind that an absolute sine qua non for growth is the presence of water. How necessary is an ade- quate supply is evidenced by the farmer's worry about rain, and the laboratory experience of every worker in physiological botany. Secondly, all growth in volume is preceded by the pro- duction of various (hydrophilic or emulsion) colloids. But not only are various colloids produced, but conditions which par- ticularly favor the absorption of water by these colloids are also instituted. It is the rule, for example, that the growing tips of plants contain much acid. Many of them are sour to the taste and will even turn congo-red, blue. The role of acids in making various emulsion colloids swell is familiar to us from previous considerations. We have no difficulty now in under- standing the observation long familiar to the plant physiologists (to whom we are indebted for most of our knowledge of growth) that there exist in the tips of plants three well-defined regions of growth.^ At the extreme tip is found a region of rapid cell division .with comparatively slow growth. Here is occurring a deposition of colloid material. Below this is found a region iSee Davenport: Experimental Morphology, 283, New York (1908). 374 CEDEMA AND NEPHRITIS exhibiting great growth In this but httle increase in colloid material is noted, but the greatest absorption of water. Why such a process should be found to consume much less time than the synthesis of colloid material in the tip explains itself. The third region again shows little or no increase in volume, but abounds in the changes collectively termed " differentiation." In plant cells a part of this differentiation consists in the forma- tion of cellulose walls. As cellulose constitutes a colloid that is not affected by acids, bases, salts and various non-electrolytes at concentrations compatible with life, changes in the volume of adult cells (such as growth phenomena) are impossible. Only the colloid material within the cellulose membrane can be affected by these substances, in consequence of which it may shrink away from the cellulose membrane (plasmolysis) or swell to burst it (plasmoptysis). The colloid conception of water absorption also gives us the means of understanding the mechanism of certain growth curvatures and curvatures due to tropisms of various kinds as manifested in plants and animals. The remarks that follow apply particularly to plants, in which Sachs first worked out the general problem of the tropisms, though they are just as applicable to many animals whose tropisms Loeb has shown to be identical with those demonstrated by Sachs in plants. In consequence of the directive action of various external stimuli (Ught, heat, chemicals, electricity, water) the growing parts of plants bend and grow toward or away from the source of the stimulus (positive and negative tropisms). GroAvth curva- tures may also evidence themselves in consequence of differences in the intensity of the action of external stimuli. Various explana- tions have been given of how these curvatures are brought about. In most the effect of an increased growth, as evidenced particularly through the presence of an increased amount of water in the convex portion of the plant stem or root, or the animal organism, over that of the convex portion, best explains the observed phenom- enon. The question at issue now is how such an increased absorption of water by one side of a stem, for example, is brought about. Osmotic forces have been considered, but they are in- adequate from both a qualitative and a quantitative standpoint. The phenomenon is quite easily understood on the basis of the absorption of water by colloids. We know first of all that tropic BIOLOGICAL APPLICATIONS' 375 curvatures both in plants and in animals are confined almost entirely to the actively growing parts, and of these particularly to those regions in which various emulsion colloids are being produced most energetically (as in the tips of roots and stems where synthetic changes are most active). We know further, from the experimental studies of F. Czapek,i that under the " stimulus " of a tropism the chemistry of the stimulated proto- plasm becomes entirely different from that of the unstimulated. Between these chemical differences, the (hydrophiUc) emulsion colloids and an available source of water all the conditions are offered which lead to inequality in the swelling of the two sides of the vegetable or animal organism, with a consequent tiumng toward or away from the source of the stimulus. It also becomes intelligible why the older portions of a plant usually take no part in these tropic curvatures. The cells constituting them are surrounded by a (colloid) framework (such as cellulose) which is not affected by the slight chemical changes (low con- centrations of acids, alkalies and variations in the distribution of various salts) that are capable of affecting so markedly the gen- eral body of younger cells and the cell contents of the older ones.^ Through these more or less rigid cell walls both the expansion and the contraction of the adult cell is markedly hindered. Oiu- remarks show that in the last analysis various external stimuli produce their effects through chemical changes which they induce in the growing protoplasm. The effects of these external conditions come, therefore, to be referred to just such local chemical differences which we have long recognized as underlying the local irregularities in growth originating within the plant or animal itself. How important in this problem must be the production in different parts of the growing organ- ism of different colloids (albumin, glycogen, starch, cellulose, Upoids, with their qualitative and quantitative differences in their capacity for holding water) or, with the same colloid, the localized production of acids, alkalies and salts is readily apparent. We can also see why, barring specific chemical effects, the action of electrolytes on growth should be so much greater than that >F Czapek: Ber. d. deut. Bot. Gesellsch., 15, 516 (1898). ^ See, for example, the experiments of Louis Kahlenbbbg and Rodney Tetje: Botanical Gazette, 22, 81 (1896), on the effects of acids, alkaUes and salts on growth. 376 (EDEMA AND NEPHRITIS of non-electrolytes. Electrolytes affect colloids in comparatively low concentrations, while most of the non-electrolytes do not.^ These ideas on growth can be tested and many of its phe- nomena mimicked in the laboratory with a few colloids and various electrolytes and non-electrolytes. What must happen in these experiments can very naturally be foreseen, though the results are nevertheless interesting. With the use of cylinders, strips and leaves of gelatin various phenomena considered charac- teristic of the tropisms resulting from the action of chemicals, light, heat, etc., and certain irregularities in growth resulting from internal causes can easily be imitated in the laboratory. When such gelatin preparations are painted with a little acid on one side and are then dipped in water, beautiful negative curvatures are produced. If acidified gelatin is used and one side is painted with an alkali or a neutral salt, positive curvatures result. Or if a mixture of gelatin with egg albumin is employed a negative curvature results when a weak acid is employed, while a positive results if a stronger one (nitric acid) or a salt capable of coagulating the albumin is applied. When any dry acid (tartaric, oxalic) is stirred into gelatin in an irregular way and strips are then cut out of it and moistened with water, complicated curvatures, spirals and other irregularities in growth, such as characterize flowers, for example, can easily be obtained. In conclusion, let attention be called to the ready explanation which the colloid conception of water absorption offers of the ways and means by which certain plants and animals protect themselves from loss of water. Aside from certain gross advan- tages of external form, protective covering, etc., it is known that plants possess internal mechanisms by which they protect themselves from loss when water becomes scarce. It is such mechanisms that enable the plants of the deserts and the dunes to maintain their existence. Certain aquatic strains of animal and vegetable life are also possessed of such mechanisms, other- wise they could not withstand transference from fresh water to sea water, and vice versa. Through the work of van Ryssel- ^ These colloid-chemical views on growth have found valuable support and development in the work of K. Gedroiz (Russ. Joum. f. exp. Land- wirtsch., 11, 66 (1910)) and G. A. Borowikow (Biochem. Zeitschr., 48, 230 (1913); 50, 119 (1913)) which shows that the absorption of water by plants and their growth is governed by the laws controlling water absorption in simple colloids. BIOLOGICAL APPLICATIONS 377 BERGHE ^ we have learned that when water is scarce certain plants convert some of their starch into oxalic acid. Those types of plants which under natural conditions are most Hable to suffer from lack of water (the succulents) seem all to possess the interesting property of reducing their output of carbon dioxid, while producing at the same time various organic acids as soon as subjected to unfavorable conditions for growth. These phenomena of acid production have generally been interpreted as meaning that by such methods the plant increases the number of soluble molecules in its cell contents and so increases its osmotic pressure. A more correct explanation, it seems to me, is this — through the production of these acids the capacity of the plant colloids for holding water is increased, so that the agencies operat- ing to rob it of this water are counteracted. A question that awaits an answer in the case of animals is whether a like produc- tion of acids is responsible here also for the maintenance of a normal water content, as when a fish, for example, born in fresh water moves out to sea. The important help that the absorption of water by colloids can render the general problem of the ways and means by which the movement of sap can be accomplished and maintained, in trees for example, needs no specific comment. VI ON THE CONTRACTION OF CATGUT AND THE NATURE OF MUSCLE CONTRACTION 2 The problem of muscular contraction naturally divides itself into three parts — a study of the physical changes that characterize the muscular contraction, a study of the chemical changes that underlie this phenomenon, and lastly, that which is usually regarded as the most important part of the whole, the means by which the chemical energy regarded as the source of the muscular contraction is converted into the mechanical. The physical changes of muscular contraction have by all odds 1 Quoted by Hobee: Physikalische Chemie d. Zelle u. d. Gewebe, Zweite Auflage, 63, Leipzig (1906). 2 William H. Strietmann and Martin H. Fischer: Kolloid Zeitschrift, 10, 65 (1912). 378 OEDEMA AND NEPHRITIS received the greatest amount of study; next in order stand tlie chemical. Least agreement exists at the present time in the matter of how the chemical changes lead to the mechanical. It is more particularly toward the solution of this phase of the problem that these paragraphs are intended to contribute. What we have to say is best begun by detailing the results of a few experiments on the contraction of catgut. 1. Observations on the Contraction of Catgut (1) As is well known since the classical studies of T. W. Engelmann/ it is possible to make raw catgut undergo alternately a marked shortening and an elongation by changing the char- acter of the surroundings in which the catgut is placed. Engel- MANN found that catgut suspended in water shortened greatly when the water was heated, to lengthen once more when this was subsequently cooled. The following experiments show how such alternate contractions and relaxations may be brought about by other changes in the surroundings of the catgut. (2) We prepared the catgut used in our studies from the com- mercial raw catgut sold to surgeons, or from violin strings, the material employed by Engelmann. The catgut strings were soaked in distilled water before being used, after which they were split into as thin strands as possible. One such strand may be used for an experiment, but, in order to get greater con- tractile force, it is best to use several, as shown in Fig. 104. Here four strands of catgut of uniform diameter have been fas- tened to the glass rod. It is best to wrap the glass rod with silk thread in order to keep the strands from slipping. The four strands are then gathered together at the bottom with a silk thread and the whole is fitted into a muscle lever such as physiologists use, arranged to write on a recording drum. The whole is set up in such a way as to make it possible to bathe the catgut strands with any desired solution without in any way disturbing the apparatus, as indicated in Fig. 105. (3) When a single strand of catgut, or a set arranged as described, has been permitted to absorb as much water as it will by being kept in distilled water (or in a " physiological " salt 'T. W. Engblmann: Pfluger's Archiv, 7, 155 (1873). Ueber denUrspmng der Muskelkraft, Leipzig (1893). BIOLOGICAL APPLICATIONS 379 solution) no change occurs in the catgut, as evidenced by any movement of the lever, over long periods of time. The point of the lever writes a straight line. If we remove the beaker holding the distilled water and substitute for it another contain- ing a dilute acid of some kind (lactic or hydrochloric is best) it is noticed that after a slight latent period the strands begin Figure 104. Figure 105. to contract and the lever point writes a curve as shown in Fig. 106, I. After obtaining a maximal amount of shortening, a horizontal line is written as long as the catgut remains in the acid solution. If this is now taken away (indicated by the right- hand arrow in Fig. 106) and the distilled water is replaced, the lever point begins to fall, and slowly returns to its old base- line level. The height of the contraction is dependent in an interesting 380 (EDEMA AND NEPHRITIS way upon the strength of the acid solution. Curve I of Fig. 106 was obtained by passing from distilled water to n/40 hydrochloric acid and then back to distilled water. Curve II was obtained in an identical way with n/60 hydrochloric acid, and Curve III with n/80 acid. These curves indicate that the greater the concentration of the acid, the higher the contraction. To this there is, however, an upper limit, n/40 hydrochloric acid representing very nearly the optimal one for the contrac- Minutes Figure 108. tion of catgut strands. Curve IV, which is the lowest of the series in Fig. 106, was obtained with n/20 hydrochloric acid. (4) The state of the catgut is of importance in determining the height of the contraction. Freshly soaked catgut gives the highest contractions. If the catgut is allowed to remain in dis- tilled water for several days, the height of the contraction as obtained with a given concentration of acid becomes progressively less. This is shown in Fig. 107, where are recorded the con- traction curves obtained with the same strands of catgut used Minutes Figure 107. to produce Fig. 106, after they had been kept in water for five days. As the recording lever, the weight carried, etc., were the same in both series of experiments, the two sets of curves may be compared with each other directly. Curves I, II, and III in Fig. 107 were obtained with the same concentrations of hydro- chloric acid as those similarly marked in Fig. 106. -What is the nature of the changes in the catgut induced by prolonged immer- sion in distilled water we are not prepared to say, although we are most inclined to think them due to the digestion of the protein material, catalyzed in part no doubt by the ferments derived BIOLOGICAL APPLICATIONS 381 from the bacteria which get into the vessels holding the catgut in our ordinary laboratory experiments. (5) Fig. 108 shows the effect of the thickness of the catgut strand upon the contraction. In this experiment only single strands of the same length could of course be used. The con- centration of acid employed was n/40 hydrochloric in each case; c indicates the curve obtained with the thickest fiber, a that with the thinnest. It is readily apparent that the contraction results the more rapidly the thinner the fiber. Curve a is not as high as the other two. As the weight lifted was the same with all three fibers, such a result is easily accounted for on the basis of the relatively higher stretching force applied to the fiber in the case of a than in the case of the other two. Minutes ' Figure 108. (6) We noted above that catgut strands relax entirely when the acid solution in which they have contracted is replaced by distilled water. To get a complete relaxation, however, takes a long time. In other words, the fiber maintains a residuum of contraction. What this amounts to, after repeated transitions of the catgut from water to acid and back to water, is indicated in Fig. 109. On the first passing from water (base line) into n/40 hydrochloric acid the maximal contraction indicated by the first rise of the line is obtained. This is a record taken on a still drum. The first hori- zontal plateau records the maximal contraction, made by rotating the drum slightly forward. The fall in the record obtained on changing to water fails to reach the original base line. On Fig^Jm 109. again passing to acid the second rise is obtained, which it will be noted is higher than the first contraction. The second fall on immersion in water does not even reach the low point previously attained, and so on for a series as indicated in the figure. Then for 382 (EDEMA AND NEPHRITIS a long period the contractions and relaxations remain equal, and there are no appreciable changes in the levels of the maximal and minimal points attained. But if the series of changes from acid into water and back again were kept up long enough, it is clear that the maximal points attained would become progressively lower and the amount of residual contraction diminish, due to changes suffered by the catgut (digestion?). Justification for such a conclusion is found in the differences observable in the contrac- tion curves obtained from fresh and old catgut (Figs. 106 and 107). (7) In Fig. 110 is shown the effect of adding various amounts of a neutral salt upon the contraction of catgut in an acid solu- tion. Curve I shows the contraction of a series of strands when immersed in n/50 hydrochloric acid. The moment of immersion is indicated by the arrow a. At c the acid is replaced by water and relaxation results. Curve II was obtained by immersing the catgut at the point a, not in a pure acid, but in one containing in addition 0.,25 per cent sodium chlorid, at the point & in a pure 0.25 per cent sodium chlorid solution, the relaxation occurring as indicated. Curves III, IV, V and VI were obtained in identical fashion by alternate immersion of the catgut in n/40 normal hydrochloric acid containing respectively 0.5, 0.75, 1.0 and 3.0 per cent sodium chlorid, and then in 0.5, 0.75, 1.0 and 3.0 per cent pure sodium chlorid. In Curve VI it will be noted that the power of the acid in bringing about a contrac- tion has been suppressed entirely. As a matter of fact the fibers have contracted even less than in pure water (the curve lies slightly below the base line). (8) The series of curves in Fig. Ill show an extremely interest- ing contrast to those of Fig. 110. Curve I is again a contraction followed by a relaxation as obtained by alternate immersion of the catgut in n/50 hydrochloric acid and in pure water. Curves II, III, IV, V, and VI were obtained by immersion in acid of the same concentration, but the solutions contained in addition respectively 0.25, 0.5, 0.75, 1.0 and 5.0 per cent sodium chlorid. At points lying between the arrows b and c, these solutions were replaced by pure water. In every case further contrac- tion is obtained before the relaxation sets in, which occurred immediately in Fig. 110 when we passed to salt solutions instead of pure water. BIOLOGICAL APPLICATIONS 383 a >-. S 3 ^ > 384 (EDEMA AND NEPHRITIS Results identical with those portrayed in Figs. 110 and 111 are obtained if any other acid (such as lactic) is used in place of the hydrochloric acid, or any other salt (such as sodium lactate) takes the place of the sodium chlorid. Neither is it necessary that the salt and the acid used have a common ion. Any salt will depress the contractions obtainable in any acid provided they do not react chemically with each other to undergo double decomposition. %HC1 . Ringer- , H^O I I solution I Minutes Figure 112. (9) In Fig. 112 the facts already noted in Fig. Ill are brought out in a shghtly different way. In the first portion of the curve is noted the contraction obtained in a pure acid solution. Between the two points marked Ringer solution the catgut strands were immersed in this solution, after which water was substituted for it. As is readily apparent, one obtains under such circumstances a second contraction which is practically as high as that obtained initially. Vizo HCl V,„HC1 XoHCl x%oHCl '■ Minutes Figure 113. (10) In Fig. 113 is shown the effect of immersion in suc- cessively greater concentrations of the same acid. In passing from one to the other a greater and greater contraction is obtained until a maximal one is reached^in the highest (optimal) concen- tration of the acid. (11) Thus far the relaxations of the catgut after immersion in an acid have been obtained by using either water or a neutral salt. The relaxation occurs much more rapidly and the base line is regained sooner if for the neutral salt is substituted one that has the power of combining with the acid used to induce the BIOLOGICAL APPLICATIONS 385 contraction. The effect of this is shown in Fig. 114. In this case the contraction was obtained in n/2 hydrochloric acid, the relaxation in m/5 sodium bicarbonate solution. 2. Interpretation of Experimental Findings It requires no imagination to see in Fig. 114 a duplicate of the tracing obtained when ordinary striated muscle is made to contract. But before we discuss this further let us see with what general phenomena in colloid chemistry we may correlate the above described experimental results.^ Catgut chemically considered is a protein, and its general physico-chemical reactions betray its colloid character. The fact that it swells in water at once serves to class it with the lyophilic or emulsion colloids, or, as water is the absorbed sub- stance, with the hydrophilic col- loids. Merely superficial exam- ination suffices, therefore, to place catgut in a group with gelatin, ' Minutes ' fibrin, gluten and serum albumin. Figure 114. But it behaves like these pro- tein colloids in various other directions also. When we observe that on immersion in a dilute acid the catgut fiber contracts, we note at the same time that it does this by a process of swelling; it becomes thicker and shorter. The same thing is noted in the case of gelatin, fibrin, gluten or serum albumin. Gelatin, fibrin or gluten swells more in any dilute acid than in pure water, and the viscosity of serum albumin rises when acid is added to it. If the acid is washed out of these colloids, they resume their original form, as does the catgut when water replaces the acid solution. Just as gelatin, fibrin and gluten show within certain limits an increase in the amount of swelling with every increase in . the concentration of the acid surrounding them, so also do we note an increased height of contraction in catgut when the acid concentration surrounding this increases. The fact that catgut does not at once return to a previous state when the conditions about it are changed has its analogue in ' See page 43. 386 (EDEMA AND NEPHRITIS the way in which fibrin, gelatin, etc., only slowly recover from the effects of a previous surrounding, when a new one is substi- tuted for it (hysteresis). With a given concentration of acid, the amount of swelling attained by gelatin, fibrin, gluten or serum albumin is reduced by the presence of any salt (even neutral salts, and such having no ion in common with the acid), and this reduction in the swelling is the greater, the stronger the concentration of the added salt. The parallel of this is found in the reduction of the height of the contraction of catgut in any acid solution, when any salt is added, the reduction being the greater the higher the concentra- tion of the added salt. When fibrin has been allowed to swell to its maximum in a mixture of any acid with a salt, and is then placed in pure water, an initial increased swelling of the fibrin is noted, before the decrease sets in which brings the fibrin back to the degree of swelling characteristic of immersion in pure water. It is as though the acid were united more firmly to the protein colloid than are the salts. In spite of the greater diffusion velocity of acids over salts, the salts nevertheless seem to get out of colloid proteins more rapidly than do the acids, an observa- tion not without biological importance, nor without interest for the theory of the colloid state. This behavior of fibrin also has its analogue in the already observed characteristics in the contraction of catgut when changed from a salt-acid mixture to pure water. Point for point, therefore, the contraction and relaxation of catgut {the absorption and secretion of water by catgut) is identical with the taking up and giving off of water by various other colloid proteins. It is easily seen how these experiments on catgut contractions correlate themselves with the experiments of Engelmann. What happens is identical in both instances, namely, an absorption and secretion of water by the proteins composing the catgut, only while Engelmann used an increase in temperature to make the catgut swell, we used, for purposes that will become evident immediately, various acids. BIOLOGICAL APPLICATIONS 387 3. On the Analogy between the Described Contractions of Catgut and the Contraction of Striated Muscle It is easily seen how similar are many of the curves illustrating this paper with the curves obtained and familiar to every physi- ologist when striated muscle contracts. Fig. 114 could easily be mistaken for the record of an ordinary muscle twitch. In Fig. 113 we observe a series of successively higher contractions that remind us of the result when a series of inadequate stimuli are thrown at proper intervals into a striated muscle. Fig. 112 illustrates a phenomenon of rigor in catgut that Edward B. Meigs has described in muscle. If a frog's muscle is immersed in a weak acid it goes into a state of contin- ued contraction; if it is then placed in Ringeh solution it relaxes, to contract a second time if it is subsequently placed in water. The series of curves shown in Fig. 110 (and the first half of the curves of Fig. Ill) illustrate in catgut what is called fatigue in striated muscle. The last half of the curves of Fig. Ill show again the contractions obtained in moving from an acid solution contain- ing salt, into pure water as already referred to in Fig. 112. Fig. 109 illustrates the staircase phenomenon familiar from muscle physi- ology. Not only are the successive contractions of the catgut fiber progressively higher, but the fiber does not relax perfectly; there remains the residual contraction or increased tone familiar to us from muscle preparations. Fig. 108 shows how a catgut fibril may remain continuously contracted (tetanus). When we compare Figs. 106 and 107, and note how the same catgut fiber undergoes changes in its state which alter markedly its power to contract under given conditions, we recognize the analogue of the importance of the state of the muscle in our physiological experiments as determining the character of its contraction. The question now arises whether these physical analogies between the contraction curves written by muscle, and those written by catgut strings as described above constitute merely a happy coincidence, or whether the two processes are really in essence the same; in other words, is the contraction of striated muscle a simple problem in colloid chemistry just as we found the contraction of catgut to be? This is our belief. The aniso- tropic substance of the muscle corresponds to the catgut threads; 388 (EDEMA AND NEPHRITIS the isotropic substance or sarcoplasm to the water that surrounds the catgut threads. But does our analogy between the contraction process in catgut and in muscle extend beyond these physical likenesses; in other words, are the chemical surroundings that we used to make catgut contract, identical with those that make striated muscle contract? This, we think, is also the case. In fact, we sur- rounded our catgut with the very chemical conditions which our present-day physiology holds to exist in muscle in the various phases of its contraction. That a muscle produces acid during an ordinary contraction constitutes one of the classic facts of our physiology. This state- ment has only recently been generalized to the extent of saying that whenever a muscle is found to contract, evidence of acid production in the muscle exists. The contraction of rigor mortis is associated with the production of acid in the muscle, a fact which made L. Hermann, in calling attention to the analogies that exist between the contraction of muscle in rigor mortis and the ordinary muscle contraction, venture the suggestion that the ordinary single twitch was due to a temporary production of acid. More recently, particularly through the work of Fletcher, Hopkins and Meigs, it has been shown that in water rigor, heat rigor, chloroform rigor, etc., the contractions noted are also always associated with the production of acid in the muscle. But not only is the production of acid associated with every contraction of muscle, it is the cause of this, as has been shown particularly well by McDotjgall and Meigs. McDougall studied the effects of acids and various other conditions on the length of the isolated contractile elements of insect-wing muscle. He found these to shorten whenever he brought them in contact with any very dilute acid. With increasing concentration of the acid he found that an optimum was reached, beyond which a lessened contraction was observable. It will be recalled that we described the same phenomenon in catgut. When the muscle elements were removed from the acid solution to pure water, relaxation set in. The relaxation occurred more rapidly if instead of being placed in distilled water the muscle elements were placed in a sodium chlorid solution. The more highly concentrated this was the more rapidly did the relaxation set in. These facts also have their analogues in catgut. BIOLOGICAL APPLICATIONS 389 Meigs has greatly amplified the experimental observations of McDouGALL and described practically identical findings in frog's muscle. From these remarks it is clear that the chemical conditions which we described above as effective in producing and modi- fying the contraction of catgut are identical with those which do the same in striated muscle, wherefore we conclude that the phenomenon of contraction in muscle is entirely a problem in colloid chemistry. If this conclusion is justified, then let us review briefly some of our current theories of muscular contraction with an eye to discovering which of them are most nearly correct. In so doing we shall find that the proponents of these theories erred not so much through a failure to recognize that muscular contraction represented a colloid problem, but rather in that they did not consider this explanation adequate or capable of account- ing for more than a small part of the essential phenomena of contraction. 4. Historical and Critical Remarks For the first great step toward the formulation of a colloid theory of contraction we are indebted to Franz Hofmeistbr,^ that old master who has done so much to establish the biological importance of the colloid state. Hofmeister built upon the fact that protoplasm consists of a series of bodies which are capa- ble of imbibing water, and pointed out how in the processes under- lying the phenomena of imbibition a migration of water and the approximation of two points (contraction) that are sur- rounded by envelopes of water must occur whenever the imbibi- tion capacity of the one is increased at the cost of the other. In this way he tried to account for all the special types of proto- plasmic contraction as observed in different animal and plant forms. The missing element in Hofmeistbr's theory — which he him- self points out — is that he could not explain why the colloids suffered the changes which make for the contraction; in other words, the nature of the chemical changes that induced the phys- ^F. Hofmeister: Die Lehre von der Pflanzenzelle, Leipzig (1867). Not accessible in the original. 390 (EDEMA AND NEPHRITIS ical. For a first suggestion in this direction we are indebted to T. W. Engelmann.^ Engelmann started from the well-known fact that during muscular contraction the carbohydrates and fats disappear from the muscle, while carbon dioxid, water, etc., appear in their place. This chemical change is associated with the liberation of heat, and this fact Engelmann utilized to construct upon it his thermodynamic theory of muscular contraction. Briefly formu- lated, Engelmann believes that the muscular contraction is initiated by a chemical change in the carbohydrates (and fats) of the muscle which results in the liberation of heat; this heat acting upon the contractile elements contained in the muscle (the anisotropic substance) makes them absorb the isotropic substance and so swell and shorten. The physical half of this theory, it will be noted, is also an imbibition theory of the nature of Hofmeister's. To support this contention, Engelmann devised his now famous experiment in which he showed how cat- gut — which is also anisotropic — contracts in water when its tem- perature is raised, to relax again when the temperature falls. If the catgut strand is only momentarily heated, a contraction curve is obtained which is identical in appearance with a single muscle twitch. Engelmann's theory has been attacked on many sides, to our minds often with scant justice when the substituted theories are weighed in the balance against his. The best argument against it are furnished by two facts: First, the amount of heat produced during an ordinary muscular contraction is not sufficient to make anisotropic substance, of the nature of that found in muscle, shorten enough to explain a muscular contraction. Second, a contraction of muscle occurs under circumstances in which there may be no production of heat whatsoever. But even after all this is granted, the great fact remains that Engelmann was the first to create a satisfactory model of the muscular contraction out of materials which may be subjected to physico-chemical analysis, and so to remove the whole problem from a realm of speculation and terminology into one of reason and fact. As will be evident later, even in the matter of making a change in temperature responsible for the physical phenomena of contraction, he was ' T. W, Engelmann: See reference on page 378. BIOLOGICAL APPLICATIONS 391 not entirely wrong; he only failed to pick the most powerful explosive out of a series lying before him. The work of L. Hermann constitutes a valuable contribution to the establishment of a colloid theory of muscle contraction in several directions. Hermann emphasized very clearly the many analogies both from a chemical and a physical standpoint that exist between the ordinary muscular contraction and the various rigors. As " coagulation " is an obvious sign in the rigors, the question of whether the ordinary muscular contraction is a " temporary coagulation, or a kind of coagulation," has often been argued since Hermann's writings. Hermann took the signs of coagu- lation and the contraction of muscle in rigor to represent evi- dences of one and the same process, and believed both of them to be due to the formation of acid in the muscle which occurs in all the rigors. In such a belief he was in part right, in part wrong. In making the production of acid responsible for both he was right, but to understand properly what happened beyond this point was impossible then, for colloid chemistry had not as yet developed sufficiently. We know now that the obvious signs of any " coagulation " such as that which characterizes the rigors can only be asso- ciated with a loss of water by the " coagulated " colloid.^ As the muscular contraction consists of an absorption of water, just the reverse of " coagulation," it is clear that the " coagulation " and the contraction observed in muscle in rigor must be entirely separate processes. What happens in muscle is identical with the development of a clouding in the cornea of an eye simultaneously with the swelling of the enucleated eye when this is placed in acid- ulated water,2 or the development of a " cloudy swelling " in any of the parenchymatous organs when these are exposed to the same conditions.^ Two colloids at least are involved in the process, and while the one is behaving like gelatin, which swells in acidulated water, the other behaves like casein, which under the same circum- stances is precipitated. In rigor the anisotropic substance swells 'See Wolfgang Pauli: KoUoid-Zeitschrift, 7, 241 (1910). Pauli and Handovskt: Biochem. Zeitschrift, 18, 340 (1910). H. Handovsky: KoI- loid-Zeitschrift, 7, 183, 267 (1910); Fortschritte in der KoUoidchemie der Eiweisskorper, Dresden (1911). Karl Schorr: Cited by Patjli and Hand- ovskt. 2 See page 664; Martin H. Fischer: Pfluger's Arohiv, 127, 40 (1909). 'Seepage 455; Martin H. Fischer: KoUoid-Zeitschrift, 8, 159 (1911). 392 (EDEMA AND NEPHRITIS under the influence of the add and leads to the muscular contraction, while under the same circumstances another colloid is being precipi- tated (or, to use Hermann's word, " coagulated ") which gives the muscle an opaque appearance. As we shall see later, the loss of water by the colloid which is being " coagulated " no doubt yields that necessary for the swelling (contraction) of the other. Whether a rigor is reversible or not depends entirely upon whether the precipitation of the colloid involved is reversible or not; whether, in other words, removal of the condition which has made the colloid precipitate permits this to go back into solution. Depending upon the means employed to produce the rigor and the length of time it has acted, the colloid precipitations may or may not be reversible, and so the rigor. This matter of rigor can, in a sense, also be mimicked on cat- gut. If we allow a chromium salt to act upon the catgut along with any acid, then we get not only a shortening of the catgut, but a permanent one. While maintaining that of acid production is responsible for the permanent contraction in rigor, Hermann ^ made the further valuable suggestion that a temporary production of acid might account for the normal muscular contraction. But this remained a mere suggestion with Hermann. The idea that the production of acid is responsible for the muscular contraction either under normal circumstances or in rigor has been particularly clearly enunciated by William McDougall.^ This author holds the anisotropic substance (the sarcomeres or contractile elements of the muscle) to be built up of tubules " having delicate walls arid containing a fluid or viscid substance." The contraction he holds to be due to an absorption of fluid by these tubules " determined by the setting free of lactic acid in the fluid con- tents of the sarcomere, aided perhaps by an increase in the osmotic equivalent of these fluid contents through an increase in the number of molecules in solution. Then so long as the acid remains present in the fluid of the sarcomere, the additional fluid absorbed will be retained and the state of contraction will continue. But as soon as the acid escapes from the sarco- ' L. Hermann: Hermann's Handbuch der Physiologie, 1, 255 (1879). ^William McDougall: Journal of Anatomy and Physiology, 32, 187 (1898). BIOLOGICAL APPLICATIONS 393 mere the additional fluid will also escape with it into the sar- coplasm and allow relaxation to take place." With McDougall's description of the histology of striated muscle we are not immediately concerned; in passing we would only point out that much of the discussion as to whether a histo- logical structure is " solid " or " liquid " is purposeless, for animal and plant structures are chiefly colloid in composition, and the colloids that compose living matter combine in one the properties usually cited as characteristic of both the solid (maintenance of form) and the liquid state (surface tension, diffusion of dis- solved substances). It is clear that McDougall's ideas readily permit one to see why a single muscle twitch, a tetanus, or a rigor due to death, acid or water, all have the phenomenon of contraction in common. Underlying all of them is the production of acid in the muscle and depending upon whether this acid production is only tem- porary or permanent we have either a temporary or a continued state of contraction. McDouGALL worked with isolated muscle fibrils. If these are placed in a weak solution of any acid (acetic or lactic) they swell and shorten. If they are then placed in distilled water and the acid is washed out of them they relax again. When the acid exceeds a certain optimal concentration the shortening becomes less marked. If any salt is present in the dilute solution of the acid, the contraction is lessened, or may not appear at all. If fibrils that have undergone no marked contraction in a solution containing both acid and salt are transferred to pure water, they undergo a rapid shortening. We need not re-emphasize that these statements are point for point identical with those we made above on the contraction and relaxation of catgut under similar circum- stances. The theoretical views of McDougall have found excellent experimental support and have won precision through the careful studies of Edward B. Meigs.^ This author has not only collected the evidence which shows that an acid production underlies every phenomenon of contraction as observed in striated muscle, but he was the first to recognize and clearly express the fact that we deal in this problem (in part only, accord- 'E. B. Meigs: Zeitschr. f. allg. Physiologie, 8,81 (1908); Am. Jour. Physiol., 22, 477 (1908); ibid., 26, 191 (1910); Jour. Physiol., 39, 385 (1909). 394 (EDEMA AND NEPHRITIS ing to Meigs) with a colloid phenomenon, and that the acid owes its action to the fact that it makes certain colloids of the muscle swell. Since Meigs' writings McDougall ^ has also expressed this idea in unequivocal terms. With this colloid view of the muscular contraction we heart- ily concur. The criticisms we have to make of McDougall and Meigs' ideas are that on both theoretical and experimental grounds they do not consider the colloid conception entirely adequate. McDougall believes that in the process of con- traction osmotic effects play a part in addition to the colloid, while Meigs thinks, in his analysis of the nature of water absorp- tion by striated muscle, that osmotic phenomena are con- cerned here. No experiments are cited byMcDouGALL to sup- port his osmotic hypothesis, and as Meigs, who is the best champion of McDougall's ideas, agrees that the swelling of the contractile elements in muscle (the essence of contraction) is a colloid phenomenon, we may consider it settled that at least so far Meigs holds that osmotic phenomena do not play a r61e. In maintaining that the " living " muscle is surrounded by osmotic membranes, Meigs calls attention to the curve of water absorption exhibited by a muscle immersed in distilled water. Such a muscle rapidly attains a maximal swelling, then for a period loses in weight, gains in weight a second time, and then slowly loses again.^ The curve representing the second gain in weight comes at the same time and accompanies the contraction of the striated muscle, and this Meigs is willing to accept as a process of colloid swelling (swelling of the contractile elements under the influence of an acid). But the first swelling Meigs does not consider as of the same type. He here follows the older belief of E. Overton, that osmotic membranes exist about the " living " muscle cell.^ When excised and placed in distilled water, these osmotic membranes are destroyed, in part owing to the accumulation of acid within the muscle, in 1 William McDougall: Quarterly Jour. Exp. Physiol., 3, 53 (1910). ^ See page 113 of this volume; also Martin H. Fischer: Pfluger's Archiv, 124, 69 (1908). E. B. Meigs: American Journal of Physiology, 26, 191 (1910). ' The same erroneous view is held by R. Beutner on the basis of experi- ments carried out under Jacques Loeb's direction. Biochem. Zeitschr., 39, 280 (1912). BIOLOGICAL APPLICATIONS 395 part due to differences in osmotic concentration inside and outside the muscle cell which lead to their rupture. When the membranes are thus destroyed, the fluid behind them is allowed to escape, and so the muscle loses temporarily in weight. But this temporary loss in weight can be interpreted more simply as a phenomenon in colloid chemistry. The muscle contains several colloids, and the maximal swelling and pre- cipitation points for any given set of conditions are not the same for all these colloids. Under the influence of an acid, for example, the maximal swelling of the one may therefore be attained and exceeded sooner than that of another, and so a swelhng of one colloid in the muscle may have reached and gone beyond its maximum (an increase followed by a decrease in weight) before another has attained its maximum. As a matter of fact we know that just such a relationship must exist between the dif- ferent colloids in a striated muscle when this contracts nor- mally. There is no free water in the body; it is all held in com- bination with the colloids of the tissues.^ If one colloid element in an organism swells (say the anisotropic substance), it can do this only as it first robs some other element of its content of water. It would be eminently useful, therefore, if the condi- tions which on the one hand make for a swelling of the aniso- tropic substance, on the other make for the shrinkage (giving up of water) of another (isotropic substance). To our mind all that characterizes the phenomena of water absorption and of contraction, or the loss of water and of relaxation in muscle, together with the various phenomena of " coagulation " observed in the rigors, represent but simple expressions of the effect of various acids and salts on that mixture of the several protein colloids which make up the muscle. We propose shortly to deal further with this subject. Here we would only direct attention once more to Fig. 112 and the apparently complicated series of reactions that may be obtained from a simple catgut fibril when exposed to the action of water, acids and salts. It is reactions of this type in muscle cells that have given rise to the highly complicated beliefs regarding the existence of membranes, etc., in and about them. As a matter of fact, we have no more reason for postulating their existence in muscle than in our catgut. 'See page 252; also Martin H. Fischer: KoUoidchem. Beihefte, 2, 304 (1911). 396 (EDEMA AND NEPHRITIS Much of the confusion that exists to-day in this whole prob- lem of contraction, water absorption, irritability, etc., as observed in muscle, arises from the fact that various authors have too carelessly passed from observations made on one to conclusions regarding another, instead of studying each phenomenon sepa- rately. Association of phenomena does not make them identical. Just as we learned that the signs of coagulation observed in rigor mortis are not identical with the phenomena of contraction observed in the same condition, so also does water absorption not parallel loss of irritability, or loss of irritability mean a loss of the power of contraction. When the original report of these observations on catgut was in press, the colloid-chemical theory of muscular contraction upon which they bear received valuable and independent support through the work of von FIjrth and Lenk ^ on rigor mortis. In a care- ful and convincing study of this problem these authors show that the postmortem contraction of muscle is a colloid process and influenced by various external conditions in the same way and in the same direction as the swelling of proteins. Somewhat later Wolfgang Pauli ^ discussed the general problem of muscular contraction and in bringing fresh support for the colloid theory of contraction further illuminated the subject by a critical dis- cussion of the chemical changes which induce the colloid ones. Independently of these authors, J. Grober'' has shown how the rate of swelling in simple colloids approximates the rate of the muscular contraction and quite recently Rudolf Arnold* has studied the water absorption and the contraction of different kinds of human muscle in a way which brings new and corrob- orative evidence of their essential colloid-chemical character. ' VON FtjRTH and Lenk: Biochem. Zeitsohr., 33, 341 (1911). ^Wolfgang Pauli: KoUoidchemie der Muskelkontraktion, Dresden (1912). ' J. Grober: Munch, med. Woohensohr., 2433 (1912). ' Rudolf Arnold: Kolloidchem. Beihefte, 5, 411 (1914). PART SIX NEPHRITIS PART SIX NEPHRITIS THE THESIS As apparent to the most casual student, our ideas regard- ing the nature and cause of nephritis are to-day in a state of chaos. The reasons for this are not far to seek. While physi- ology, pathology and clinical medicine have all contributed toward the analysis of the problem, little or no effort has been made by the various workers in each of these fields to find com- mon ground with those in another. Such effort needs to be made, for, as pointed out above, we step in no abrupt manner from the physiology of the kidney into its pathology, or from laboratory findings into the practical problems of everyday medicine. We need in these pages to get a definition of nephritis which is sufiiciently broad, and so we shall use this much-abused term in its ordinary clinical sense. It becomes therefore a convenient heading under which to consider those clinical pictures which are characterized by the appearance of casts and albumin in the urine; by certain morphological changes in the kidney; by a change in the amount of water put out by it; by changes in the absolute and relative amounts of dissolved substances given off in the urine, and by such associated phenomena as oedema, increased blood pressure and cardiac hypertrophy. How these all fit together will develop later. I need not be reminded that the term nephritis with its implied meaning of an inflammation of the kidney is a misnomer, 399 400 CEDEMA AND NEPHRITIS because in the "non-purulent inflammations of the kidney" which constitute the accepted cf the nephritides, the ordinary patho- logical evidences of inflammation are largely missing. Termin- ological discussions do not change nor yet analyze the well- recognized pathological conditions with which we are dealing. That which we as clinicians have come to regard as a clinical entity, and call nephritis, represents in reality the aggregate of a number of changes each of which must be treated separately if we would come to a satisfactory understanding of what is included under the clinical term. At least silently, the necessity for such a division of the subject has, as a matter of fact, long been recognized, for have we not largely given up the discussion of nephritis and taken up more and more that of albuminuria, anuria, oedema, chlorid retention — all of them parts of nephritis? Yet the persistence of the term nephritis in spite of our daily efforts in medicine to become scientifically more precise seems to be not without reason; it is the one term by which we are enabled to express the fact that the albuminuria, the anuria, etc., nearly always appear as associated phenomena. But from such a constancy in association we are enabled to draw an impor- tant conclusion — they must all have a common cause. The fol- lowing pages attempt to show what this is. To render our argument clear, we will at once state our gen- eral conclusion: All the changes that characterize nephritis are colloid-chemical in nature and due to a common cause — the abnormal production or accumulation of acid and of substances which in their action upon colloids behave like acid, in the cells of the kidney. To the action of these upon the colloid structures that make up the kidney are due the albuminuria, the specific morphological changes noted in the kidneys, the associated production of casts, the quantitative variations in the amount of urine secreted, the quantitative varia- tions in the amounts of dissolved substances secreted, as well as the other signs of nephritis which appear in direct connection with the kidney. The alleged consequences of kidney disease such as osdema, high blood pressure, uremia, etc., are not consequences, but accom- panying signs and symptoms which demand separate discussion and analysis. We shall now take up the proofs for these contentions in order. It is convenient to consider first the chemical factors NEPHRITIS 401 which bring about the colloid changes, and of these we shall lay main stress on the abnormal production or accumulation of acid in the kidney. Not only does this seem to be the most important, but our remarks concerning it may serve as an out- line by which the value of any other factor in the problem may subsequently be tested. If our thesis is correct we must be able to show that: 1. There is evidence of an abnormal production or accumula- tion of acid in the kidney, or of conditions predisposing thereto in every case of nephritis; and conversely that: 2. Any means which leads to an increased production or favors the accumulation of acid in the kidney results in nephritis. II AN ABNORMAL PRODUCTION OR ACCUMULATION OF ACID IN THE KIDNEY OCCURS IN EVERY CASE OF NEPHRITIS §1 If nephritis results whenever the acid content of the kidney is sufficiently increased then evidently the maintenance of its normal state must be intimately associated with maintenance of neutrality in it. We are therefore, first of all, interested in the fact that (exclusive of the gastric juice, the urine, and less pos- itively, the sweat, vaginal secretion, and alimentary contents when fat is fed) the fluids and tissues composing the normal mammal are to all intents and purposes neutral in reaction and are capable of maintaining this neutrality against the introduction of con- siderable acid into them. In the terms of our modern physical chemistry, an acid reaction is due to the presence of free hydrogen ions, an alkaline reaction to the presence of free hydroxyl ions. A neutral reac- tion means, therefore, one of two things: either neither of these ions are present, or else just as many of the one as of the other, so that they balance each other (the latter is the case in the body). The neutral reaction of the blood (and from this it has been gen- erally assumed that the tissues themselves are also neutral in reaction) seems at first sight a rather surprising fact when con- sidered in the hght of our older teachings that the body fluids 402 (EDEMA AND NEPHRITIS and the cells are all " alkaline." But these older teachings are erroneous, for they are based upon an improper interpreta- tion of the results obtained with titration methods. The blood, for example, has generally been held to be " alkaline," because it is capable of neutralizing acid. But as we know now, the power of a solution to neutralize an acid is no index of its content of free, that is to say, active, hydroxyl ions which is the true measure of its alkalinity. For a proper measure of this hydroxyl ion concentration in the blood (or in the tissues) all the determinations are valueless that antedate the fundamental measurements of P. Fraenkel,^ G. Farkas,^ and Rudolf HoBER^ who first used proper physico-chemical methods (so- called gas chains) in the biological study of this problem. The observations of all these authors agree in pronouncing the normal blood neutral in reaction; as neutral as pure distilled water. Of further interest is the fact that this state of neutrality of the blood {and of the tissues) is maintained against the introduc- tion of considerable acid or alkali into them. When exposed to the action of an acid, it is found that the normal hydroxyl ion concentration of the blood drops with the progressive introduc- tion of acid into it in the form of a curve, which falls only very slowly at first, and then more rapidly. Just why and how the state of neutrality is thus maintained does not at this particular moment interest us, but it may not be amiss to point out that two factors are involved in the process. The first lies in the fact that such salts as sodium carbonate and disodium hydrogen phosphate are capable of uniting with acids (carbonic and phosphoric acids) to form salts having a higher hydrogen con- tent (sodium bicarbonate and sodium dihydrogen phosphate), but which in their dissociation yield few more hydrogen ions than the salts from which they were originally formed and which were present in the blood to start with. In other words, there is only a slight increase in the concentration of the hydrogen ions (increase in hydrogen ion acidity) in spite of the considerable introduction of acid into the system. (L. J. Henderson.) * ' P. Fraenkel: Pfliiger's Archiv, 96, 601 (1903). 'G. Faekas: Pfluger's Archiv, 98, 551 (1903); Archiv f. (Anat. und) Physiol., Supplement, 617 (1903). 'Rudolf Hobbr: Pfluger's Archiv, 81, 522 (1900); 99, 572 (1903). *L. J. Henderson: American Journal of Physiology, 15, 257 (1906); 21, 169 (1908); 21, 427 (1908); Ergebnisse d. Physiologic, 8, 257 (1909), where extensive references to the literature will be found. NEPHRITIS 403 The other and perhaps lesser element for the maintenance of neutrality resides in the amphoteric character (that is to say their power of combining either with acids or alkalies) of the col- loids found in the blood and tissues. The albumins, for example, can unite with considerable quantities of acid (or alkali) with- out any decided change in their behavior toward indicators. The presence of certain colloids in any system will therefore serve to delay the increase in the concentration of the hydro- gen ions when an acid is added to this system.^ But let not the impression be gained from these remarks that blood or the tis- sues are not sensitive to even very minute additions of add {or alkali) to them. Such an increase in the concentration of the car- bonic acid as occurs when normal arterial blood becomes venous is already sufficient to reduce the hydroxyl ion concentration in the latter to one-half that existing in normal arterial blood. How profoundly even such a change affects the state of the colloids we will have occasion to discuss later. For the present we are con- tent with making the point that the blood and (presumably) the tissues are neutral in reaction and that they are capable of maintaining this neutrality within rather wide limits, even when subjected to the action of an acid. § 2 As modern physico-chemical studies have reduced what we formerly regarded as the " alkalinity " of the blood to a point where we may call it neutral, so also have they reduced the normal " acidity " of the urine from what we used to assume this to be. Just as the neutralizing power of the blood for acids is no indication of its true reaction, so also is the amount of alkali with which a given specimen of urine will combine no measure of its true, that is to say, active, acidity. To gage this properly the concentration of the hydrogen ions in it must be determined and this was not done until L. von Rhorer ^ and Rudolf Hober ^ 1 J. Sjoquist: Skand. Arch. f. Physiol., 5, 277 (1895). Otto Cohnhbim: Zeitschr. f. Biol., 33, 489 (1896). K. Spieo and W. Pemsel: Zeitschr. f. Physiol. Chem., 26, 233 (1898). S. Bugarszky and L. Liebermann: Pflii- ger's Arch., 72, 51 (1898). T. B. Robektson: Jour, of Physical Chem., 11, 542 (1907); 12, 473 (1908). 2L. VON Rhorer: Pfluger's Archiv, 86, 586 (1901). 'Rudolf Hober: Hofmeister's Beitrage, 3, 525 (1903). 404 (EDEMA AND NEPHRITIS applied the principle of the gas chain to the physico-chemical analysis of the urine. Table LXXVI, taken from Hobee/ indicates what is the concentration of the hydrogen ions in a series of normal morning urines. TABLE LXXVI Normal Urine Hydrogen ion acidity (10 -'-Ch). Titration acidity. 0.58 0.62 0.50 0.46 0.31 0.046 0.034 0.042 0.069 0.075 It would support our idea of the cause of nephritis if it could be shown that this acidity of the urine increases in conditions associated with the urinary findings characteristic of such kidney disease. How strikingly true this is, is clearly evident from the analyses of nephritic urines given in Table LXXVII also taken from HoBBR and made, of course, without thought of using them for such purposes as we do here. TABLE LXXVII Abnormal Urine Hydrogen ion acidity C10-«-Ch). Titration acidity. Remarks. 2.34 1.50 0.84 1.10 2.20 2.10 0.56 0.67 0.0191 0.018 1 0.027 ( O.O20J O.O22I 020 1 0.014 1 0.050/ Interstitial nephritis. Acute nephritis. Chronic interstitial nephritis. As is seen on comparing the two tables, the active acidity of the urine of nephritics may be more than four times that of the normal urine. But Table LXXVII already suffices to betray another fact. The highest acidities occur in the acute forms of nephritis, in other words, in the same forms in which we find most albumin, the largest number of casts, the greatest decrease in the 1 Rudolf Hober: Physikalische Chemie d. Zelle u. d. Gewebe, Zweite Aufl., 158, Leipzig (1906). NEPHRITIS 405 urinary output, etc. The lowest values are found in the chronic interstitial forms, in other words, in the very types in which albumin is found in smallest amounts, or at times not at all. The degree of the albuminuria and the other evidences of kidney disease therefore tend to follow the degree of acidity. We shall have occasion to return to this question later. Let us now look at the columns in these tables that record the titration acidities. It is such determinations that we find recorded in large number in clinical studies of nephritis. When the individual titration acidities in the above tables are com- pared with their corresponding hydrogen ion acidities, it is readily apparent that the two values do not even approximately parallel each other. What is learned when the titration acidity of the urine is determined, is its capacity to neutralize alkali. Under otherwise constant conditions it is clear that this titration acidity of the urine must grow with every increase in the amount of acid in the urine. The uniformly higher titration acidity of the urine in nephritis, as shown not only in tables LXXVI and LXXVII, but in the scores that may be found in any of the larger monographs on nephritis, becomes further evidence, there- fore, in favor of our contention that an abnormal production or an abnormal accumulation of acid occurs in the kidney when thus affected. §3 In the same way that we use the increased capacity of the urine for neutralizing alkali as evidence for the presence of abnormally large amounts of acid in it (and so in the kidney cells from which this comes), so also may we use the decreased capacity of the blood for taking up acid as evidence in the same direction. The titration values of the blood, which the earlier clinical observers looked upon as indices of its " alkalinity," may be drawn upon for evidence to show that in the nephritides there exists a decreased power of the blood to neutralize acids. As studied particularly by Rudolf von Jaksch,i W. H. Rumpf,^ E. Peiper 3 and F. Kraus * a decrease in the acid capacity of 1 R. VON Jaksch: Zeitschr. f. klin. Medioin, 13, 350 (1887). 2W. H. Rumpp: Centralbl. f. klin. Medioin, 12, 441 (1881). 'E. Peiper: Virohow's Arch., 116, 337 (1889). < F. Kraus: Zeitschr. f. Heilkunde, 10, 106 (1889); Archiv f. exp. Path. u. Pharm., 26, 181 (1889). 406 CEDEMA AND NEPHRITIS the blood is noted in no conditions more strikingly than in nephritis and its oft associated "uremia." §4 Our argument thus far has shown that in nephritis there is a great increase in the hydrogen ion acidity of the urine, and that in both the urine and the blood there occur changes in the titration values which clearly indicate that both are holding a more than normal amount of acid. Our knowledge of physi- cal chemistry (the laws of chemical equilibrium) permits us to utilize these facts as evidence indicating that the kidney itself, in other words, everything which lies between the urine on the one hand and the blood on the other must, under such circum- stances, also shows an increased acid content. But it would strengthen this view if we could bring more direct proof in support of this deduction. It would be well, of course, if we could obtain a direct measure of the hydrogen ion concen- tration in the kidney. Gas-chain methods are naturally not applicable to sohd organs, and to apply them to the expressed juice of the kidney would be to introduce so many errors into the whole prolbem as to render the conclusions valueless. We can, however, obtain material help by using indicators. Proof of an increase in the amount of acid held by the kidney cells in conditions associated with the urinary findings of nephritis is furnished by the following facts: In 1885 H. Dreser ^ described a series of experiments on the excretion of dyes by the kidney which differed from the preceding studies of this subject as first made by R. Heiden- HAiN ^ and M. Nussbaum,^ in that he utilized the results of his experiments in an attempt to get an answer to the question as to where in the kidney the acid of the urine is secreted. Dreser made chief use of acid fuchsin which he injected in 5 to 10 per cent solutions (amounts not stated) into the dorsal lymph sacs of frogs. This dye has the property of being red in aqueous solution only in the presence of an acid; in an alkaline solution it becomes practically colorless (yellow). Dreser therefore rea- soned that the presence of a red color in any tissue after the in- >H. Dreser: Zeitschr. f. Biol., 21, 41 (1885); ibid., 22, 56 (1886). 2R. Heidenhain: Pfltiger's Archiv, 9, 1 (1875). ' M. Nussbaum: Pfltiger's Archiv, 16, 141 (1878). NEPHRITIS 407 jection of this dye into the circulation of an animal was evidence of an acid reaction in that tissue. The first fact noted by Dresbr that is of interest to us is that after a single dose of acid fuchsin the urine is found shortly thereafter to become brilUantly red. If the kidney from such an animal is examined no stained cells are noted anywhere in the kidney. To interpret this fact we would have to say that normally the urine is acid in reaction, but the cells of the normal kidney are not. The following may serve to corrobo- rate this finding of Dreser: Experiment 55. — Three frogs, weighing 35 grams each, are injected, respectively, with 0.25, 0.5, and 1.0 cc. of an aqueous 1 per cent acid fuchsin solution, into the dorsal lymph sac. All are seen to secrete a red-colored urine before being killed. They are killed respectively after 1, 1|, and 4^ hours. On autopsy, red urine is found in the bladder of each animal. The kidneys are not stained. They are rapidly removed from the freshly killed animals, frozen with liquid carbon dioxid (on a Babdeen freezing microtome, where the gas does not come in con- tact with the tissue) and sectioned. The sections are immediately transferred to a sUde (without being brought in contact with water or any other medium except air), covered with a cover slip, and examined under the microscope. None of the kidney tissues is seen to be stained. To be sure that the freezing plays no part in the findings, a parallel series of free-hand sections and crush preparations of the kidneys are made. No stained cells are found. When the uncolored sections are touched with very dilute acetic acid they are seen gradually to assume a pink color. Acid fuchsin is therefore present in the kidney tissues, but as cut from the body the reaction of this organ is not such as to allow its red color to appear. The pink tinge visible in the kidney after being touched with acid includes the glomeruli. Experiment 56. — To show that what was said for the frog holds also for the mammal, two young rabbits, weighing, respectively, 184 and 189 grams, received into the ear veins 2 and 4 cc, respectively, of a 1 per cent aqueous acid fuchsin solution. At the end of thirty and thirty-five minutes, respectively, they were killed by a blow on the head and immediately autopsied. Light red urine was found in the bladder of the first, deep red urine in that of the second. The appearance of the kidneys in both animals was entirely normal, and no dye was visible in the kidneys either macroscopically or microscopically. When a little very dilute acetic acid was permitted to flow under the cover slips, the sections turned uniformly pink, Dreser noted no staining of the frog's kidney until he had repeated his acid fuchsin injections several times. Then he 408 CEDEMA AND NEPHRITIS found that the cells of the convoluted and of the straight tubules began to stain red. He interpreted this finding by saying that from the long-continued effort on the part of these cells to excrete the dye, they become fatigued and so some of the dye remains behind to be discovered on subsequent section of the kidney. From all these facts Dreseb concluded that the acid constituents of the normal urine are " secreted " by the convoluted tubules, and that since the glomeruli and their capsules remain unstained, the " urine " coming from these must be " alkaline " in reaction, to change to an acid reaction after passing by the convoluted tubules. Whether such conclusions are really justified we shall have occasion to discuss later. No one can quarrel with the simple experimental finding that acid fuchsin does not stain the normal kidney, and does do this after repeated and long-continued injections. Such staining of the kidney Dreser still regards as "physiological." Strictly speaking, and for reasons that will be apparent as we go on, I shall myself regard it as " pathological." What Dreser calls the "fatigue " of the cells of those portions of the kidney which stain after repeated injections of the acid fuchsin, we are perfectly safe in regarding as the first evidences of an abnormal acid content in these cells, and we may hold that the repeated injection of this dye is itself responsible for such a condition. Acid fuchsin is a weak acid, and must produce the same effects upon the kidney that we know are produced by the injection of any other acid.^ After the injection of acids we note regularly all the signs of a nephritis, and that these were not absent in Drbser's experiments is clearly evidenced by the " anuria " which this author so often noted in his frogs. But Dreser ^ describes yet another experiment which shows that an abnormal production or storage of acid occurs in the kidney in nephritis. The kidney of the frog receives a blood supply, it will be remembered, from two sources — through the renal artery, as in mammals, and through a sort of portal system analogous to that existing in the liver. The blood from both these sources mixes to leave the kidney by way of the renal vein. Dreser noted that if acid fuchsin is injected into the abdominal vein an hour after the renal artery has been tied, the convoluted '■ See page 415. "R. Dreser: Zeitsohr. f. Biol., 21, 53 (1885). NEPHRITIS 409 tubules stain red. As already pointed out, no such red staining of the cells is noted if the dye is so injected without ligation of the renal artery. Dreser interprets his finding in the terms of physiology, but that we deal here with a pathological condition of the kidney — a nephritis — is betrayed not only by the fact noted by Dreser, that kidneys so treated secrete no urine, but by the evidence furnished below,^ that after occlusion of the arterial blood supply to the kidney, acid develops in this organ, the kidney swells, the water secretion falls and casts and albumin appear in the urine. Further tinctorial evidence of an abnormal production or accumulation of acid in the kidney in nephritis is furnished by certain experiments of R. Heidenhain, M. Nussbaum and P. GRtJTZNER with sodium indigosulphonate. This dye behaves similarly to acid fuchsin. It is deep blue or indigo in an acid solution and yellow in an alkaline one. The somewhat con- tradictory conclusions of these authors, based on their studies with this dye, are easily put in order if we try to separate those of their findings which are pathological from the physiological. In my own experiments on rabbits and frogs, I have, first of all, never been able to confirm any but the conclusion of Nuss- baum,^ that no part of the normal kidney stains with sodium indigosulphonate. This corroborates the finding obtained with acid fuchsin — the normal kidney does not contain sufficient acid to bring out the blue color. Experiment 57. — -Four frogs, weighing 30 grams each, are injected respectively, with 0.25, 0.5, 1.0, and 0.25 cc. of a 1 per cent aqueous sodium indigosulphonate solution into the dorsal lymph sac. Blue urine is voided by each of the animals before being killed. After, respectively, forty minutes, fifty minutes, seventy minutes, and 3^ hours, their heads are cut off and they are autopsied. Blue urine is found in the bladders of the last three. Macroscopic examination shows no color anywhere in the kidneys of these animals, and microscopic examina- tion of frozen sections only confirms this fact. Experiment 58. — Three rabbits from the same litter, and weigh- ing 497, 575, and 447 grams, respectively, receive, respectively, through the ear vein, 1, 2, and 5 cc. of a 1 per cent aqueous sodium indigosul- phonate solution. They are killed by a blow on the head one hour after being injected. Blue urine is found in the bladder of each. This ' See pages 225, 424 and 529. 2M. Nussbaum: Pfluger's Archiv, 16, 141 (1878). 410 OEDEMA AND NEPHRITIS is also present in the ureter of the third. The kidneys are entirely unstained in the first two, and no color is found anywhere in the frozen sections prepared from these kidneys. The kidney of the third animal has a mottled blue appearance superficially, and one section shows some blue streaks radiating toward the pelvis of the kidney. Fro?en sections show no dye anywhere in the kidney substance proper. The blue streaks are due to dye found in the lumina of a few of the collecting tubules. In apparent contradiction to this simple conclusion that the normal kidney does not stain with sodium indigosulphonate, stand the classical experiments of Heidenhain/ who found certain portions of the kidney, notably, again, the convoluted tubules, to stain when the " secretion of the urine was sufficiently depressed." Heidenhain brought about the desired reduction in the secretion of urine by such procedures as transverse section of the spinal cord in the neck. But as he himself noted, this produces an enormous fall in blood pressure. Such a fall does not, however, leave the kidney in a normal condition — it spells not alone an anuria, but an albuminuria, and casts, in other words, a " nephritis." The staining of the kidney under these circum- stances is again evidence of an abnormal production or accumulation of acid in this organ, a conclusion that we shall shortly be able to corroborate by entirely different methods. Both Heidenhain and Dreser have laid special stress on the fact that the convoluted tubules stain under the conditions offered in their experiments, while the glomeruli remain unstained, because it is upon this fact chiefly that they (and their followers) have based their conclusion that the different parts of the urin- iferous tubule in its course from the glomerulus to the pelvis of the kidney have different functions. As generally held, these different parts are supposed to secrete into (or, according to Carl Ludwig, absorb from) the mother urine — the liquor postu- lated by W. Bowman to be separated from the blood in its passage through the glomeruli — as this flows down the uriniferous tubules, the different substances which serve to characterize the urine. I do not myself question the probability that the different por- tions of the uriniferous tubules have different functions, but that this is so is not proved by these particular experiments. Strictly speaking, the findings of Dreser and Heidenhain only show 1 R. Heidenhain: Pfliiger's Archiv, 9, 1 (1876); Hermann's Handbuch d. Physiol., 5, 346, Leipzig (1883). NEPHRITIS 411 that, under the conditions of their experiments, the neutrality mechan- ism existing in the convoluted tubules is broken down- more easily than that existing, for exam-pie, in the glomeruli. That this approxi- mates more nearly a correct interpretation of the observed phe- nomena is, as a matter of fact, indicated by the following: By simply continuing the conditions which were mentioned as effective in leading to a staining of the convoluted tubules, we get a staining of the glomeruli. Evidence for the correctness of this conclusion can be adduced even from some cursory experi- ments mentioned by Heidenhain and Grijtzner. As pointed out above, the conditions which lead to a staining of certain portions of the kidney with acid fuchsin or sodium indigosulpho- nate (excessive acid injection, ligation of renal artery, gross falls in blood pressure) are conditions which we can show by other means to be such as are associated with an abnormal production or accumulation of acid in the kidney. No matter how we inter- fere with a proper blood supply to the kidney we get such a pro- duction of acid. It does not, therefore, surprise us that when GBtJTZNER ^ produced circulatory disturbances in the kidney by injecting gum arable, he noted not only the development of anuria and albuminuria, but found at the same time that the glomeruli and their capsules now stained with sodium indigo- sulphonate. Quite as simply can we interpret Heidenhain's ^ finding that the glomerular tufts stain with sodium indigosul- phonate when the ureters are ligated. When this is done the urine is dammed back and accumulates in the space between the glomerular tuft and the parietal layer of the capsule, in con- sequence of which the capillaries composing the tuft are com- pressed, so that the normal circulation of blood cannot now occur through them. Under these circumstances an abnormal pro- duction or accumulation of acid in the cells of the glomerulus and the capsule is rendered possible and so the tissues making up these structures now stain. One can further test the soundness of the reasoning detailed here that a staining of the kidney as a whole or in part marks the presence of an abnormally high acid content by working with excised kidney. Slices of fresh kidney kept in dilute solu- ip. Gbutznek: Pfluger's Archiv, 24, 461 (1882). 2 R. Heidenhain: Hermann's Handbuch d. Physiol., 5, 372, Leipzig, (1883). 412 CEDEMA AND NEPHRITIS tions of sodium indigosulphonate or acid fuchsin stain only very slowly and very slightly. But let a trace of acid be added and all parts of the section may be made to stain a deep, blue in a few minutes. In the same way a section of tissue from a kidney that has been dead some time (and so contains post- mortem acids) stains readily, and, let it be noted, in all its parts. It will be recalled by anyone familiar with such studies as those of Heidenhain, Dresek or the numerous investigators who since their day have adopted similar experimental methods, that these studies are intended to throw light on the problem of secretion by the kidney cells. This process of secretion is, of course, a dynamic one, made up of two parts, the one concerned with the taking up from the blood of the substance to be secreted, the other with the giving off of this same substance in the urine. The problems involved here are discussed in detail later, but it may not be amiss to point out even now that what is so often done, namely, the regarding of a mere staining of some or all of the cells of an organ as dependable evidence indicating that the dye is " secreted " by these cells, is entirely wrong. The presence of a dye in a cell does not mean this; nor when cells stain unequally does it mean that those most deeply stained are most involved in this process. It may mean just the reverse. The staining of the excised kidneys described above shows this very clearly. A kidney touched with a little acid, or one showing postmortem change, stains better than a normal kidney, and this without any hope of subsequently " secreting " the absorbed dye. Again, a kidney rendered " nephritic " by ligation of its arterial blood supply stains better than a normal one, and yet no one would maintain that a nephritic kidney " secretes " all dissolved sub- stances better than a healthy one. What really happens in the excised kidneys, or in the " neph- ritic " kidneys contained in the still living animal, represents but an isolated expression of the general laws that we to-day know to underlie all that is comprised in the physical chemistry of the process of dyeing. The kidney cells in the experiments that have been detailed are stained for the same reason, and their staining reactions mean the same thing, as when any ordinary lyophilic colloid such as fibrin or gelatin takes up add fuchsin or sodium indigosulphonate. If these colloids are seen to be stained red or blue, it means that they contain, under the conditions of NEPHRITIS 413 the experiment, a certain minimum of acid. But with a given concentration of the dye the depth of the staining becomes a measure of the acid content, for a given colloid will absorb the more of any so-called " acid stain " the higher the con- centration of the acid in the colloid. Other things being equal, the kidney cells must stain the more intensely with acid fuchsin or sodium indigosulphonate, the higher the acid concentration developed in them. To this whole question we shall have to return later. §5 Yet other Unas of evidence may be adduced to prove that in nephritis there is an abnormal production or accumulation of acid in the kidney. To some of these we return later.^ When the carnivorous animal (including man) is subjected to intoxication with acid, it meets this to begin with, by neutraliz- ing the acid with the fixed bases of its body. But when these are heavily drawn upon the organism has a reserve mechanism which enables it still further to tolerate an acid intoxication — it converts an increasing amount of its protein material into ammonia and uses this to neutralize the acid. The high relative and absolute ammonia excretion so common in many of ike nephritides or in patients likely to show albumin and casts in the urine with a diminished water output, as in starvation, after the anesthesias, after poisoning with phosphorus, arsenic, lead, etc., becomes evidence, therefore, of the existence of abnormally great amounts of acid in them. (Munzer, Palma, Aeaki, Badt, Laub, etc.) The low carbonic acid content of their blood is further proof in the same direction.^ When any of the fixed acids are introduced into the body, or are produced there (lactic, diacetic, betaoxy- butyric, abnormally large amounts of sulphuric, phosphoric, etc.), either in the course of normal or deranged metaboUsm these tend to drive off the volatile carbonic acid from the blood exactly as in test-tube experiments. The low carbonic acid con- tent of the blood so common in nephritis and in many of the 1 See page 638. 2 See H. Meter: Arch. f. exp. Path. u. Phann., 14, 313 (1881); 17, 304 (1883). See also Lewis, Rtffbl, Wolf, Cotton, Evans and Barcboft: Heart, 5, 45 (1913); Jour. Physiol., 46, 53 (1913); Yandbll Henderson: Jour. Am. Med. Assoc, 63, 318 (1914). 414 CEDEMA AND NEPHRITIS intoxications accompanied by casts, albumin, etc., in the urine is explained in this way. Quite recently interesting colloid-chemical evidence of an abnormal production or accumulation of acid in the body in nephritis has been brought by F. von Hoefft.^ The coagu- lation temperature of proteins is reduced, their alcohol precipi- tability and electrical conductivity increased whenever their acid content is raised.^ The blood of patients with nephritis shows all these changes. The work of A. W. Sellaeds ^ on the tolerance of patients to administration of alkali (sodium bicarbonate) before they react to the point of secreting a neutral urine is also of interest here. While normal individuals require some 5 to 10 grams, patients with recognized " acidoses," produced by feeding acid or in diabetes, were found to need more than these amounts (30 grams) before their urines turned neutral. Sellabds pro- poses the use of such alkali feeding by way of settling whether the " nephropathies " are due to " acidosis." Without question- ing the justice of Sellards' classification of his patients — for as I have insisted it is most important to know whether all of a kidney is diseased or only pieces in it — he finds that five of his thirteen cases showed a normal tolerance while in all the rest it was increased. In one patient he did not get neutral urine even after injecting 60 grams of sodium bicarbonate intraven- ously, and in a second the urine was still acid after 130 grams. Sellards' work has frequently been quoted as evidence against my views, but if I am any judge, it is simply a way of saying backwards that there is evidence of an abnormal production or accumulation of acid in the body in nephritis. Ill ANY MEANS WHICH LEADS TO AN INCREASED PRODUC- TION OR ACCUMULATION OF ACID IN THE KIDNEY IS A MEANS OF PRODUCING NEPHRITIS We are now ready to discuss the converse of what has gone before, and so try to show that any means by which we can bring I F. VON Hoefft: KoUoid Zeitschr., 13, 278 (1913). ' See page 108. ^A. W. Sellaeds: Johns Hopkins Hosp. Bull., 23, 289 (1912); 26, 141 (1914). NEPHRITIS 415 ■about an abnormal production or accumulation of acid in the kidney constitutes a method of producing the signs of nephritis. The simplest way of increasing the acid content of the kidney consists, of course, in the introduc- tion into this organ of an add of some kind. This is done most easily by injecting the acid, either in solution in water or in a " phys- iological " salt solution, directly into the general circulation of an animal. For this purpose I used, in my own experiments, a large- sized aspirating syringe with a two-way valve, rubber tubing and a hypodermic needle, as illustrated in Fig. 115. The acid solution warmed to 37° C. is sucked into the syringe through the tube a. After turning the valve v it can be ejected, on lowering the plunger, through the tube b, which ends in the hypodermic needle n. The needle is inserted into the ear vein of a rabbit and is held in place by a couple of small artery forceps. As the acid is injected intravenously, one observes the normally alkaline urine of the rabbit to become neutral and then to turn acid, and as this acidity rises, albumin appears in the urine. The following experi- ments dealing with the effects of such intravenous acid injections will serve to illustrate this point. Let it be noted that in addition to the appearance of albumin in the urine, this comes to contain various casts, epithelial cells, blood corpuscles and hemoglobin. By Figure 115. 416 (EDEMA AND NEPHRITIS comparing the urinary output in these animals with that shown by normal animals,^ it is seen that this is decreased. Evi- dences of oedema are also not wanting; animals injected with an acid do not excrete the water that is injected with this acid as does a normal animal that is given water only, in the form of A " physiological " salt solution. The water when injected unth an add is retained in the body, but to this phase of the problem of nephritis we shall need to return later. For the present it is clear that there develop all the most typical signs of an acute nephritis when acid in sufficient amount is injected into an animal. ExPEEiMENT 59. — Belgian hare; weight 1870 grams. Has been fed corn, oats, hay, and cabbage. Urine obtained by gentle manual pressure over the bladder.^ In the time of the experiment there are injected, at 37° C. and at a uniform rate, with the exceptions noted, 291 cc. of the following mixture: 300 cc. n/20 HCl+20 cc. 2/m NaCl. Time._ Amount of urine in cc. Remarks. 1.20 1.45 2 00 " Ol 6 OJ Turbid, yellow, no albumin, no casts. 2. IS 2.30 2.45 3.00 Few drops 1.9 Injection into ear vein begun. Turbid, yellow, no albumin, no casts. Clear, yellow, faint trace albumin, no casts. 1.5 Clear, brownish tinge, albumin present, few red blood cor- puscles, isolated kidney cells, no casts. 3.15 3 30 1.61 1 3/ Smoky urine, albumin, isolated granular and epithelial casts. 3.45 4.00 4.15 4.45 5.00 2.2 4.8 1 21.0 i Smoky urine, albumin, isolated granular and epithelial casts. Injection interrupted for 2J^ minutes. Smoky urine, albumin, isolated granular and epithelial casta. Injection interrupted for five minutes. Smoky urine, albumin, isolated granular and epithelial casts. Hemoglobinuria. Injection interrupted for ten minutes. Animal dies. 1 See pages 290 and 296. ^ In these experiments on nephritis the greatest care is necessary not to injure the lower urinary passages and so get a bleeding which might, through the presence of albumin and blood in the urine, lead to the erroneous con- clusion that a nephritis is at hand when only some bleeding is occurring into the bladder or urethra. Manual pressure over the bladder must be made with gentleness, and care must be taken not so to crowd the bladder into the pelvis as to kink the urethra. Only the smallest soft rubber catheter, well vaselined, must be introduced. If these precautions are not followed, fallacious, if not worthless, results are obtained. When an animal dies or is killed, the lower urinary passages must be examined for hemorrhagic points. NEPHRITIS 417 Total urine secreted since beginning injection 34.3 cc. AiUopsy. — Weight of animal 2135 grams! No free fluid in peri- toneal, pericardial, or pleural cavities. Kidneys slightly bluish, and bleed freely on section. Nothing about them is strikingly abnormal. Experiment 60. — Belgian hare; weight 2008 grams. Has been fed a mixed diet of corn, oats, hay, and cabbage. Urine obtained by gentle pressure over bladder. During the course of the experiment there are injected at 37° C, and at a uniform rate with the excep- tion noted, 90 cc. of the following mixture: 90 cc. n/10 HCl -|-6 cc. 2/m NaCl. Time. Amount of urine in cc. Remarks. 3 35 3 40 Injection into ear vein begun. Turbid, yellow, alkaline to litmua. No albumin, no casts. Clearer, trace of albumin present. Urine smoky, albumin increasing. Injection stopped for fifteen minutes as animal threatens to die. 3.5S 4.05 4.20 4 40 6.0 1.7 0.8 4.45 4.47 Few drops Bloody, much albumin, red blood corpuscles, great numbers of granular casts of various sizes. Total urine secreted since commencing injection 8.5 cc. Autopsy. — Weight 2078.5. No free fluid in the peritoneal, pleural, or pericardial cavities. Kidneys slightly swelled. Under the cap- sule appear tiny hemorrhagic points. Experiment 61. — Belgian hare; weight 2259 grams. Diet unknown, as he has just been received in the laboratory. Urine obtained with a catheter. In the course of the experiment 75 cc. of the following solu- tion are injected intravenously at a uniform rate, with the exception noted: 75 cc. n/10 HCl -|-5 cc. 2/m NaCl. Time. Amount of urine in cc. Remarks. 11.30 11.45 12.00 12.15 12.30 12 40 7.0 19.4 2.01 0.4} 1.6 Tied to animal board. Urine thick, chrome yellow, no albumin. Thick, chrome yellow, no albumin, alkaline to litmus paper. Injection into vein of ear begun. Thick, chrome yellow, no albumin, alkaline to litmus. Clearer, pinkish tinge, albumin present. 12.45 1.45 3.7 11.5 Urine distinctly red, much albumin, many casts. Urine turbid, red, shows spectrum of oxyhemoglobin, filled with albumin, casts, (epithelial, granular and mixed), epithelial cells and red blood corpuscles. Animal released in good condition, returned to hutch. 418 (EDEMA AND NEPHRITIS Total urine secreted since beginning injection 19.2 cc. 5.30 5.0 f per cathe- ter. 37.0 per cathe- ter. Clear, yellow, acid. Casts and albumin still present. 11.00 f Next i morning [ Dark amber, thick, faintly acid, clear. Microscopic examina- tion shows many squamous epithelial cells and isolated casts. Carefully filtered urine shows a trace of albumin. These nephritides produced experimentally in animals have a perfect parallel in the albuminurias with blood, casts, and a diminished urinary output to the point of cessation which are observed from time to time in human beings who have inhaled or swallowed by accident or intent sufficient quantities of various acids. §2 It will be retorted by some that to inject acid intravenously is so " abnormal " that it and its consequent nephritis has noth- ing in common with the albuminurias and nephritides observed in human beings. To meet this criticism it is only necessary to examine the urine in conditions in which large amounts of acid are produced " physiologically " within the body itself. As is well known, large amounts of acid (especially lactic, acid) are produced in the muscles when these contract. If the muscle works under physiological conditions and not too fast, the acid as formed may be largely oxidized in situ. But if the muscle works more rapidly then more acid is produced than can be oxidized in the muscles and so in the higher animals some passes unchanged into the blood, with this to the kidneys, and out in the urine.^ It is evident that the opportunities for such an accumulation of acid in the body become the greater the more rapidly and the harder the musculature of the body works, and we should add, the more defective the oxygen supply to the working muscles, for this element is necessary for the proper oxidation of the acid in the body. Now such a combination of hard work with a (temporarily) defective oxygen supply to the active muscles is furnished whenever the organism engages in exercise that calls for more than usual effort. We are therefore 'Trasaburo Araki: Zeitschr. f. physiol. Chemie, 19, 422 (1894), where references to his earher papers may be found. Hoppb-Seyler: ibid., 19, 476 (1894). Fletcher and Hopkins: Journal of Physiology, 35, 247 (1907). NEPHRITIS 419 not surprised to find that soldiers after prolonged marches, women in labor, Marathon runners, etc., show albumin, casts, blood, etc., in the urine when examined after such exertions.^ The amount of exercise needed to bring about such albuminurias is really surprisingly low, as is indicated by the following: Experiment 62. — Seven trained athletes just before entering upon a game of basket ball were asked to void their urine into a series of flasks. At the end of the game, which lasted 1| hours, they voided their urine a second time into a second series of flasks. Helleb's Figure 116. Figure 117. test was then applied to the various specimens of urine. While none of the "players showed any trace of albumin in his urine before the play, all gave marked reactions after the game. The results of the tests applied to the urines voided after the game are shown in Figs. 116 and 117. The first four tubes are photographed against a white background, the three of Fig. 117 against a black. The faint albumin ring present in the tube on the extreme right of Fig. 117 scarcely shows in the photo- graph. Interestingly enough, this specimen of urine came from a player who was in the game but five minutes. Experiment 63. — Five trained athletes shortly before engaging in a match game of basket ball void their urine into a series of flasks. 1 W. Lbubb: Virchow's Arch, 72, 145 (1878). G. Edlbfsen: Centralbt. f. d. med. Wissensch., 762 (1879). C. von Noordbn: Aroh. f. klin. Med., 38, 205 (1886). 420 CEDEMA AND NEPHRITIS All the urine voided during the succeeding li hours during which the game is played is collected in a parallel series of flasks. In none of the control urines with the exception of that of Player IV are there found albumin or casts. This player had found it necessary before coming to the game to rush about town making train and street-car connections and had moreover had a " cold " for three days previ- ously. After the game all the players showed an albuminuria and a great many granular, hyaline, and mixed casts. The albumin and the casts in the previously affected individual were markedly increased. The findings are illustrated in Fig. 118 and in the appended Table Figure 118. LXXVIII. The five tubes on the right show the results of applying the cold nitric acid test to the urine after the game. The tube on the extreme left shows the albuminuria existing in Player IV even before entering the game. The quantitative estimations in the Esbach tubes were carried out in the ordinary way using Tsuchiya's ^ phos- photungstic acid reagent. The photograph was made after the tubes had stood for only six hours. The readings in the table were made after twenty-four hours. iTsuchita: Centralbl. f. inn. Med.,' 29, 105 (1908). Phosphotungstic acid, 1.5 grams. Concentrated hydrochloric acid, 5 cc. Alcohol enough to make 100 cc. NEPHRITIS 421 TABLE LXXVIII Before the Game Player Amount of urine in cc. Nitric acid test. Casts. I II III IV V 60 158 5 47 134 Negative Negative Negative Poaitive Negative None None None Occasional granular and hyaline None After the Game (1| Hour Period). Player. Amount of urine in cc. Nitric acid test. Casts. EsBACH reading with phospho- tungstic acid. Albumin excreted in grams. I II III IV V 168 69 35 30 94 Positive in all Many hyaline, granular and mixed casts present in all. 0.6 3.25 2.3 5.0 2.75 0.111 0.224 0.080 0.150 0.268 Av. 0.163 A remarkably short period of hard athletic work suffices to produce a great albuminuria, as the following taken from many such observations shows: Experiment 64. — B , a well-trained and expert Univer- sity runner ran a quarter-mile race. Before starting he voided 54 cc. of urine which on exami- nation showed no albumin . After his race (time: 58 seconds!) he voided 59 cc. of urine in which much albumin was found. In Fig. 119 are shown the results of the albumin tests as applied to the two samples of urine. In the two tubes on the right the cold nitric acid test has been applied to the urines; in the tube on the left a quantitative estimation has been carried out in an Esbach tube with Tsuchiya's phospho- tungstic acid reagent. Figure 119. 422 CEDEMA AND NEPHRITIS §3 A condition in the body analogous to that produced vol- untarily by the athlete in his athletic activities is created through any uncompensated heart lesion or any disease of the lung of such a character as to interfere materially with the proper aeration of the blood. Under these circumstances there is not produced the excessive amount of acid by extra mus- cular exertion, but the oxidation of such amounts as are normally present has been decreased by not permitting the nor- mal amount of oxygen to get to the tissues of the body. The end result is, of course, the same. A defectively functioning heart or a sufficiently disabled lung interferes first of all with the proper escape of carbonic acid from the blood (and so from the cells in which this is produced).^ But they do more than this, they place the organism as a whole in a state of lack of oxygen, and as a necessary consequence of this we know from the studies of Teasaburo Ahaki,^ Hermann Zillessen,^ and P. VON Terray* that we get an abnormal production and accumulation of other acids, notably lactic and oxalic acids, in the tissues. Heart or lung lesions therefore are potent to lead to that same abnormally high acid content of the cells of the kidney that we previously found created through the direct injection of acids, or the hard work of the athlete, and so we are prepared to find in these pathological states of the heart and lung that albuminuria with casts, and a defective secretion of water is again a common consequence. As a matter of fact the associa- tion of "nephritis " or "Bright's disease " with heart lesions of the most varied kinds, or pathological conditions in the lung (manual compression of the thorax, pleurisy with effusion) that reduce its ventilation area sufficiently, is so constantly observed that it is taken for granted clinically. 1 Strassbitbg; Pfltiger's Arch., 6, 94 (1873); A. Ewald: Arch. f. (Anat. und) Physiol., 663 (1873); 123 (1876). 2T. Araki: Zeitschr. f. physiol. Chemie, 15, 335 and 546 (1891); 16, 453 (1892); 17, 311 (1893); 19, 422 (1894). 3 H. Zillessen: Zeitschr. f. physiol. Chemie, 15, 387 (1891). «P. vonTerray: Pfliiger's Arch., 65, 393 (1896). NEPHRITIS 423 It requires no special comment to recognize that a whole series of pathological states such as the severer anemias, carbon monoxid poisoning/ and epileptic seizures, which at first sight seem to have nothing in common with each other, contain within themselves all the elements necessary for the development of the signs of a nephritis. The severe anemias (leukemia or per- nicious anemia) merely constitute further ways of interfering with a proper oxygen supply to the tissues. Both are accom- panied by an abnormal storage and production of acid in the tissues as evidenced by Felix Hoppe-Seyler's ^ and T. Irasawa's ^ chemical analyses of the urine, and R. von Jaksch's * titrations of the blood in cases of severe anemia. An abnormal acid production in carbon monoxid poisoning has been proved by T. Araki,^ E. Munzer and P. Palma; ^ in epilepsy (severe muscular exertion with defective breathing) by Araki and E. Mendel. As clinicians well know, the finding of albumin, casts, blood, etc , in the urine in any of these pathological states is usual. §5 The etiological importance of " cold " (in the strict sense of the word as a lowering of the body temperature and unac- companied by an infection) in the production of an acute nephritis, or in the lighting up of a chronic one that has slumbered for a time, has always been insisted upon by earlier observers. This view finds a rigid scientific support in our present knowledge of the physiological effects of low temperature upon the warm- blooded animals. Of these none is more characteristic than the rise in the acid content of the cells of an animal so exposed.'^ ' G. Thompson: Trans. Assoc. Am. Physicians, 1902; William Ravine: Personal Communication. 2F. Hoppb-Seylee: Zeitsohr. f. physiol. Chemie, 19, 473. (1894). s T. Irasawa: Zeitschr. f. physiol. Chemie, 15, 380 (1891). * R. VON Jaksch: Klinische Diagnostik, Funfte Aufi., 2, Berlin (1901.) 6T. Akaki: Zeitschr. f. physiol. Chemie, 15, 335 (1891). 'E. Munzer and P. Palma: Prager Zeitschr. f. Heilk., 15 (1894). 'See Araki: Zeitschr. f. physiol. Chemie, 16, 453 (1892). On the basis of this same acid production we can with ease explain the precipitation of an attack of hemoglobinuria in the cases of so-called paroxysmal hemo- 424 CEDEMA AND NEPHRITIS How potent is this element of cold in leading to the signs of a nephritis has been well brought out by R. D. Kennedy,^ who in Northern Michigan (Calumet) during extremely cold weather found albumin in 40 per cent of all patients examined who had been exposed to it. Thirteen of fourteen physicians in the hospital had albumin and casts in their urines some time through the winter. The only exception was an eye speciaUst who worked indoors. But even more trivial exposures to cold suffice to bring about these consequences. A cold bath, for example, leads, in not a few individuals, to the appearance of albumin in the urine. §6 Thus far we have discussed only general conditions — con- ditions affecting the whole animal — that are capable of indu- cing an abnormal storage or production of acid in the body, and so of inducing a nephritis. We shall now consider a series of more local conditions that bring about the same result. Instead of interfering with the normal action of the heart or lungs an effective state of lack of oxygen in the kidney can, of course, be induced by direct interference with the normal blood flow through this organ. Experimentally such a condition is easily established by total or partial ligation of either the arterial or the venous blood supply of this organ, a state that has its clinical par- allel in such affections as partial or complete occlusion of the renal vessels through arteriosclerosis, thrombosis, embolism or the pres- globinuria when these patients take a cold bath, are exposed to cold, etc. The acid produced under these circumstances rises to the point where it leads to hemolysis of the patient's red blood corpuscles. This view is sup- ported by the fact that it is possible to precipitate an attack of hemoglo- binuria for diagnostic purposes quite as easily through temporary obstruc- tion of the circulation in the arm by applying a band about it (accumulation of carbonic acid and production of other acids due to a lack of oxygen) as through the customary immersion of the extremities in cold water. The essential nature of the paroxysmal hemoglobinurias would seem to reside in the lesser resistance which the red blood corpuscles of such patients have to such a hemolytic agent as an acid. The resistance is enormously increased by the addition of various salts to the blood, as Oscar Berghausbn has shown. This fact is not only of theoretical interest, as I have tried to show in discussing the nature of hemolysis (see page 364 or Fischer: KoUoid Zeitschr., 5, 146 (1909)) but of practical use in the treatment of these cases of hemoglobinuria which need a diet rich in alkalies, calcium salts, etc. iR. D. Kennedy: Personal Communication (1912). NEPHRITIS 425 sure of tumors, etc., upon these vessels. But as the experiments of T. Aeaki and H. Zillessen have shown, such an interference with the normal blood supply (oxygen supply) to any of the parenchym- atous organs is followed immediately by the accumulation of aeids in the affected tissues. Do we find that in such local cir- culatory disturbances of the kidney we get an albuminuria? That we do is, of course, known to everyone — it constitutes, since Max Herrmann's ^ experimental studies, one of the classi- cal facts of pathological physiology; it is attested to by the experience of the medical diagnostician; it is the bugbear of sur- geons who operate on the kidney and find a temporary closure of the renal vessels expedient or necessary.^ §7 Instead of interfering directly with the oxygen supply to the kidney by procediu'es which interfere with the blood supply to this organ, we can bring about the same result in a more subtle way by giving the kidney parenchyma its normal oxygen supply, but by so interfering with the chemistry (enzymotic processes) of the cells themselves that make up the kidney as to render these incapable of utilizing in proper form the oxygen that is freely supplied them. So far as the end result is concerned, it matters little, of course, whether we interfere with the normal oxidation, say, of the carbohydrates of the living cell to carbonic acid by shutting off the oxygen supply to the cell and so halting the decomposition of the carbohydrates when these have been changed to lactic, oxalic, formic and other acids (saccharinic acids) ; ^ or whether we do nothing about the oxygen supply but 'Max Hekkmann: Sitzungsber. d. Wiener Acad. Math.-phys. Klasse, 65 (1861). 2 For a discussion of the methods to be employed in combating the evil consequences of such temporary closure see the section dealing with the treatment of nephritis. ' The chemical aspects of this problem of the formation of acids from carbohydrates in the absence of oxygen are discussed by Felix Hoppe-Seyler : Berichte d. deut. chem. Gesellsoh., 4, 346 (1871); H. Kiliani: ibid., 15, 701 (1882); DucLAUx: Compt. rend., 94, 169; Schutzenbebgeb: ibid., 76, 470; BucHNER, Meisenhbimeb, and Schade: Berichte d. deut. chem. Gesellsch., 39, 4217 (1906); J. U. Nep: Liebig's Annalen, 357, 214 (1907). The biochemical aspects of this same problem are discussed in the papers on lack of oxygen already referred to on pages 193 and 230. 426 CEDEMA AND NEPHRITIS introduce something into the cell which prevents the oxidation of the lactic acid as formed (or more probably its mother sub- stance, glycerin aldehyd) to carbonic acid.^ The cells of the living body in the end get into the same state whether they have their oxygen supply cut off or whether this is not interfered with, but they are " poisoned " in such a way as to be unable to utilize this oxygen as normally. As has been shown particularly well by T. Araki, a large number of poisons lead to the same state of lack of oxygen, with its associated abnormal production and accumulation of acids in the tissues, as do the grosser interferences with the oxygen supply to the various organs or the body as a whole, that have already been described. And so it cannot surprise us to discover that Aeaki's list of poisons — poisons utilized to show that an abnor- mal add production is the constant accompaniment of a state of lack of oxygen in the tissues no matter how produced — is identical with the list of poisons familiar to any laboratory or clinical worker who has busied himself with the problem of the toxic nephritides: metaUic salts, such as those of arsenic, uranium, chromium and lead; alkaloids, such as morphin, cocain, veratrin and strychnin; anes- thetics, such as alcohol, acetone, ether and chloroform; unclas- sified poisons, such as amyl nitrite, the cyanids and phosphorus. § 8 In concluding this section we need to discuss the albuminurias encountered in three conditions which not only are readily inter- pretable on the basis of our contention that albuminuria results whenever abnormally great amounts of acid accumulate in the kidney, but give this contention valuable support. Since Rudolph Virchow's description of the condition fifty years ago, the albuminuria of the newborn constitutes a matter of common knowledge to every pediatrist. It occurs in perfectly healthy infants as a transitory phenomenon, is regarded as " phys- iological," and to it ordinarily no clinical importance is attached. Whence comes it? The condition is most commonly found in "hard " labors, when the cord prolapses, in breech presentations, etc., all of them conditions which mean a state of more than the ' In this connection see the interesting work of R. T. Woodyatt: Jour. Am. Med. Assoc. 55, 2109 (1910). NEPHRITIS 427 normal lack of oxygen in the organism of the child during the process of its birth. Even normal labor means, of course, a decided interference with the circulation of the infant — is it not in this fact and the associated accumulation of carbonic acid and other acids in the blood that the cause of the first respiration is to be sought, as ZuNTZ has shown? Difficult labors mean in toto only a more than usual interference with the circulation of the child. It is entirely a matter of definition as to just how much of this we will accept as " physiological." But when we have thus connected the development of the albuminuria with a dis- turbance in the general circulation of the child then we have made it, at the same time, a mere subheading of the albuminurias discussed in § 3 of this section (page 422), and the albuminuria is " physiological " only as we will accept httle or great inter- ference with the circulation in the infant during its birth as " physiological." Albuminuria is a common accompaniment of salt starvation, be this a complete salt starvation or only such a partial one as is induced by eliminating completely the sodium chlorid from the food. Under this same heading is to be classed the albu- minuria consequent upon the excessive consumption of water low in salts. The latter washes the salts out of the body ^ and so leads indirectly to the same state as that induced by a lack of salts in the diet. The effect of a salt-free diet is twofold. In the first place it leads to the accumulation of acids in the tissues.^ Other things being equal, we have on this basis alone therefore a reason for the appearance of the urinary findings characteristic of nephritis when salts are withheld from the diet. But the salts act in yet another way. As already discussed, and as we shall see in greater detail later, many of the changes induced in colloids by acid may be greatly inhibited through the presence of all salts, even neutral salts incapable of an effect that might be construed as due to a mere neutralization of the acid. Through the with- drawal of salts from the tissues, whether by salt starvation or through leaching these out with water, we favor, therefore, the development of the signs of a nephritis in two ways: not only do we render possible an abnormal production or accumulation of 1 See page 324. 2G. Btjngb: Zeitschr. f. Biol., 10, 111(1874); see also J. Forster: ibid., 9, 297, 369 (1873); N. Lunin: Zeitsohr. f. physiol. Chemie, 5, 31 (1881). 428 (EDEMA AND NEPHRITIS acids in the tissues, but we take away at the same time the action of the salts in reducing the effect of the acids. IV NEPHRITIS DUE TO OTHER THAN ACID CAUSES The colloid changes in the kidney which are characteristic of nephritis and which we shall discuss in greater detail later, such, for example, as the sweUing of the kidney, are inducible, as previously noted, by other substances besides acids. Any agency thus capable of increasing the hydration capacity of a protein colloid and under physiological or pathological cir- cumstances conceivably active in a kidney may in this way become a cause of the nephritic signs. Of the various ones which might be mentioned (alkalies, urea, pyridin, certain amins, etc.) which we touched upon in the discussion of oedema, we shall here consider only the first, namely, the alkalies, in illus- tration of this point. It so happens that the sum total of the chemical changes that go on in the living animal organism are of such a character as to threaten it chiefly from the acid side. Even under normal con- ditions, the tissues have to guard themselves against becoming acid. Is not carbonic acid among the chief end products of the oxi- dation of our foodstuffs? The normal tendency of the tissues to run over to the acid side is enormously increased under various patho- logical conditions, and as we shall find these conditions to be just such as are likely to lead to a nephritis, the discussion of this sub- ject will naturally claim our chief attention. An abnormally high alkali content in the cells under ordinary circumstances is scarcely possible, and when it is induced artificially it is difficult to main- tain, for the normal acid production (carbonic acid production) in the living cell tends quickly to neutralize it. This question is, therefore, scarcely to be considered in our further analysis of the problem of nephritis. Still from a theoretical standpoint and in poison cases it is quite as important as that upon which we shall lay the greater stress. We should, on the basis of our colloid conceptio7is of nephritis, be able to induce this condition experimentally quite as easily through alkalies as through adds. As the following experiments show, this is actually the case. NEPHRITIS 429 Experiment 65. — Belgian hare; weight 2085 grams. Has been fed hay, oats, corn and cabbage. In the course of the experiment there are injected intravenously at a uniform rate 125 cc. of the fol- lowing mixture: 150 cc. n/10 NaOH+10 cc. 2/m NaCl. Time. Amount of urine in cc. Remarks. 2.35 3.15 3.30 3.45 4.00 4.15 4.30 4.45 4.58 31.0 0.7 1.2 8.4 6.0 1.2 0.7 0.4 0.4 + . 6 con- tained in cathe- ter. Catheterized. Dark amber, acid to litmus paper. No albu- min. No casts. Weighed. Placed in animal board. Injection into ear begun. No albumin. No casts. Acid in reaction. Urine clearer. Acid in reaction (?) Trace of albumin (?). Milky, alkaline to litmus. Faint trace of albumin. Milky, alkaline to litmus. Isolated casts. Faint trace of albumin. Milky, alkaline to litmus. More albumin. Many long hya- line casts with coarsely granular material sticking to them. Milky, alkaline to litmus. Much albumin. Filled with casts. Filled with casts. Bloody tinge to urine. Milky, alkaUne to litmus. Much albumin. Filled with casts. Bloody tinge to urine. Animal dies. 1 . gram of feces lost. It is noted that the albumin reactions as obtained with cold nitric acid applied to the filtered acidi- fied urine are not as intense as in the albuminurias induced by acid injections. (Less albumin?). Total urine since beginning injection 18.9 cc. Autopsy. — Weight 2187 grams! No fluid in the cavities. Intes- tinal contents seem somewhat more fluid than usual. Kidneys are firm, apparently somewhat swelled, and do not bleed easily. Experiment 66. — White rabbit; weight 1911 grams. Fed hay, oats, corn, and greens. In the course of the experiment there are injected at a uniform rate 185 cc. of the following mixture: 225 cc. n/10 NaOH-l-15 cc. 2/m NaCl. Time. Amount of urine in cc. Remarks. 2.15 85.0 Catheterized. Turbid, dark amber, acid. No albumin, no casts. 2.30 0.7 Turbid, dark amber, acid. No albumin, no casts. 2.45 0.7 Weighed. Injection into ear vein begun. Urine as before. 3.00 0.2 Urine as before. 3.15 1.0 Neutral to litmus. Clearer. Small amount of albumin. Many hyaline casts. Some have coarse granules in them. 3.30 6.4 Urine clear as water. Some albumin. Many hyaline casts. Some have coarse granules in them. 3.45 15.5 Urine clear as water. Only a few oasts can be found. Albumin present. 4.00 22.0 Weakly alkaline. Albutain present. Isolated casts only can be found. 4.15 24.5 Albumin present. No casts can be found. The urine has a pink tinge (hemoglobinuria). No red blood corpuscles microscopically. 4.25 Injection stopped. Faintly alkaline. Clear, pink, no casts, no red blood corpus- 4.30 23.0 ties. Albumin present. Animal released. Seems entirely normal, and eats at once. Total urine since beginning injection, 92.6 cc. Weight 2000 grams! 430 (EDEMA AND NEPHKITIS ExPEBiMENT 67. — White rabbit: weight 2177 grams. Fed hay, oats, corn, and cabbage. In the course of the experiment there are injected at a uniform rate 240 cc. of the following mixture: 225 cc. n/20 NaOH+15 cc. 2/m NaCl. Injection made into ear vein. Time. Amount of urine in cc. Remarks. 1.50 10.0 Catheterized. Turbid, yellow urine. No albumin. No easts. 2.00 Weighed, 2.15 2.30 2.45 1 drop No albumin. Injection begun. 0.2 Albumin present. Filled with casts, mainly hyaline in char- acter, but some are finely granular. Much squamous epi- lium and cell detritus. 3.00 0.5 Alkaline to litmus. Albumin and casts as before, but all the casts are hyaline except for coarse, granular material con- tained in or attached to some. 3.15 0.2 Strongly alkaline. Albumin and casts as before. 3.30 2 drops Strongly alkaline. Albumin and casts as before. The urine has a pinkish tinge (hemoglobinuria). 3.45 1.0 Strongly alkaline. Albumin and casts as before. Urine pink- ish (hemoglobinuria). Red blood corpuscles are found and two microscopic blood coagula. This bleeding is attributed to traumatism (animal struggled and whipped catheter about). 4.00 2.8 Urine strongly alkaline. The animal has begun to shiver (acid production!) during the last fifteen minutes. The previously warm ears are pale and cold. 4.15 13.0 1 The urine becomes faintly alkaline, then scarcely affects either 4.30 12.0 red or blue litmus. The urine is clear like water except for 4.45 16.0 > a clouding due to (traumatic) blood. Careful search of the 5.00 19.0 sedimented urine reveals only an occasional cast. The 5.15 23.0 J animal is shivering constantly. It is killed. Total urine since beginning injection, 87.7 cc. Autopsy. — Weight 2326 grams! The peritoneal, pleural, and pericardial cavities are dry. The kidneys are soft and bleed a normal amount. A few pinpoint hemorrhagic spots are found in the bladder. V THE ALBUMINURIA 1. Introductory Remarks Having discussed the evidence which shows that, an abnormal production or accumulation of acid occurs in the kidney in every case of nephritis, and, conversely, that whenever such is brought about, the signs of nephritis become manifest, we need now to say how this one factor is able to produce the various objective signs which as clinicians we have come to regard as characteristic of this pathological entity. The first to be considered is the albuminuria. NEPHKITIS 431 The urine of man or of the various animals that serve us for experimental purposes does not under normal circumstances contain albumin in an amount that betrays itself when any of our ordinary laboratory tests are applied to it. By special methods it is possible to show that even such normal urine con- tains faint traces of albumin, but it is generally held that this is of no pathological significance and has behind it a no more serious cause (it is thought) than the shedding and destruction of a few cells from the tract through which the urine has to pass from the uriniferous tubules into the outer world. An albuminuria, q,s we shall use the term, will, therefore, have a meaning only as apphed to the presence of albimiin beyond this normal amount, and, we ought to add, of renal origin and not from somewhere below this organ. Nor has the mechanism of the albuminuria which we are discussing anything in common with the albu- minuria consequent upon gross destructive lesions in the kidney as when small or large blood vessels are ruptured, allowing their entire contents to escape into the urine. Our current hypotheses regarding the cause of albuminuria are familiar to everyone and are notoriously unsatisfactory. It is generally held that the (chief) albumin of albimiinuria is serum albumin, that it is derived from the blood, and that it is under normal circumstances prevented from going over into the urine by the kidney structures which lie between the urine and the blood. Some twenty years ago R. Heidenhain attempted to express the whole situation in satisfactory physico-chemical terms. He pointed out the colloid nature of the blood albumins, and called to mind Thomas Graham's fundamental differentia- tion between the colloids which do not diffuse through animal membranes and the crystalloids which do this readily. On this basis he maintained that the latter appeared in the urine because they could readily diffuse through the animal membrane that separates the urine from the blood, while albumin is absent because this colloid body cannot diffuse through such a mem- brane. We have not since Heidenhain's considerations gotten beyond this view. Simple and apparently satisfactory as this explanation is, it cannot stand the pressure of a little analysis. In nephritis this membrane is, of course, still present, and yet in this patho- logical state the albumin appears in the urine. To meet this 432 (EDEMA AND NEPHRITIS fact it has been generally maintained, and, let us add, without any experimental support whatsoever, that the "permeability" of the urinary membrane for albumin has been altered, so that it now lets this through. As a matter of fact, we have not even had offered us any parallel from the pages of physical chemistry for such a change in the permeability of any " membrane " that in the laboratory corresponds with such as we might have in the body, nor, so far as I know, has anyone attempted to say just what chemical or physico-chemical agent is responsible for the changes in permeability postulated in the case of the kidney. A first error in this theory of albuminuria (which represents the epitome of our present conceptions regarding its nature) arises from the fact that the albumin found in the urine is looked upon as coming from the blood. Such a belief has been entertained because it has been found that the albumin present in the urine shows a series of reactions which are identical with those obtained from serum albumin. But this does not yet prove that the albumin of albuminuria has come directly from the blood. Such a conclusion overlooks the important fact that the albumins contained in the kidney itself, in other words the albumins contained in the secreting membrane separating the urine from the blood, also show these reactions. None of the albumin reactions used in these tests is " specific." They only represent certain group reactions which colloid chemistry has shown us to be common to a large number of the protein colloids of animal origin. Such considerations carry with them the important conclusion that the albumin of albuminuria need not come from the blood at all {except indirectly); it may come from the urinary membrane itself. That, as a matter of fact, it does come from this will appear more distinctly as we proceed. Albuminuria results whenever conditions are offered in the body which permit the solid colloid membrane that separates the blood from the urine to go into solution in the urine. The chief reason why this occurs in nephritis resides in the fact that acids are produced which render the colloid membrane "soluble." To make clear what is meant by this conclusion we need but recall our previous remarks on the general structure of the kidney ^ and introduce some observations on this question of the " solubility " of colloids. Many scattered but important facts regarding this problem » See page 282. NEPHRITIS 433 are found in the literature of colloid chemistry, especially if we bear in mind that the " solution " of a protein is, in all prob- ability, not a simple affair. We have learned how under the influence of an acid, the protein particles are hydrated and in increasing amount with increasing concentration. But under the same influence, as the " dissolved " state is approached, new properties are likely to be exhibited by the protein as indicated not only by a fall in its viscosity but by changes in its diffusibility, Brownian movement, susceptibility to precipitation, etc. The protein moves from a state in which it is markedly colloid toward the crystalloid side. Technically put, its degree of dispersion is increased. An increase in swelling followed later by a decrease, an increase in viscosity giving way to a decrease, an increase in diffusibility, or in Brownian movement may all therefore be regarded as evidences for an increased " solubility." If this is borne in mind then Thomas Graham becomes one of the first students in this field, for he noted that the addition of acetic acid to egg albumin increased its diffusibility. On the other hand, E. von Reg^iczy ^ found the addition of sodium chlorid to delay its diffusion. Particularly important for our purposes are the studies of T. B. Wood and W. B. Hardy,^ who found plant protein (glu- ten) to maintain its " cohesiveness " (remain solid) in water and neutral media but to " disintegrate " and " dissolve " when a little acid was added. The solution in acid depended upon the nature and the concentration of the acid and was at all times inhibited by the addition of salts. While all salts (including sodium chlorid) showed this behavior some were relatively more powerful than others — the bivalent and tri- valent radicals being more active, generally speaking, than the monovalent ones. As previously emphasized, the urinary membrane (the kidney itself) is composed, in the main, of a mixture of hydrophilic protein colloids which, as we learned, are capable of existing in two fairly well-defined states: in a solid or gel state and in a liquid or sol state. A familiar illustration of such existence IE. VON REGiczY: Pfluger's Arch., 34, 431 (1884). "T. B. Wood and W. B. Hardy: Proc. Roy. Soc, London Series B, 81, 38 (1908). 434 (EDEMA AND NEPHRITIS in two states is offered by ordinary gelatin. Under certain conditions this appears in the form of a stiff jelly, under others as a " solution." In the same way fibrin represents the gel form of the sol fibrinogen, paracasein (casein) the gel of casein (caseinogen), ordinary soft rubber, the gel of a " dissolved " rubber, etc. It is generally recognized that a rather close relationship exists between the gel state in which the colloid is " swelled " and the sol state of this same colloid in which it is " dissolved," and yet the transition from the one state into the other is not necessarily or in all examples a perfectly smooth affair. We need only to call attention to the fact that ordinary gelatin, for example, when thrown into cold water merely swells up — it enters the gel state. But in this state it remains, one might say almost indefinitely; to mere appearance it does not go into solution at all as would, for example, the crystals of any salt thus thrown into the solvent. But if the temperature of the water is raised then the gelatin goes into solution rapidly — it passes over into the sol state. A change in temperature in this case is necessary to accompUsh its " solution." As to our mind albuminuria represents Just such a passage of a colloid in the gel state (the proteins of the urinary membrane) over into a colloid in the sol state (the proteins contained in the urine of the albuminuric individual), let us study the conditions favoring such a transition in more detail, paying especial atten- tion to such changes in surroundings as we might imagine could come into play in the cells of the living animal. I studied fibrin and gelatin in this regard. The following facts regarding them are of importance in the further development of our subject. 2. Observations on the " Solution " of Colloid (Protein) Gels (a) Fibrin. — When well-washed fibrin that has been thor- oughly dried and then powdered in a mortar is thrown into water it . swells up somewhat. Even though the vessel is thor- oughly shaken practically none of the protein goes into solution in the water. The matter is easily tested by filtering the water off the fibrin and treating the filtrate in any of the acepted ways for albumin. One must only be careful to use a fine filter or else not powder the fibrin so thoroughly that gross particles of it NEPHRITIS 435 can pass through the pores. By similar means it can be shown that the fibrin will not dissolve appreciably in any solution of the ordinary neutral salts. In a solution of any acid {or alkali) it not only swells up more than in water, but it goes into solution. Within certain hmits more and more fibrin goes into " solution " with every increase in the concentration of the acid (or the alkali). But in this matter there seems to exist an optimum above which the progressive increase in " solution " stops and gives way to a fall. So far as my present experiments indicate, the optimum for " solution " seems to coincide with the optimum for " swell- ing." There is, moreover, an upper limit to the total amount of fibrin that goes into solution in a given volume of the solvent. Under given conditions one has quite as much albumin in " solution " after shaking a mixture for two or three hours as after two or three days. In a given concentration of add {or alkali) the amount of fibrin that " dissolves " is markedly decreased through the addition of any neutral salt. With a progressive increase in the concentra- tion of the salt there is a progressive decrease in the amount of fibrin dissolved. But the character of the salt is not immaterial. When equimolar solutions of different salts are compared, some act more powerfully than others, but on the basis of my exper- iments as thus far carried out it is unsafe to state definitely the order in which the various salt radicals affect the " solution " of the solid gel. The order seems to be identical with that in which they affect the swelling of fibrin. Monovalent radicals are, as a group, less powerful in decreasing the " solution " of fibrin in an acid (or an alkali) than are bivalent ones, and these than trivalent radicals. What has been said will be rendered clearer by introducing a few typical experiments. Fig. 120 indicates the general way in which these experiments were performed. Weighed amounts of powdered fibrin were introduced into measured volumes of various solutions contained in Erlbnmeyer flasks which were then placed in a shaking machine and shaken for various periods of time. At the expiration of this time the fibrin was allowed to settle, and the supernatant liquid was decanted off into a filter-Uned funnel and received into a second flask. After stirring the filtrate, a measured volume was taken and the amount of " dissolved " albumin contained in it determined 436 OEDEMA AND NEPHRITIS quantitatively through precipitation with phosphotungstic acid ' and measurement of the heights of the precipitate either in the graduated Esbach albuminometer tubes or cahbrated test-tubes. As the experiments are purely comparative in character I have contented myself in these pages with simply photographing the results of a few of such as have a direct bearing upon our subject. Figure 120. Experiment 68. — 0.5 gram of powdered fibrin was shaken up for five hours in each, of the following solutions : 1. 50 cc. n/125 HCl 2. 50 cc. n/80 HCl 3. 50 cc. n/50 HCl. 4. 50 cc. n/25 HCl. 5. 50 cc. n/10 HCl. 6. 50 cc. H2O. The appearance of the fibrin in each of the flasks at the end of this time is shown in Fig. 120. In the first three fiasks (1, 2, 3) there is progressive increase in the swelling ,of the fibrin with the progressive increase in the concentration of the acid. Beyond this point (flasks 4 and 5) there is a decrease in the swelling in spite of the further increase in the concentration of the acid. The least amount of swelling is noted in flask 6, which contains water only. The solution of the flbrin is indicated in Fig. 121. From left to right these tubes correspond with ' The phosphotungstic acid reagent had the following composition: Phosphotungstic acid, 100 grams. Sulphuric acid (sp.gr. 1.84), 100 grams. Water enough to make 1000 cc. NEPHRITIS 437 the flasks of Fig. 120. No precipitate of albumin is seen in the tube on the extreme right, indicating that the fibrin did not go into solution appreciably in the water (neutral reaction). All the remaining tubes show a precipitate of albumin. FlGtTRB 121. Experiment 69.— 0.5 gram of powdered fibrin was put into each of the following solutions and shaken for five hours: 1. 10 cc. n/10 HCl+40 cc. H^O. 2. 10 cc. n/10 HCl+40 cc. m/8 NaCl. 3. 10 cc. n/10 HCl+40 cc. m/6 NaCl. 4. 10 cc. n/10 HCl+40 cc. m/4 NaCl. 5. 50 cc. H2O. The amount of albumin that went into solution is indicated in Fig. 122. No precipitate is seen in the tube on the extreme right (water). Most albumin is found in the first tube (pure acid solution). It is evident that the presence of the sodium chlorid reduces the amount of the albumin that goes into solution. The amount of this reduction is the greater the higher the concentration of the salt. ExPEraMENT 70. — 0.5 gram of powdered fibrin was placed in each of four flasks containing the following solutions and shaken for 5 hours: 1. 10 cc. n/10 HCl+40 cc. H2O. 2. 10 cc. n/10 HCl+40 cc. m/8 Na^SOi. 3. 10 cc. n/10 HCl+40 cc. m/8 MgS04. 4. 10 cc. n/10 HCl+40 cc. m/8 CuSOi. 438 (EDEMA AND NEPHRITIS Figure 122. Figure 123. NEPHRITIS 439 After filtering, the amount of albumin dissolved in the supernatant liquid found above the fibrin in each of the flasks was determined by- mixing 20 cc. of filtrate with 14 cc. phosphotungstic acid. The result is shown in Fig. 123. As is readily apparent, each of the salts markedly reduced the amount of albumin that was dissolved. Figure 124. Experiment 71. — 0.5 gram of powdered fibrin was introduced into each of five flasks containing the following solutions and shaken for five hours : 1. 10 cc. n/10 HCl+40 cc. m/8 sodium acetate. 2. 10 cc. n/10 HCl+40 cc. m/8 sodium nitrate. 3. 10 cc. n/10 HCl+40 cc. m/8 sodium sulphate. 4. 10 cc. n/10 HCl+40 cc. m/8 sodium citrate. 5. 10 cc. n/10 HCl+40 cc. H2O. The relative amounts of albumin found dissolved in each of these mixtures at the end of this time are indicated ia Fig. 124. As is again evident, most albunun was dissolved by the pure acid solution. Each of the salts decreased through its presence the amount thus dissolved. (b) Gelatin. — What has been said regarding the " solution " of fibrin holds almost word for word for the " solution " of gelatin. The best commercial gelatin shows some solubility in water. The commercial product, as is well known, has a decidedly acid reaction. When, instead of being placed in water, gelatin is dropped into solutions of acids (or alkalies) this solubility of the 440 (EDEMA AND NEPHRITIS (commercial) gelatin is greatly increased. The presence of neutral salts in the acid (or alkali) solution decreases the amount of the gelatin that will go into solution. As in the case of fibrin, we note here again a progressive decrease in the amount that " dissolves " with every increase in the concentration of the added salt. With a given concentration the amount of such a decrease varies with the salt employed, and here again it seems that monovalent salt radicals do not as a group decrease the " solubility " of the gelatin as much as bivalent, or these as much as trivalent ones. The following experiments may serve in illlistration of what has been said: Experiment 72. — The following solutions were prepared: 1. 100 cc. H2O. 2. 100 cc. n/1000 HCl. 3. 100 cc. n/500 HCl. 4. 100 cc. n/200 HCl. 6. 100 cc. n/100 HCL 6. 100 cc. n/75 HCl. 7. 100 cc. n/50 HCl. 8. 100 cc. n/40 HCl. 9. 100 cc. n/30 HCl. Five leaves of dry gelatin, each measuring 85 by I5 em., weigh- ing altogether 0.5 gram, and obtained by cutting them out of the central portions of the large gelatin leaves that are obtained com- mercially, were dropped into each of these solutions. From time to time the dishes containing the solutions with their gelatin leaves were agitated so as to keep them from adhering to the sides, and aid the solution of the gelatin. All the vessels were treated exactly aUke. The degree of solution of the gelatin after twenty-eight hours in these various solutions is indicated in Fig. 125. As the photograph shows, least gelatin is dissolved in the pure water. With the increase in the concentration of the acid there is a progressive increase in the amount of dissolved gelatin, but only up to a certain point, after which it falls in spite of the continued further increase in the concentration of the acid. ExPEBiMENT 73. — In the manner just described, 5 leaves of dry gelatin, weighing in toto 0.5 gram, and of the same surface were placed in each of the following solutions : 1. 100 cc. H2O. 2. 15 cc. n/10 HCl-l-85 cc. H2O. 3. 15 cc. n/10 HCl-l-21 cc. 2/m NaCl-|-82i cc. H2O. 4. 15 cc. n/10 HCl-l- 5 cc. 2/m NaCl-f-SO cc. H2O. 5. 15 cc. n/10 HCl-hlO cc. 2/m NaCl-|-75 cc. H2O. 6. 15 cc. n/10 HCl-t-15 cc. 2/m NaCl-|-70 cc. H2O. NEPHRITIS 441 The relative degrees of solution of the gelatin after a residence in these mixtures of eighteen hours is indicated in Fig. 126. The addition of the salt has decreased the amount of the gelatin that goes into solu- tion in the hydrochloric acid, and this the more the higher the concen- tration of the added salt. Figure 125. Experiment 74. — Five leaves of dry gelatin weighing altogether 0.4 gram and having the same surface were placed in each of the fol- lowing solutions : 1. 15 cc. n/10 HCl-l-85 cc. 2. 15 cc. n/10 HCl+lO cc. 3. 15 cc. n/10 HCH-10 cc. 4. 15 cc. n/10 HCl+10 cc. 5. 15 cc. n/10 HCH-40 cc. +45 cc. 6. 15 cc. n/10 HCl-l-40 cc. 7. 15 cc. n/10 HCl+10 cc. H2O. m/1 sodium acetate+75 cc. H2O. m/1 sodium chlorid +75 cc. H2O. m/1 sodium nitrate +75 cc. H2O. m/4 disodium hydrogen phosphate H2O. m/4 sodium sulphate+45 cc. H2O. m/1 sodium citrate +75 cc. H2O. The relative amounts of gelatin dissolved in these various solu- tions after the gelatin leaves had with occasional agitation remained in them for 19^ hours are indicated in Fig. 127. As is readily evident, each of the salts decreases by its presence the amount of gelatin dis- solved in the acid solution. The bivalent and trivalent acid radicals are more powerful in this respect than the monovalent ones, with the exception of the acetate. The intermediate position taken by this radical, in this matter of the solution of the gelatin, corresponds with the intermediate position occupied by this same radical in the swelling of the colloid under similar circumstances. 442 CEDEMA AND NEPHRITIS Figure 126. Figure 127. NEPHRITIS , 443 These experiments show that the " solution " of two typical emulsion colloids (protein gels) is intimately connected with the character of the medium surrounding them. Acids (and alkalies) favor their solution, while various other substances (notably salts) either do not affect it at all, or when present in conjunction with an acid (or alkali) depress the amount that would have been " dissolved " if the acid (or alkali) had been present alone. Let us now recall that what lies between the urine on the one hand and the blood on the other (the kidney) is made up physico-chemically of just such colloid gels as these under dis- cussion. Furthermore, we learned above that under normal circumstances, in " health," in other words, conditions are such in the kidney that this gel state is maintained. But in nephritis the acid content of the kidney is increased, in con- sequence of which conditions are offered which permit these protein colloids to pass over into the sol state and so escape with the urine. Albuminuria results whenever some or all of the colloid gels that constitute the urinary membrane go into " solution " in the urine, and this is made possible under the same conditions which permit fibrin or gelatin gels to " dissolve " in water. We shall find further evidence in support of this belief as we proceed. If it is true that albuminuria represents merely a " solution " of the kidney proteins in the urine, if, in other words, it does not come from the blood (except in that indirect way in which the proteins of any cell come originally from the blood), then albu- minuria cannot be that strange and specific thing which as clin- icians we are likely to think it. Any cell must, under conditions similar to those existing in the kidney when this is nephritic, be capable of serving as a source of albumin to a surrounding liquid medium, and so be capable of being responsible for a state which in the kidney goes by the name of " albuminuria." A little thought will show that such actually is the case. Every worker in the biological sciences is familiar with the ancient fact that " dead " organisms allow the escape of protein from them. A frog or fish hving in its aquarium does not impart a protein reaction to the water. But let it die and in a few hours the previously clear water gives a positive result when tested for albumin, and this reaction becomes the more intense as time goes on. What happens is that after death the tissues 444 (EDEMA AND NEPHRITIS develop the familiar postmortem acids and under their influence some of the proteins now go into " solution " in the surrounding medium. Such " solution " occurs whether the organism is a simple or complex one. Only when a simple organism possessing no circulatory system thus serves as a source of albumin then we can no longer talk about a " filtering " off of albumin, an " increased permeability of vessel walls," etc., — as we should not in the discussion of the albuminuria of nephritis. But we need not wander so far away from the mammals, or in fact the living animal itself, in order to show that " albu- minuria " is not the specific thing we think it. As surgeons well know, the normal intestinal juices scarcely yield an albumin test, yet the fluid contained in a strangulated hernia or a volvulus is rich in albumin. Here the interference with the circulation to the gut, produced through the strangulation or the twist, has placed a section of the bowel in ^ state of lack of oxygen; it develops in consequence an abnormally high acid content, and so some of the proteins of the gut wall go into " solution " — in other words, we get in the bowel what in the kidney is called albuminuria. Analogous conditions come to pass in the parenchymatous organs. When these are placed under circumstances which lead to an increase in their acid content, a state analogous to the albu- minuria of the kidney results. The lymph coming from a muscle that is made to work hard has a higher albumin content than that coming from this same muscle when at rest, and when the circulation through the hver is impeded (I would say oxygen supply through the hepatic artery is interfered with), either through ligation of the inferior vena cava or obturation of the thoracic aorta, the albumin content of the lymph coming from this organ begins to rise, as E. H. Starling ^ has clearly shown. In glau- coma the albumin content of the fluid of the anterior chamber rises, and in oedema of the brain and cord the cerebro-spinal fluid shows a more than usual amount of protein.^ 1 E. H. Starling: Jour. Physiol., 16, 224 (1894); 17, 30 (1895); see also Bayliss and Stabling: ibid., 16, 159 (1894). ''Edmund M. Baehr: Personal communication. NEPHRITIS 445 3. Critical Remarks It has been argued by some of my critics that this theory of albuminuria fails because the amounts of albumin given off by a kidney may be so large that were it all due to solution the whole kidney would be lost in a short time. Before this plausible objection is accepted it is well to consider the following rather obvious facts. In the first place, the albuminuria which we are discussing and the mechanism of which is a matter of debate is only that which is observed even though the blood vessels are intact. No one questions the vascular origin of the protein derived from frankly ruptured and oozing blood vessels or from red and white blood corpuscles which have escaped through the substance of the kidney into the urine and died there (dia- pedesis with subsequent " solution " of the escaped cells). We are likely to find the largest amounts of albumin without gross blood vessel lesions in the earlier stages of the acute types of nephritis, say in the pregnancy or other toxic nephritides. Suppose we choose a figure high enough to suit everybody and say that 20 grams of albumin are being given off to the liter of urine. Such kidneys are not likely to be secreting more than 100 or 200 CO., but suppose they are putting out 500. Even so, only 10 grams of albumin are being lost daily and we have 300 or more grams of kidney tissue to work on. But such albuminurias are neither common nor do they last long — the patient either gets over them in a very few days or dies. In the chronic types of nephritis we encounter no such figures. But even those that are observed cannot without revision be credited to that essential albuminuria which alone needs dis- cussion. Thus, in chronic interstitial nephritis associated with vascular disease the blood vessels in the kidney commonly rupture and bleed as they do elsewhere in the body; hemorrhage by diapedesis occurs in all types of kidney disease; and leucocytes not infrequently wander into the kidney tissues and through them into the urine. If the urine has not the right salt concen- tration, or is a little acid or alkaline, all these cellular elements are destroyed and the dissolved protein from this source is added to the exuded blood plasma, and both together are added to that derived from solution of the kidney itself. In this way the later figure may be pushed to any height, but to say that such an 446 CEDEMA AND NEPHRITIS amount of albumin has come from the kidney itself may suit a critic, but it is wrong. Aside from the fact that if a patient lost a gram of protein daily it would take months to destroy enough of his kidney substance to make him aware of it, this presupposes that a kidney has no regenerative powers, which, as a matter of fact, it has equally with other parenchymatous organs. A testicle, for instance, produces enough sperm daily to total several times its own weight in a year, and yet at the end of that time it has not disappeared. We need in this place to consider also an editorial criticism of the Journal of the American Medical Association.^ Its head- ing " A Controverted Theory of Nephritis " has been chosen a little broadly, for the paragraph itself comments only on the work of G. Salus,^ who in connection with his serological studies touches on the solution theory of albuminuria. Salus found that he could develop a precipitin for human blood serum by using albuminous urine as an antigen. He concludes cor- rectly from this, that the albumin in his urines contained blood proteins. But then no one has ever disputed this fact, for hem- orrhage through gross rupture of the blood vessels and by dia- pedesis is common in all types of nephritis. On the other hand, he found that he could get no response (more accurately stated, but one in ten trials) when he added the antiserum prepared from albuminous urine to a solution of tissue proteins extracted from the kidney. He cites this as evi- dence against the presence of dissolved kidney protein in the urine. In the face of the fact that it is difficult to prepare a specific antiserum even when kidney substance is "used directly, such findings are hardly conclusive. The colloid chemists, more- over, know how alterable in consequence of mere laboratory handling are the reactions of proteins, and so some argument will be necessary to make those of solid organs and of organ extracts synonymous in their minds. Salus himself recognizes these difficulties but the editorial writer seems unaware of them. 1 Editorial: Jour. Am. Med. Assoc, 62, 1971 (1914). ''G. Salus: Biochem. Zeitschr,, 60, 1 (1914). NEPHRITIS 447 VI THE MORPHOLOGICAL CHANGES IN THE KIDNEY 1. Introduction Anyone who has on the one hand busied himself with the clinical, or as we might better say, the biochemical, aspects of nephritis, on the other with the morphological aspects of this same problem, as this has been developed for us during the last two or three decades, must be struck not alone by the fact that the two have grown up practically independently of each other, but that they have made but sUght effort to find common ground. As a matter of fact, when we attempt to find a connection between the comparatively simple biochemical characteristics of nephritis and the elaborate morphological analyses of the organs from patients who have chnically shown the biochemical marks of a nephritis, this is at first sight not easy. Even if we ignore the fact that much of that which is supposed to characterize nephritis morphologically has nothing to do with the albuminuria, the changes in the secretion of water, the changes in the secre- tion of dissolved substances, etc., which are the distinguishing marks of a nephritis biochemically, there still remains an apparent lack of connection between the facts, to which any clinician or pathologist will testify, namely, that individuals may die of an acute Beight's disease and show surprisingly little macroscopic or microscopic change in the kidney, while others, never affected with any symptoms referable to the urinary system, may show on autopsy the infant-sized kidneys of chronic interstitial nephritis. And yet if we will but free our minds from the erroneous con- clusions to which the temptations of elaborate fixing and stain- ing methods and high power microscopes have led us, it is an easy matter to see that all the morphological changes that occur in a kidney, the seat of an acute or chronic nephritis, are fun- damentally simple in character, and that they are easily brought into connection with the clinical manifestations of the disease. We will discover at the same time that the essential morpholog- ical changes of acute and chronic nephritis were recognized and a satisfactory classification of the nephritides on morphological 448 (EDEMA AND NEPHRITIS grounds was made decades ago, more especially by Weigeet,* and that a classification of the nephritides on the basis of patholog- ical physiology brings us in these modem days back to yet older teachings, to those of F. T. Frerichs,^ for example, who regarded all the nephritides to be in essence the same. 2. Classification of the Nephritides. Correlation of the Mor- phological Changes in the Kidneys with Some Clinical Manifestations There is but one kind of nephritis — parenchymatous nephritis. How could there be any other? It is the function of the kidney to yield a secretion which we call urine, and this function is exhibited by the parenchyma of which the kidney is composed. A disturbed kidney function can come to pass only as the parenchyma has been involved. Histological examination shows that the parenchyma is not everywhere the same, and it is pre- sumable therefore that the different parts play different roles, but since the physiologists have not yet settled what are these differences in the functions of the glomeruh, the convoluted tubules, the collecting tubules, etc., a more detailed classifica- tion into glomerular, tubular, etc., types is, to say the least, premature. We know not a single experimental procedure or pathological process which involves exclusively only one of these structures, and we cannot in consequence do more than speculate on their function. It is evident that a pathological process may involve a whole kidney, in which case we may speak of a generalized parenchy- matous nephritis, or it may involve only smaller or larger patches, leaving healthy kidney between, in which case we may speak of a focal or spotty parenchymatous nephritis. Either of these types may, of course, be acute or chronic. If the agencies attack- ing the kidney are removed, and if the damage done the paren- chyma has not been too great, then, evidently, the involved cells may recover, in other words, the normal state of the kidney be re-established. Expressed more technically, recovery occurs if the changes induced in the kidney remain of a reversible * The most accessible of Weiqert's papers on nephritis appear in Vir- chow's Archiv during the years 1860 to 1875. * F. T. Fbbrichs : Die Bright'sche Nierenkrankheit, Braunschweig (1851) . NEPHRITIS 449 type. But if for any reason, say through prolonged or par- ticularly intense action of the agencies producing the nephritis, irreversible changes occur in the kidney parenchyma, then the involved cells die and are absorbed. If the defect is not or can- not be made good by regeneration of new cells, then that portion of the' kidney is gone and in its place may appear nothing more than a httle scar tissue. All these possibilities of injury with recovery, or injury with death and loss of the involved part may and do occur in nephritis whether it involves a whole kidney or only a patch in it. Let us consider first the generaUzed parenchymatous type of nephritis consequent, say, upon an acute intoxication of some kind. It is possible, first of all, for such a kidney to recover entirely. But if such a fortunate ending is not attained, death of the whole kidney is not the only alternative. Larger or smaller pieces may die and be replaced by connective tissue while the remainder of the kidney recovers. There will ultimately result then a kidney which as far as it goes contains normal parenchyma, but in diminished amount, a so-called secondarily contracted or sclerosed kidney, a so-called chronic interstitial nephritis secondary to generalized parenchymatous nephritis, the " small red kidney " of the pathologists. Diagrammatically represented, the process may be illustrated by reference to Fig. 128. The rather uniform effect of a poison of some sort circulating through a kidney is illustrated by the black shading under a. If pieces of this kidney die and are absorbed while the remainder recovers we get ultimately the secondarily contracted kidney represented under b. If the parenchjTnatous nephritis is of the focal or spotty type as represented diagrammatically in Fig. 129 a, the changes in the parenchjTna may again be either reversible (curable) or irreversible (incurable). If they are irreversible the involved patches will again die and be replaced by connective tissue. The ultimate picture is shown in 6 of Fig. 128, and again approximates that previously described. Healthy kidney substance remains to make up the bulk of the kidney which, however, is diminished in amount and has connective tissue scattered through it. We have again a " small red kidney," in other words, again a chronic interstitial nephritis. But because some have assumed — falsely as we shall see — that the connective tissue was laid down first 450 OEDEMA AND NEPHRITIS and that the death and disappearance of parts of the kidney occurred later, this pathological entity has been called a primarily contracted kidney or a primary chronic interstital nephritis. From a morphological classification point of view this kidney is about the same as the secondarily contracted kidney previously dis- cussed. It will make matters a little clearer if we try at once to con- nect up this- simple classification with the clinical aspects char- acteristic of the described kidney states. If a poison capable of inducing nephritic changes circulates in the body of a patient, let us say the toxins of a scarlet fever, the toxins of a pregnancy nephritis, or bichlorid of mer- cury, it will, in passing through the kidney, tend on the whole Figure 128. Figure 129. to affect the entire kidney at once and more or less uniformly. For this reason examination in the course of a surgical opera- tion or postmortem reveals the swollen kidney of the so-called generalized parenchymatous nephritis. Since the whole kidney is involved, we encounter under these circumstances the greatest interference with function and therefore the greatest decrease in water output — maybe to the point of complete suppression. At the same time such urine as is secreted is heavily charged with albumin and casts. If the causes operating to produce the nephritis pass away, the kidney recovers and so the urinary output rises again and casts and albumin diminish, all, maybe, in the space of a few days. But if pieces of the kidney die, the evidences of destruction in the kidney as betrayed by casts and albumin may last longer, but even here if the causes operating to produce the nephritis are ultimately overcome, such portions NEPHRITIS 451 of the kidney as are left may recover and the patient with his secondarily contracted kidneys — his morphologically chronic interstitial nephritis — may live indefinitely. The reason for this resides in the fact that one-quarter of our total kidney substance is easily sufficient to maintain us in health and happiness, and if this amount has been saved the patient need show no signs or symptoms which will allow a diagnosis of his true condition. An autopsy or examination of the kidneys in the course of a surgical operation may offer the first occasion for recognizing the kidney state. It is hardly possible for a soluble poison to enter the kidney and not affect it rather uniformly. A circulating poison can hardly, therefore, give rise to a spotty or focal parenchymatous nephritis. For such we need a spotty cause. Such is offered, for example, by the changes of vascular disease.^ We shall return to this problem later, but it may be emphasized here that all evidence shows the vascular disease to be primary and the kidney disease secondary to it.^ Vascular disease attacks particularly the smallest blood vessels. (When it attacks the larger blood vessels it does this by attacking their vasovasorum.) Since the blood vessels of the kidney do not escape, this organ may, of course, be affected. In consequence of the thickening of the vascular walls and the oft-accompanying thrombotic changes, one piece of the kidney after another is deprived of its blood supply. As this happens they show the changes character- istic of nephritis, and since the arteries involved are end arteries, the kidney changes are largely irreversible, and piece after piece of the kidney dies and disappears while connective tissue takes its place. The portions of kidney involved in this localized de- struction of kidney parenchyma show all the signs characteristic ^ Under this caption I include all the pathological processes capable of affecting the blood vessels, no matter what their assumed causes, be they frank infections as in syphilis, or " degenerations " aa in atheroma, arterio- sclerosis, etc., popularly regarded as consequent upon attack from that old guard, alcohol, hard work, gout, auto-intoxication, and a meat diet. For further remarks on vascular disease, see pages 482, 494 and 499. ' See in this connection, Hauch's beautful x-ray pictures of the blood vessels of healthy and diseased kidneys. In vascular disease involving this organ the lumina of the blood vessels becomes smaller and the vessels supplying a given area progressively less in number. When such changes are suflBciently advanced the involved kidney tissues die. Hauch: Fortschr. Rontgenstrahl., 20, 172 (1913). 452 OEDEMA AND NEPHRITIS of 'parenchymatous nephritis. Between these localized areas of parenchymatous nephritis the kidney tissue is healthy. When, now, we again remember that less than one-fourth of the total kidney substance is necessary for the maintenance of life, it is easy to see why a patient with chronic interstitial nephritis runs along in a fairly normal way. The destruction of the kidney occurs so very slowly that little albumin appears in the urine, and casts only in small numbers. So this patient may also die without ever having become conscious of his kidney state. If a diagnosis is made for him it is done very largely on the basis of findings referable to his vascular disease (palpable blood ves- sels, high blood pressure, cardiac hypertrophy) which as we shall see are not secondary to his kidney disease, but expressive of his vascular condition. We shall have occasion to return to all this later. For the present it is sufficient merely to emphasize the fact that the chronic interstitial nephritis associated with vascular disease is in essence also a parenchymatous nephritis — a slow-going but pro- gressive localized parenchymatous nephritis resulting in death and loss of the involved portions of the kidney and ultimately in a picture which is best described by calling it an atrophy of the kidney. The patient with chronic interstitial nephritis is, therefore, in the same position as an animal that has had its kidney substance progres- sively diminished in amount by successive operations and abla- tions of kidney parenchyma. The man who' has gone through life without marked signs or symptoms of kidney disease, who dies of other causes than kidney disease, and shows on the autopsy table what, as morphologists, we call chronic interstitial nephritis, is simply like the animal that has siiffered a great reduction in total kidney substance, but has not yet reached the physiological minimum compatible with life for that animal under the con- ditions under which it has to live. What is left of kidney paren- chyma to man or animal is still physiologically active and physi- ologically adequate. Such a biological contention finds its morphological support in the fact that the parenchyma of such (morphologically) chronic interstitial types of nephritis shows little or no change either macroscopically or microscopically (" small red kidney "). The presence of the connective tissue in the kidney is an accident; it is scar tissue, and whatever im- portance we may care to attach to it morphologically, this is NEPHRITIS 453 no more expressive of the physiological state of the kidney parenchyma that is left than the scar which repairs and sei'ves to reunite the ruptured ends of a muscle is any index of the physi- ological efficiency of that muscle. With this we have disposed of the apparent differences between parenchymatous nephritis and what is called chronic interstitial nephritis. While infections of the kidney are not ordinarily considered under the nephritides, they might as well be, for the result is the same, of course, whether the function of the kidney is impaired because toxins • are carried into it by the blood stream or they are manufactured on the spot by micro-organisms present in the kidney. The infections may give rise to either a generalized or focal type of nephritis. If tubercle bacilli, for example, or some of the pus formers are sown into a kidney, patches of nephritis result which give rise to albumin and casts in the urine in proportion to the amount of kidney involved. If sufficient healthy tissue remains between these patches the total urinary output need not be much changed. On the other hand, if the involved patches increase in size or become so numerous as to take up the major portion of each kidney, then albumin and casts must increase and urinary secretion diminish even to the point of complete suppression perhaps. If destruction of kidney tissue results with replacement by scar tissue these infectious cases also yield secondarily contracted kidneys (chronic interstitial nephritis). Let us complete this discussion by referring once more to Figs. 128 and 129. Under b of Fig. 128 is represented the atrophic remains of a healthy kidney which is called, morphologically, chronic interstitial nephritis. Such a kidney is, of course, as sub- ject to attack by any of the causes which may underlie a general- ized parenchymatous nephritis as is a normal kidney. When this occurs the normal or " increased " urinary output so often observed in the morphologically chronic interstitial nephritides gives way to a diminished one with many casts and much albumin. This is frequently the terminal picture in the chronic interstitial types of nephritis found associated with vascular disease and its accompanying cardiac hypertrophy and high blood pressure. The chronic interstitial nephritis is produced as already described. But while the patient is living on his remnants of kidney, his 454 (EDEMA AND NEPHKITIS heart muscle begins to fail or the sclerosis of the main arteries leading into the kidney reaches a fatal limit, and subject to the inadequate blood supply resulting from such the remaining kidney cells die— the " small red kidney " passes over into the " small gray " one. We shall find much evidence to support these simple views as we proceed. What will strike my readers at this time as the most obvious shortcoming in my insistence that the chronic interstitial types are also but parenchymatous nephritides, is my ignoring of the fact that with the frankly parenchymatous types we are likely to find associated a generalized oedema, while with some of the recognized types of chronic interstitial nephritis there goes no oedema, but an increased blood pressure and cardiac hyper- trophy. I am ignoring these things only temporarily, however. But even here let me point out what will be proved later, that these signs, while likely to be associated with these types of nephritis, are not secondary to the kidney disease as generally taught, but due to entirely different causes.^ Let us now discuss the morphological changes observed in the nephritic kidney seriatim. The pathologist accepts the fol- lowing as characteristic of the parenchymatous lesions whether they involve patches or the whole of a kidney. For their recog- nition no elaborate histological technique is at all necessary. 1. An increase in the size of the involved portions of the kidney, traceable on the examination of fresh, unfixed and unstained cells, back to an increase in the size of the individual cells and tissues composing the kidney. 2. A loss of the normal color of parts or all of the kidney which assume a less glistening, drier and more opaque (boiled) look. On microscopic examination this change is found to be associated with the appearance of granular substances in the cells of the affected portions of the kidney. This change in color, taken in conjunction with the increase in the size of the kidney, con- stitutes the " cloudy swelling " of the pathologists. 3. The appearance of blood corpuscles extravascularly. They may be found in the tissues of the kidney itself, or in the spaces about the glomerular tufts and in the uriniferous tubules. 4. Evidences of a falling apart of the kidney as a whole and of a disintegration of the individual cells of the kidney. Under iSee page 482. NEPHRITIS 455 this heading are grouped not only the gross destructive lesions observed in the kidney, such as the rupture of capillary tufts, but the separation of individual and groups of cells from their attachments in the glomeruli. Bowman's capsule and the urin- iferous tubules (formation of casts). This catalogue of morphological changes holds both for the acute parenchymatous nephritides and for the chronic forms. The chronic show all the changes of the acute with certain others added to them, notably a " fatty degeneration," and the develop- ment of a certain amount of scar tissue. As the last two take us into fields not immediately connected with our problem of nephritis, we shall not discuss them in detail. 3. The Changes in the Size and in the Color of the Kidney in Nephritis (Cloudy Swelling) ^ While we shall later find ourselves compelled to discuss these two changes in the kidney separately, we will first take them up together because it is in this form, under the caption of cloudy swelling, that they have been chiefly discussed by the pathologists. As is familiarly known, we are indebted to Rudolph Vibchow not alone for a first clean-cut description of this cloudy swelling as it occurs in the kidney (and other parenchymatous organs), but for a first attempt to analyze its nature. Virchow held cloudy swelling to be " a kind of acute hypertrophy with tendency to degeneration," a phrase which has found its way into even our most modem text-books of pathology. But while such a phrase still serves many as a satisfactory characterization of the condition from a biological standpoint, it means nothing, of course, from the standpoint of its physico-chemical analysis. Toward the physico-chemical analysis of cloudy swelling Vikchow con- tributed the important suggestion that the cause of the granule formation in the cells is due to a change in their albuminous constitution. He based this conclusion upon the fact that the granules are soluble in acids and alkalies, and not in ether, thereby distinguishing them from fat deposits in the cells (fatty degenera- 1 Martin H. Fischer: Kolloid Zeitschr., 8, 159 (1911). 456 (EDEMA AND NEPHRITIS tion) which at times mimic in general appearance cells affected by cloudy swelhng. For the increase in the size of the cells ViRCHOw gave only the biological explanation of an " increased irritation " of the affected cells, caused, for example, by the products of an infectious disease, in consequence of which they were made to take up " excessive amounts of nutrient material." That cloudy swelling represents a change in the albuminous constitution of the cell seems never to have been questioned. Eduard Rindfleisch 1 accepted this belief and, moreover, expressed himself of the opinion that cloudy swelling was " pass- ive " in its nature and due to " a kind of corrosive action in con- sequence of which the albuminous matters, held in solution by the protoplasm, undergo coagulation and become visible as minute granules." In 1882 Julius Cohnheim ^ subjected ViR- CHOw's teachings to a rigorous critique. That the process of cloudy swelling involved the albuminous constituents of the cell he did not question, but he perpetuated a conclusion (erroneous as we shall see) of Virchow, when he wrote: " Of course we must deal here with a protein that is different from that which is nor- mally present in the cell protoplasm ... as we could not other- wise account for the optical difference." But Cohnheim, too, expressed the possibility of cloudy swelling representing " a spontaneous precipitation in solid form, or the coagulation of a previously fluid protein." What underlies such a change in the albuminous constitution of the cell, he did not attempt to say, but he showed very conclusively that the causes proposed by earlier writers were questionable if not entirely inadequate. Thus he showed that the fever accompanying the various infections liable to be accompanied by a cloudy swelling could not by itself be the cause of the change, by calling attention to the well-known fact that cloudy swelling may be absent in cases that have run a high fever, or present in conditions not associated with an abnormal rise in temperature. In such a half -hypothetical state did the subject of cloudy swelling remain until 1901, for in spite of various discussions of the subject, no clear-cut advance was made either toward 'Eduard Rindfleisch: Pathological Histology. Translated by Baxter, 30, London (1872). 2 Julius Cohnheim: AUgemeine Pathologie, Zweite Auflage, 1, 662; 2, 570, Berlin (1882). NEPHRITIS 457 defining more precisely what cloudy swelling is, nor yet in dis- covering a something common to all conditions associated with cloudy swelling, which might justly be regarded as its funda- mental "cause." At this time H. J. Hamburger ^ reported a series of observations on isolated liver, kidney and spleen cells which served to estabhsh more firmly what can justly be regarded as little more than lucky speculation on the part of the earher writers. Hamburger applied to these cells observations pre- viously made on red and white blood corpuscles. In a study of the latter he had found that various acids, including carbonic acid, bring about an exchange of substances, including water, between the red and white blood corpuscles and the serum in which they are contained. Under the influence of acids all these cells take up water from their surroundings. He paralleled this with the findings of previous observers that, in fevers of the most varied origins, acids are produced and the " alkalinity " of the blood is reduced, and so concluded that in this acid production resided the cause for the enlargement of the cells in cloudy swelling. He debates why acids bring about the enlargement and con- cludes that changes leading in the aggregate to an increase in the osmotic pressure of the cell contents are mainly respon- sible. Hamburger then points out that the white opaque appearance of isolated kidney, liver and spleen cells exposed to dilute acids is identical with that of cells affected with cloudy swelling and discovered postmortem. Cells treated with an acid are studded with granules, as are the cells showing a cloudy sweUing that are found postmortem, and to prove that the gran- ules are similar in character in both, and represent albumin precipitates, he calls attention to the fact that the granules which he has made appear through a weak acid dissolve again as the acid concentration is increased. Hamburger found an analogue of the production of the granules in the isolated paren- chyma cells in the precipitation of albumin from a diluted blood serum when an acid is added to this. The first great value of Hamburger's studies resides in the fact that he has detailed experiments which show that all the necessary elements for cloudy swelling reside in the parenchyma ■^ H. J. Hamburger: Osmotisoher Druck und lonenlehre, 3, 49, Wies- baden (1904), where references to his earlier articles may be found. See also Karl Landsteiner: Zeigler's Beitrage, 33, 237 (1903). 458 CEDEMA AND NEPHRITIS cells themselves, and that he has pointed out that, what is added through an infectious disease (or, as we might say, in order to make our contention more pointed, any condition which is capable of inducing a nephritis) may be nothing more than a little acid.- This simple reasoning of Hamburger does away with the bio- logical terminology that has so long been apphed to the subject of cloudy swelling, and renders possible an attack upon the problem in the light of the simpler concepts of physics and chemistry. Since Hamburger's work I know of no contributions to the subject of cloudy swelling which have either questioned the correctness of his view, that the increased absorption of water by the cell affected with cloudy swelling represents an osmotic phenomenon, nor yet any which have adduced further evidence in support of the protein precipitation idea of the granule forma- tion in this condition. As the subject is intimately connected with our problem of the morphological changes occurring in nephritis, I felt that it could to advantage be restudied, espe- cially since the acquisitions of colloid chemistry — the chem- istry of the very substances of which the kidney is composed — have furnished us with data and theoretical deductions that are of immediate applicability in the analysis of this problem. By utilizing these we shall find ourselves in a position to give a simpler physico-chemical explanation for the increased water absorption by the tissues in cloudy swelhng than is contained in the unsatisfactory osmotic explanation of this part of the phenomenon, and at the same time we shall learn how the clouding ' of the parenchymatous organs follows the same laws as the pre- cipitation of such a simple colloid as casein. In this way we shall find a ready explanation of the first two of the morphological changes in the kidney catalogued above and characteristic of nephritis, namely, the increase in the size of the parenchymatous elements and their change in color. At the same time we shall find that both arise from the same cause, namely, the abnormal production and accumulation of acid in the kidney, which we have previously tried to show to lie at the base of all the nephritides. NEPHRITIS 459 §2 We shall first describe a series of observations on the arti- ficial production in excised kidneys of the changes characteristic of nephritis (production of cloudy swelling) which will prove themselves of service in the further analysis of our problem. The methods employed in these experiments were the same throughout. The kidneys of healthy, freshly killed rabbits and guinea pigs were used, which after being sliced were dis- tributed into bowls each containing 100 cc. of the necessary solutions. As it is impossible to give absolute values to the various grades of grayness and opacity observed in the different solutions, one can, in the description of the findings, only com- pare the appearance of a tissue in one solution with that of a similar piece in a different solution at the same time. The gen- eral conclusions from a long series of experiments may be sum- marized as follows: (a) When slices of fresh kidney are dropped into distilled water they slowly swell and at the same time become gray. A tone of gray that is readily distinguishable from the color of the normal organ appears over the cut surface some three or four hours after being dropped into the water. This gradually increases in intensity imtil, twenty-four hours after the begin- ning of ithe experiment, the tissues look decidedly gray. For a day or two longer this may continue to increase in intensity, but the change from the first twenty-four hours is not very marked. As the tissue becomes gray it shows an acid reaction to Htmus, and this acid production in even a small piece of tissue may be sufficiently great to impart an acid reaction to the sur- rounding fluid. (b) The pieces of tissue swell much more rapidly if they are placed in any dilute acid instead of in distilled water. This is shown in Fig. 130. A has simply been protected against evapo- ration. B has lain for an hour and a half in n/333 hydrochloric acid. The two pictures represent opposite faces of the same cut through the kidney. The tissues also become gray sooner in an acid solution than in distilled water. In n/200 solutions of lactic, formic, acetic, tartaric, hydrochloric, sulphuric, or nitric acids a decided cloudiness is visible in ten minutes after immersion. This cloudiness becomes gradually more marked. 460 (EDEMA AND NEPHRITIS After three hours, when the control in distilled water is just showing a grayness, the slices of tissue in the dilute acids are grayer than the controls appear the following day. The vari- ous acids show some difference in the intensity of the cloudi- ness that they produce, but this is so much a function of their concentration and the time, that a table of their relative effective- ness cannot be given to advantage. After another two hours the tissues in all the acid solutions are intensely gray. The control in pure water is about as gray as the tissues placed in the dilute acids were after ten minutes of immersion. On the following day an ultimate degree of grayness (a typical "boiled " appear- ance) is shown by all the organs in the dilute acids. Figure 130. Speaking generally, it may be said that when the effects of different concentrations of the same acid are compared, the cloudiness develops the more rapidly the greater the concentra- tion of the acid. So far as intensity is concerned, there is, how- ever, little difference. In the end every acid gives the tissues a boiled appearance. With different concentrations of nitric acid I found the boiled appearance after a ten-minute immersion in n/10 normal acid. In n/20 acid the same appearance was attained in an hour; in n/40, n/100 and n/500 in two to three hours. What has been said of nitric acid holds true in general for all acids, though there are more or less specific differences with the different acids both so far as rapidity of development and intensity of the cloudy swelling is concerned. Acetic acid NEPHRITIS , 461 is particularly interesting. With increasing concentrations of the acid there is first an increase in the rate and (in units of time)/ in the intensity of the cloudiness produced. If we observe closely, this is seen to be followed with increasing concentra- tions of acid (above n/10 acetic acid) by a stage in which the cloudiness is less than in lower concentrations. To see these successive changes one must observe especially the superficial portions of the tissues. A second clouding can now be obtained by changing to one of the " strong " acids (nitric, sulphuric, or hydrochloric) of the same or of a higher normality than that of the acetic acid which has brought about the disappearance of the first clouding. This change from cloudiness to clear- ness and back again to cloudiness, with progressive increase in the concentration of an acid, can be followed particularly well under the microscope (see (g) below). But even in the sections of tissue kept in the " stronger " acids can two such regions of cloudiness separated by one of clearness be discerned. I found, for example, that the marked cloudiness of slices of kidney, which had been kept for 1| hours in concentrations of nitric acid up to n/200, disappeared when the surface of the organ was touched with the ordinary weak acetic acid of our lab- oratory reagents, to reappear when dilute nitric acid was sub- stituted for it. The cloudiness of the tissues obtained in any of the acids listed above, if developed in not too high concentrations (below n/200), can also be made to disappear if the tissues are placed in equinormal alkali solutions, or in alkali solutions of a higher concentration. (c) Through the addition of various salts the develop- ment of a cloudiness in any acid solution can be either retarded or hastened. So far as the absorption of water is concerned, all the salts have but one effect — they decrease the amount of the swelling in the acid solution. When to n/200 hydrochloric acid enough of various potassium salts is added to make their final concentration m/20, the following is noted. After ten minutes' immersion it is plainly evident that some of the salts are accelerating the effect of the acid in producing the devel- opment of the cloudiness, while others are inhibiting it. In an hour the differences are very marked. The sulphocyanate, iodid, bromid, and nitrate all increase the cloudiness, the first 462 (EDEMA AND NEPHRITIS named being the nriost powerful in this respect. Then comes the pure acid. Following this comes the chlorid, the acetate, the tartrate, and the citrate. After three to six hours of im- mersion the differences are still more striking. In the solutions containing the first-mentioned salts the tissues are " boiled " in appearance. In the pure acid the grayness is well marked. The tissues in the solutions containing the chlorid and the acetate lag somewhat behind the pure acid. In the tartrate a faint film is only just visible over the surfaces of the organs. The sections in the solutions containing the citrate look per- fectly normal. In fact, in the two last-named solutions the organs retain an almost normal appearance for two to three days. Similar results are obtainable by using sodium or ammo- nium salts in place of the potassium salts, or lactic, formic, or nitric acid in place of the hydrochloric, except that in the latter case the absolute rate at which any degree of cloudiness is obtained is not quite the same as in hydrochloric acid. (d) Various salts accelerate or retard the development of a cloudiness in sections of kidney placed in their pure solu- tions, in just the same way as they accelerate or retard the development of a cloudiness if an acid is added at the same time, only the rate of development and the absolute intensity of the cloudiness attained is less in the pure salt solutions than in mixtures of these with any acid. In all salt solutions the kidney slices swell less than in pure water. These findings are to be interpreted by noting that the excised tissues become acid, so that the tissues placed in the pure salt solutions are really in the same state as the tissues described in the preceding paragraph — • the tissues are really in an acid solution plus certain salts- only the concentration of acid is lower in this case than in the previously described experiments. (e) Alkalies do not produce a cloudiness of kidney parenchjona in any concentration. Sodium, potassium, ammonium, and calcium hydroxids were employed in concentrations up to n/33. The superficial layers of the tissue slices " dissolve " in the hydroxids, covering the tissues with a clear, gluey mass. After two or three days the tissues lose their bright normal color, but the grayness assumed is only slight. The slices of kidney swell just as they do in acid solutions. NEPHRITIS 463 (/) The addition of any salt to the solution of an alkali does not lead to any cloudiness of the tissues, though it markedly reduces the tendency of the superfiicial layers of the tissues to go into " solution," and the swelling of the tissue fragments as a whole. I have tried without effect the chlorids, bromids, iodids, nitrates, sulphates, sulphocyanates, acetates, tartrates, and citrates of sodium, potassium, and ammonium in conjunction with the hydroxids of sodium, potassium, and ammonium. I have also tried a few strontium and barium salts with these hydroxids and calcium hydroxid, employing all in such low con- centrations as to prevent the formation of precipitates, but I got no cloudiness of the immersed tissues. (g) The macroscopic changes observed in the kidney when immersed in water, various acids, or alkalies, in salt solutions or these in combination, show a series of interesting parallels microscopically. A perfectly fresh scraping from the kidney shows the cells to possess a fairly clear protoplasm in which lie but few granules. Even after the kidney cells have been kept for twenty-four hours (simply in their own moisture, and protected against evapora- tion by being covered) they show no change from this appearance. But as soon as water touches the cells, especially if the organ has been kept for twenty-four hours, or if they are placed in any very dilute acid, a grayish film is seen to develop macroscopically, and microscopically the cells are now found thickly studded with granules. This is the typical histological picture of the cloudy swelUng described in our text-books of pathology. If now, while such cells are being observed, a httle caustic soda is allowed to run under the cover slip, the cells as a whole are seen to swell, the granules to become fainter, then fewer, and finally to disappear entirely, and if enough alkali is added the whole goes into homogeneous " solution." The granules can also be made to disappear by the addi- tion of more acid; they form, for example, in very dilute acid, and disappear again if the concentration of this same acid is raised. Most interesting is the fact that this granular appear- ance can be made to come a second time by still further increas- ing the concentration of the acid. Acetic acid will not do this, but nitric acid will do it promptly. If strong nitric acid is used this second appearance of the granules is only a temporary affair, 464 CEDEMA AND NEPHRITIS for they again disappear as the whole tissue goes into " solu- tion." With the second appearance of the granules the cells undergo a marked shrinkage from the more swollen state attained previously, but this shrinkage, like the second appearance of the granules, is also only temporary, and the cell undergoes a final enormous swelling before being " dissolved." §3 How now are we to interpret these various findings, and what light do they bring us regarding the cause and the essen- -tial nature of those changes of hke character, which we observe in the kidney in nephritis and which lead to the increase in its size and to the change in its color. Our first attention must be dedicated to the increase in the size of the cells. H. J. -Hamburger recognized very clearly that the funda- mental cause for the increase in the size of the cells affected with cloudy swelling lies in the production of acid in them. As we have already learned, evidences of an abnormal produc- tion and accumulation of acid in the kidney occurs in every case of nephritis, and so we may make this condition, which we have already made responsible for the albuminuria, responsible for this increase in the size of the kidney also. But how does an abnormal acid content manage to bring about the increased water absorption which leads to the increase in the size of the cells (and so of the kidney as a whole) in nephritis? Hamburger answered this question by attributing an indirect effect to the acid, whereby this was assumed to increase the osmotic concen- tration within the cells. The enlargement of the cells in " cloudy swelling " represents an oedema of the affected cells, and this is most easily accounted for on the basis of the colloid constitu- tion of living matter. The serious objections that can be lodged against the widely accepted belief that cells repre- sent osmotic systems cannot be raised against the view that the (lyophilic or emulsion) colloids of the tissues and their state determine the quantity of water absorbed by a cell. As previously emphasized, the amount of water that such colloids (as represented by gelatin, fibrin, and serum albumin, for example) will absorb is enormously increased if any acid is present. This fact receives incidental illustration in Fig. 120. On this basis, NEPHRITIS 465 it is easy to parallel the absorption of water, and so the enlarge- ment of the cells of the kidney when affected by nephritis, with the increased amount of water absorbed, say by a gelatin cube or some fibrin particles, when instead of being placed in water they are placed in a dilute add of some kind. In the case of gelatin and fibrin, and similarly in the case of the experiments on excised kidneys, the source of the water for the increased swelling is to be found in the solutions surrounding these colloid structures; in the case of the nephritic kidney, in the blood and lymph streams passing through the organ. There is, within certain Hmits, an increase in the amount of the swelling of such protein colloids as gelatin or fibrin with every increase in the concentration of the acid surrounding them. On this basis we can understand the increase in the swelling of the kidney cells with every increase in concentration of the acid up to a certain point. When a certain optimal concentra- tion of the acid is exceeded, the colloid swells less than in weaker solutions (see Fig. 120). This furnishes a ready interpretation of the finding detailed above, that on substituting nitric acid for a weaker solution of acetic acid, kidney and liver cells are seen to shrink. Incidentally, it is worth while emphasizing that in the great rapidity with which such cells will give off and take up water, in changing from a medium of one concen- tration to another having a lower or a higher one, lies a powerful argument against the osmotic pressure idea of water absorption in cells. I have seen these cells pass from the swollen state, in a weak acetic acid solution, to the greatly shrunken state induced by nitric acid, and through a second swollen state into " solution " in less than two seconds. Equalizations of osmotic differences either through a movement of solvent, or of dissolved substance, do not occur with such velocity. We may now turn to a consideration of the changes in the color of the kidney in nephritis, and see how these become inter- pretable on the basis of the fact that in this condition an abnor- mal amount of acid is present in the kidney. The statements made above regarding the means by which a cloudiness can be produced in the parenchymatous cells of the kidney, or regarding its rate of development, or the means by which the intensity of such a cloudiness may be increased or decreased, have all of them parallels in the ways and means 466 CEDEMA AND NEPHRITIS by which protein may be precipitated from one of its " solu- tions," or such a precipitation be hastened or retarded. The development of a cloudiness in the kidney cells follows most closely the solution and precipitation of such a colloid as casein} Casein ^ is insoluble in water. It is soluble in dilute hydroxids, in which state it is electro-negative. It is in this state that we find the body proteins normally, as Wolfgang Pauli^ has shown. In the left-hand tube of Fig. 131 is shown the perfectly clear casein solution made by saturating an alkah (NaOH) with Figure 131. casein. When a dilute acid is added to such an electro-negative protein, let us say to the solution of casein in anj hydroxid, a precipitate of the casein is thrown down, as shown in the second tube. A similar precipitation of an electro-negative colloid 'This term is used in Hammarstbn's sense and corresponds therefore with the caseinogen of Halliburton. 2 For a discussion of the general properties of casein see O. Hammarsten: Physiological Chemistry, Translated by Mandel, New York (1914); E. Laqueur and O. Sackue: Hofmeister's Beitrage, 3, 193 (1903); W A Osborne: Jour. Physiol., 27, 398 (1901); T. B. Robertson: Jour. Biol. Chem., 2, 317 (1907); L. L. van Slyke and E. B. Hart- Am Chem Jour., 33, 461 (1905). » Wolfgang Pauli: Naturwissensch. Rundschau, 21, 3 (1906). NEPHRITIS 467 occurs when our sections of kidney are immersed in any dilute acid. The development of a cloudiness in tissues immersed in water is also to be regarded as a precipitation through a dilute acid, only in this case the tissues themselves produce the acid. Similar conditions hold in nephritis, when, in consequence of the abnormal acid content of the kidney, some of the protein constituents of the cells composing this organ are precipitated. As we have already found this same acid to be responsible for an increased swelling of the tissue colloids, it is easy to see how from the two there results, when water is available, the picture we designate " cloudy swelling." But our analogy goes further than this. If we continue to add acid to the reaction mixture in which our casein began to be precipitated, the precipitate becomes heavier (as in the third and fourth tubes of Fig. 131), but soon, with still further addition of acid, the casein begins to go back into solution. This is evident in tubes five, six, and seven, the last of which is again entirely clear (the white spot at the bottom of the tube in the photograph being a highlight). This is what we observe in the kidney cells when we note the cloudiness produced in a weak solution of any acid, or that found in the nephritic kidney on autopsy, to disappear on appljdng a stronger solution of the acid (say acetic acid) to the kidney. This macroscopic change has its parallel in the microscopic disappearance of existing granules in a cell, the seat of a cloudy swelling (found either postmortem or induced artificially), when acetic acid is run under the coverslip. But the casein thus redissolved in such an acid as acetic acid can be precipitated a second time if strong nitric (or hydrochloric or sulphuric) acid is allowed to flow into the test-tube as shown in the tube on the extreme right of Fig. 131. If the protein is not present in excessive amounts, this second precipitate also disappears — we say it goes into solution in the excess of the nitric acid. It is not difficult to see that this is entirely analogous to the reappearance of granules in the kidney cells, with subsequent total solution of the affected cells, on the addition of nitric acid for example, to cells in which a previous set of granules has been made to dis- appear by the addition of acetic acid. Equinormal solutions of different acids are not equally effective in producing a precipitation of casein, neither are they 468 OEDEMA AND NEPHRITIS equally effective in producing the cloudiness of cloudy swelling. In low concentrations certain salts favor the precipitation of casein in dilute acids while others hinder this. The sulpho- cyanates and iodides quickly precipitate casein from an acid solution in heavy curds. Equimolar solutions of the bromids, nitrates, and chlorids produce only an opalescence, while in citrates the casein remains in solution. When arranged accord- ing to the intensity with which these acid radicals favor the development of a cloudiness in the kidney the order is the same. Various basic radicals, in the dilute solutions in which they have to be used to prevent their precipitation as hydroxids, do not influence the precipitation of casein. Neither do they affect the development of cell cloudiness. Kidney cells also follow the behavior of casein toward alkalies. All the alkalies make casein go into solution and, similarly, the alkalies do not produce any clouding in kidney cells. ^ Casein is not precipitated in alkaline solution by the addition of any of the ordinary salts. Neither is a cloudiness produced when any salts are added to slices of liver or kidney immersed in a dilute alkah. Point for point the analogy between the precipitation of casein and the artificial development of a cloudiness in kidney cells seems therefore to be complete, and since there exists no discoverable difference between the changes thus artificially induced in excised kidneys and those which nature produces for us in this same organ in nephritis, nor yet in the conditions leading to these changes in either case, we would seem to be justified in considering all these changes as in essence the same, and as caused fundamentally by the same circumstances. As this process of cloudy swelling represents a series of changes in the state of the cell colloids, it is clear that the employment of any methods in its study — such as fixing agents and various stains — which in themselves are capable of pro- ducing changes in the state of cell colloids, should be excluded. Nevertheless, to meet the possible objection that what has been described in these pages as cloudy swelling might really not be identical with this change as observed on the autopsy table, our 1 The slight grayness developed by slices of kidney, kept for several days in a dilute alkali, has its parallel in the turbidness which we find developed in alkaline solutions of casein, when these are kept for longer periods of time. NEPHRITIS 469 pathologist, Paul G. Woolley, generously offered to examine by approved histological methods the tissues in which I had produced cloudy swelling artificially. He reports that the pictures obtained are identical with the most extreme grades of cloudy swelhng that are encountered pathologically. In concluding these paragraphs we have to answer the final question of the relation of the swelling of the kidney cells to the clouding in them. On the basis of the fundamental work of Wolfgang Paxjli ^ and his coworkers, Hans Handovsky and Karl Schorr, this is easily done. As these investigators have shown, the swelling and solution of a protein colloid (its hydration) and its loss of water and precipitation (its dehy- dration) represent antagonistic processes and are therefore mutually exclusive. It follows from this that the swelling of the cells in a parenchymatous nephritis, and the development of a cloudiness in them, cannot possibly involve but one colloid — in other words, at least two must be involved. The con- ditions which permit the one of these to imbibe water and so to lead to an increase in the size of the cell are of such a char- acter as to lead to the precipitation of another, and so to the cloudiness Wolfgang Patjli kindly advised me to test out this idea in a model made by pouring a solution of casein (prepared by saturating sodium hydroxid with casein) into a concentrated, carefully washed gelatin (20 per cent) and allowing the whole to stiffen. When plates are cut from such a mixture they swell (absorption of water by the gelatin) and become cloudy (precipitation of casein) under the same conditions (presence of acids and various salts) as were found above to lead to a " cloudy swelling " in sUces of kidney. 1 Wo. Paitli: Kolloid Zeitschr., 7, 241 (1910); Pauli and H. Handovsky: Biochem. Zeitschr., 18, 240 (1909), 24, 239 (1910); H. Handovsky: Kol- loid Zeitschr., 7, 183, 167 (1910); Fortschritte in der Kolloidohemie der Eiweisskorper, Dresden, 1911; Kabl Schorr: Cited by Pauli and Han- dovsky. 470 (EDEMA AND NEPHRITIS 4. The Bleeding into and from the Kidney in Nephritis (Hemor- rhage by Diapedesis) The blood that appears in the urine in some cases of nephritis has a purely traumatic origin, in other words, capillaries or larger blood vessels are ruptured and the blood escapes. In large part, however, pathologists hold that blood corpuscles get from the capillaries into the urine by a process of diapedesis. Through diapedesis are also explained many of the hemorrhages into the kidney substance itself. As such bleeding does not occur from the normal kidney we become interested in its mechanism, and it becomes a part of our problem to discover why in nephritis such a process which occurs also in other pathological states should be especially prone to appear. We still lack a satisfactory explanation of the mechanism of diapedesis. Our present teachings continue to partake of the views of von Recklinghausen and Julius Arnold, who held that holes (so-called stomata) exist in the capillaries, and that through these the red blood corpuscles escape in conditions associated with a bleeding by diapedesis. But such a concep- tion, as Julius Cohnheim pointed out years ago, is grossly incorrect, for what escapes from the blood is not the whole blood, but only the red blood corpuscles, and it is inconceivable how holes which would permit the passage of the cellular elements of the blood through them should hold back the liquid portion of the blood. Cohnheim believed diapedesis to be dependent upon changes in the blood vessel walls whereby these became abnormally permeable, after which he held the blood pressure to be able to force the red blood corpuscles through them. How such an abnormal permeability was brought about he declared himself unable to explain. Hemorrhage by diapedesis, while discussed by us because present in some forms of nephritis, is really, of course, a widely distributed pathological phenomenon. As is familiarly known, it occurs in any well-marked passive congestion, produced, for example, by ligation of the veins of any of the parenchymatous organs, of the mesenteric veins, or of those coming from the leg or the ear of a rabbit or dog. But it occurs also after ligation of the arterial blood supply to a part, and I have observed it in NEPHRITIS 471 the entire absence of any circulation in the legs of frogs so ligated as to close both arteries and veins, and kept in a little water. Hemorrhage by diapedesis occurs also in conjunction with the acuter forms of inflammation no matter how induced. It is clear from these few remarks that blood pressure, which might at first sight be thought to be of some importance in squeezing the red blood corpuscles out of the blood vessels into the surrounding tissues, cannot be of great importance in this regard, for diapedesis occurs in conditions associated with a decrease in the blood pressure, or, as just pointed out, even in its entire absence. What is present in all the conditions noted is such a disturbance in the circulation as to lead to a state of lack of oxygen in the tissues, and, we have to repeat, an abnormal production and accmnulation of acids in the affected regions. And this is what we have in the kidney in nephritis. But how does this now lead to the diapedesis? The answer is not hard to find. We have already called attention to the well-known fact that the cells of the living organism represent in the main a mixture of several so-called lyophilic or emulsion colloids. Under normal circumstances in the body these are in a swollen state that is similar to that assumed by fibrin or gelatin when placed in water. If a little acid is introduced into such a colloid the absorption of water by it is enormously increased, and as we have already pointed out before, this is what happens when acid is introduced into the kidney (or into the tissues of the other parenchymatous organs, the intestine, the leg or the ear), in other words, an " oedema " develops. But this increased absorption of water makes the tissue softer (or, to put it more technically, its internal friction is decreased and its surface tension is changed) and now the red blood corpuscle which lies in contact with its surface is no longer held out by the surface layer of the tissue colloids (the blood vessel wall), but penetrates this — is really " swallowed " by the tissue. The increased fluidity of the kidney tissues, after these have been treated with a little acid in the presence of water, is readily observable under the microscope. The cells can be pushed about and molded on slight pressure in a most striking way. What makes the red blood corpuscle move through the tissues are inequalities in the stresses present in the tissue colloids. By 472 (EDEMA AND NEPHRITIS a process, the reverse of that described, the tissue which has once swallowed a red blood corpuscle may again get rid of it, though in practice such a result is hardly to be expected, for after a softened tissue that has swallowed some red blood corpuscles has a more normal circulation restored to it, it is hkely to lose its excess of acid, and so its water, so rapidly that the red blood corpuscles remain behind entangled in the tissues. As a matter of fact, we know that red blood corpuscles which have escaped into the tissues are usually absorbed indirectly after they have disintegrated. What we have said here regarding the red blood corpuscles holds also, of course, for the white blood corpuscles, only these possess in addition independent powers of movement which are lacking to the red blood corpuscles.-^ More strictly in the class with the red blood corpuscles belong the bacteria which we know may reach the kidney from any part of the body and pass through the kidney substance out into the urine. Briefly formulated, the problem of how in nephritis the red blood corpuscles pass into the tissues of the kidney or through these out into the urine, or the problem of how white blood corpuscles or bac- teria do this comes to be the problem of how one colloid body may pass through another, and of the laws that govern such a passage. No holes are necessary in order that one colloid may pass through another, and such a passage is accomplished without one colloid losing its identity in the other or leaving behind it any evidence of its passage. The matter can be prettily illustrated by letting a mercury drop or solid metals (iron fragments or shot, or these covered with colloid material as agar-agar or collodion), under the influ- ence of gravity, move in all directions through a solidified gelatin. The mercury is particularly suitable, for, while not a colloid, it has the " liquid " character possessed by the red and white corpuscles. In the body the migration of the blood corpuscles (or metal fragments, etc.) does not, of course, occur under the ' In the discussion of the migration of white blood corpuscles in inflamma- tion (chemotaxis) most emphasis is always laid upon the changes that the white blood corpuscles themselves are believed to suffer (for example, changes in surface tension), which result in their movement toward the inflammatory center. This is only half the problem. The changes in the tissues themselves (changes in viscosity, for example), produced through the action of the excitant of the inflammation, also play a role. NEPHRITIS 473 influence of gravity, but in consequence of inequalities in the pressure exerted upon the surface of these elements, occasioned through inequalities in the stresses present in the tissues (brought about in turn through local changes in the water content of the lyophihc colloids comprising the tissues). And as the question of whether a mercury drop will enter a solidified gelatin, and the rate at which it will move about in this are matters that have to do with the surface tension relation- ships that exist between the mercury and the gelatin, and the viscosity of the gelatin (in its turn, affected by concentration, temperature, acids, bases and salts), so these same factors play a r61e in diapedesis as observed in the living organism. In Fig. 132 is shown how a mercury drop is unable to penetrate a stiffened gelatin (3 per cent) at room temperature. It may be rolled about on the surface of the gelatin without entering it. If the experi- ment is repeated at the same tem- perature, with a stiffened gelatin of a somewhat lower concentration, the mercury drop enters it and falls slowly to the bottom (Fig. 133, a, b, c). By turning the tube about (Fig. 134), the mercury drop moves in all directions through the stiff gelatin in which, of course, no holes exist, and in which none remain after the mercury has passed.^ The essential change in the gelatin, which makes such pene- trability possible in this experiment, was induced through regu- FlGURB 132. 1 It is this property of coUoids which explains why small wounds made in the Uving animal close immediately. The property of colloids, which gives them such great interest biologically, is the fact that they combine in one the properties of liquids (surface tension, viscosity, diffusion of dis- solved particles) with the properties of solids (maintenance of form). 474 CEDEMA AND NEPHRITIS lation of the concentration. A similar change can be induced by raising the temperature somewhat (not to the point of melt- ing the gelatin, of course) or, in the presence of water, by adding a little acid. This approximates most closely the change that occurs in the body when in passive congestion, for example, a Figure 133. diapedesis into the oedematous tissues is noted. What happens under such circumstances can also be mimicked with some gelatin cakes and a few mercury drops. If one gelatin cake is placed in water, another in a dilute acid, the one in acid undergoes a SM-elling which after a time reaches a stage which readily admits of the passage of a mercury drop, while the control in water will not. NEPHRITIS 475 5. On the Origin and the Different Tjrpes of Tube Casts We have now to discuss how the abnormal acid content of the kidney in nephritis leads to the formation of casts. In this section we shall also leam how the various types of casts that are discovered in the urine in nephritis bear a simple relationship to each other; how, in fact, it is possible to convert one type of cast into another, and back again if we so choose, under the conditions found in the kidney and in the urine in nephritis. What must be the effect of the abnor- mal production or accumulation of acid in the kidney, so far as this problem of casts is concerned, may be determined in any one or all of several ways. We may simply leave the normal kidney, freshly removed from the body, to itself, protect it against evaporation, and study the effects of the postmortem development of acid in it. Or, we may slice the kidney into several pieces and place them in water, or, finally, we may place such slices directly into shghtly acidified water. The kidneys of guinea pigs and rabbits furnish excellent material, and it is on these that the following obser- vations were made. When we take a fresh kidney that has been cut across and squeeze it gently, we only see a Uttle blood ooze from the blood vessels. If we scrape the surface and put a little of the scrapings on a sUde, we find little more than some red blood corpuscles mixed in with a little granular material. In other words, it is difficult to obtain any kidney parenchyma cells^-they do not separate easily from their attachments. The same kidney, preserved for several days, presents a somewhat different appear- ance. The surface may not be quite so glistening when cut, and on squeezing the organ turbid points arise over the surface of the kidney which, when examined microscopically, are seen to be made up of epitheUal cells which have loosened from the kidney Figure 134. 476 (EDEMA AND NEPHRITIS tubules. These may be single, or joined together in groups, and with them are again found the red blood cells and the granular detritus that was observed in a scraping from the perfectly fresh kidney. A somewhat different picture is presented by the sections of kidney that are placed in water. These tissues become gray more quickly than the tissues that do not come in contact with water, and develop an opaque appearance. The normal kidney markings gradually become more and more obscured, and the tissues as a whole are seen to swell somewhat. The whole makes up the t5T)i- cal picture of that which the pathologists call cloudy swelling, and the nature of which we discussed in a foregoing section. The scraping from the surface of such a gray kidney shows a large nmnber of free epithehal cells, which one has no difficulty in recognizing as coming from glomerular tufts and from the urin- iferous tubules. In making the scraping one notices, moreover, that while vigorous scraping yielded little or nothing when appHed to the healthy kidney, it is no trick at all to get an abundant amount of material from the surface of a kidney that has lain in water for a day or two. One notices, moreover, that the numerous epithelial cells are swollen and studded with granules. But beside the individual epithelial cells one notices groups of these, and then casts with rounded ends of whole tubes. One has no difficulty in recognizing these as duplicates of the epithelial casts found in the urine in certain types of nephritis. But the most striking picture is that presented by the sections of kidney that we have thrown into a very weak acid of some kind. In this the cloudy swelling of the slices of kidneys already described occurs very rapidly. A gentle scraping from the surface of a kidney slice, treated with such a dilute acid (n/500 lactic, for example), shows in several hours after immersion a granular detritus, separate epithelial cells, groups of epithelial cells and casts of various kinds (Figs. 135 and 136). When the kidney is simply gently squeezed and its surface touched to a slide, and this is then examined micro- scopically, one cannot escape the impression that he is examining a centrifuged urinary specimen from a case of acute nephritis. The epithelial cells, the epithelial casts, the granular casts are all there. One misses only the hyaline cast, but this can be promptly obtained by simply adding a little stronger acid to the specimen under the microscope, when the granular casts are seen to lose their NEPHRITIS 477 Figure 135. or prolong their residence in the granules, swell somewhat more decidedly and become difficultly visible. Scat- tered nuclei may stick to the casts, but if enough acid is added, these too, go, so that only the greatly swollen, entirely hyaline " cyhndroids " of some authors remain. Or, we can assure our- selves of a generous yield of hyaline casts and cylindroids from the start if we simply in- crease the acid concen- tration into which we drop our kidney slices, solution. We can convert the granular casts into hyaline ones quite as easily through the addition of an alkali as through the addition of an acid, and if the kidney slices are from the first dropped into a dilute alkali, only hyaline casts are ob- tained. The hyaline casts produced through the acids can be converted back into granular casts, if we wish, by simply running a little salt under the cover slip. A sulphocyanate is particularly good for this purpose, but if we wish to use a salt that is more " physio- logical " in nature, Figure 136. 478 (EDEMA AND NEPHRITIS sodium nitrate or sodium chlorid will do. The hyaline casts produced through alkalies can also be converted, into gran- ular ones, though to accomplish this they must be treated with an equinormal acid. Why all these transformations are possible is, of course, readily inteUigible when the experiments on cloudy swelling as detailed in the previous sections are recalled. In Fig. 137 is shown the appearance of a gentle scraping taken from a slice of kidney that has lain in water for several hours. Figure 137 A granular cell detritus and isolated casts characterize such a specimen. Nuclear fragments are prominent, and the epithelial cells may in places still be made out. The cells are granular. In Fig. 138, a, is shown a scraping similarly prepared from a slice of kidney that had lain in n/200 acetic acid for three hours. The cast formation (falling apart of the kidney) is a far more prominent feature. In the cast occupying the central point in the photomicrograph remnants of an epithelial structure are still present. In the casts lying above this all evidences of nuclear structure have disappeared. They are filled with fine granules. NEPHRITIS 479 When these casts were treated with a stronger solution of acetic acid they became hyaline, as shown in Fig. 138, b} It is clear, therefore, that under the influence of a little acid the kidney drops apart into its morphological elements. While these are firmly cemented together in the healthy kidney (as witness the attempt to obtain them by scraping the surface of the kidney with a knife), they are separated with the greatest ease after the kidney has lain in acid for a while. The answer Figure 138. to why the kidney falls apart as it does under the influence of acid it is needless to discuss, but the view that some of the (col- loid) " cement substances " are more easily " soluble " or more easily " digested " in weak acids than other portions of the kidney at once suggests itself. Such a view finds support in our previous considerations of albuminuria and in the fact, easily observed in 1 Casts lose their granules and appear hyaline to ordinary microscopic vision before they become hyaline photographically. This is easily explained by the optical behavior of colloids. As Wolfgang Ostwald has emphasized, the ultraviolet rays affecting the photographic plate are still refracted (and the picture appears granular) by particles too small to change the path of the longer rays of ordinary white light. 480 CEDEMA AND NEPHRITIS these experiments, that the solutions in which the kidney shces he, come to contain with time progressively larger amounts of albumin. That som.e constituents of the kidney (or of any other organ) are more readily soluble in an acid than are others, is clearly enough evident under the microscope. The nuclei of the cells still retain their outlines, for example, in concentrations of acid in which the protoplasm generally has become entirely hyaliiie. The action of the acid could be aided and abetted, of course, by the various substances which in their action on the body colloids act like acids, including the enzymes. What is important to us, from the standpoint of the theory of nephritis, is the way in which the kidney falls apart. The epi- thelial cells tend to stick together while they separate in mass from their supporting membrane. This marks the origin of the urinary cast which, in clinical cases, is washed down into the bladder by the force of the secreted urine. These simple facts regarding the origin of casts, and the con- ditions under which the one type may be converted into another, are not without some clinical significance. In treatises on medi- cine and in works on clinical diagnosis much has been said, not only regarding the importance of the appearance of casts in the urine, but of the sigoificance of the different kinds of casts. It seems to me that the experiments just detailed urge caution upon , one in drawing too sweeping conclusions from such data. So far as mere numbers of ca,sts are concerned, it requires no spe- cial emphasis to realize th^t^gTeat numbers of casts present in the urine at one time, while irriicative of a more extensive involve- ment of the kidney parenchyma at that time may not be as sig- nificant as a lesser number present over longer periods of time. The aggregate destruction may in the latter case, of course, be much greater than in the former (a condition further modified in the living organism by the rate and quantity of the regenera- tion occurring in the kidney). In judging of the meaning of the character of the cast, whether epithelial, granuligir, or hyaline, one must be exceedingly care- ful. We have seen that the epithehal cast is readily convertible into either the granular or the hyaline, depending upon how much acid is present and the length of time that it is allowed to act; and the hyaline, we have seen, can be reconverted into the granular. The thought might suggest itself that we use NEPHRITIS 481 the nature of the cast as an index of the acid concentration in the kidney and so as a measure of the intensity of the nephritis. But this may not be done, for we know from autopsy findings that a nephritis need not affect all the parts of a kid- ney equally, or at the same time, and the urine represents the mixed product of the whole kidney. Moreover, the urine itself varies so in composition under different (physiological) cir- cumstances that it may alter the character of the cast in its passage through the ureter and bladder, no matter what its nature when it left the kidney. A highly acid urine would on the whole tend to yield granular or, if sufficiently high, hyaline casts. An alkaline urine would tend to yield only hyaline casts. On the other hand, the salts of the urine would tend to counteract the acid and make the casts not only smaller (loss of water by the colloid) but more granular (precipitation of the colloid). One can easily satisfy himself of these facts by providing him- self with casts from a clinical case of acute nephritis, or from such kidneys as I have described, and examining them under the microscope, while a little acid, or this in conjunction with various salts, is allowed to run under the coverslips of the preparations. In concluding this section it is well to revert for a moment to the question of albuminuria. It is possible to test the idea that albuminuria results from a " solution " of the proteins of the kidney under the influence of an acid in these experiments on the formation of casts. If we take a perfectly fresh kidney from either a rabbit or a guinea pig, cut it into several slices, and wash the pieces a few times in water or a " physiological " 0.9 per cent NaCl solution, so as to get rid of the blood, we find thereafter that the wash water gives little or no reaction for albumin. But if we permit the pieces of kidney to lie in the wash water until next day, we have no difficulty in getting the albumin reaction. Still more rapidly do we get it if we immerse the washed slices of kidney from the start in a weak acid solution. If we pipette off the sediment found about the kidney pieces and examine this under the microscope, we find at the same time various kinds of casts. But the albumin is not simply due to these, for we continue to get a marked albumin reaction after careful filtration. 482 (EDEMA AND NEPHRITIS VII ON THE ALLEGED CONSEQUENCES OF KIDNEY DISEASE We need now again to break into our main argument show- ing how the factor of acid production acting upon the colloids of the kidney leads to the signs and symptoms of nephritis to discuss the alleged consequences of kidney disease. Until we have disposed of this question we shall not be of one mind on certain points where agreement must be reached before progress can be made. Many of the clinical manifestations observed in patients having kidney disease are considered consequences of the impaired kidney function. There are consequences to such impair- ment, but, almost without exception all those most generally regarded as such do. not belong in the group. Dogmatic teaching and the inertia of time have woven here a tangled skein, but if we will consider these alleged consequences separately and logically, order can easily be established. 1. On the Relation of Vascular Disease to Nephritis § 1 It has long been recognized that vascular disease, increased blood pressure and cardiac hypertrophy are frequently asso- ciated with changes in the kidney which in the aggregate lead to the morphological picture which we call chronic interstitial nephritis. It is also quite generally accepted that such blood vessel disease, hypertrophy and increased blood pressure are consequent upon the kidney disease in the sense that impairment of function is supposed to permit poisonous substances to accu- mulate in the blood, which in addition to producing destructive lesions in the blood vessels themselves lead also to the cardiac hypertrophy and high blood pressure. This conception with all its various modifications is fundamentally wrong. Neither logic nor experiment support it and everything argues against it. The primary disturbance in chronic interstitial nephritis associated with vascular disease and changes in the heart is the vascular disease, and the changes in the kidneys, in the heart and in the other organs of the body are secondary to it. NEPHRITIS 483 No one has as yet produced in animals a chronic interstitial nephritis associated with vascular disease and a hypertrophy of the heart. By injecting various poisons into animals, such as the salts of the heavy metals, it has been possible to produce a chronic interstitial type of nephritis, but the animals show no changes in their vascular system and no hypertrophy of the heart. These kidneys really correspond to the chronic interstitial types of nephritis which we see in human beings who have passed through a generalized parenchymatous nephritis due to an intoxi- cation of some sort and in whom pieces of the kidney have been Figure 139. lost with secondary contracture. We do not as yet possess any experimental method of producing in animals vascular disease as observed in man and not until we do, need we expect to observe a cardiac hypertrophy, increased blood pressure and destructive lesions in the kidney which correspond with the changes observed in human beings. The current notion that vascular change, high blood pressure, etc., are secondary to kidney disease can easily be tested out experimentally. When we observe such signs and symptoms in a human being, the patient is, of course, still alive. He must in consequence have sufficient kidney substance available to live. It is an easy matter experimentally to reduce the kidney 484 (EDEMA AND NEPHRITIS substance of an animal down to the physiological minimum. If the current opinions were correct such animals should show vascular changes, high blood pressure and cardiac hypertrophy, but as a matter of fact they show none of them. In Figs. 139, 140, and 141 are shown the photographs of a series of rabbits in which the kidney substance has been thus reduced. A wedge has first been taken out of one kidney so as to reduce its volume one-half or even more, and then after complete recovery from the effects of this operation the whole of the opposite kidney has been removed. The animals have therefore but one-fourth to one- FlGURE 140. eighth their total kidney substance left. We have had these ani- mals in the laboratory upwards of two and one-half years. They are our breeding stock and are perfectly normal. In Fig. 139 are shown three breeding males, in Figs. 140 and 141 two mothers with their families. Lest it be charged that such facts hold only for the herbivora we insert Figs. 142 and 143, in which are shown a dog and some rats which have undergone a similar reduction in kidney tissue and which are none the worse for it. There is only one way in which the results of these experiments can be interpreted. Loss of kidney function does not lead to vascular disease, high blood pressure, cardiac hypertrophy or the other signs so frequently attributed to chronic interstitial nephritis. NEPHRITIS 485 Once we begin to look upon the vascular disease as the primary cause of the nephritis, the cardiac hypertrophy, etc., we encoun- ter no difficulty in interpreting these. As is well known, the primary changes in vascular disease occur in the smaller blood m vessels. When vascular disease attacks the large blood vessels it is through the small blood vessels which supply their coats. In consequence of the changes occurring in the small vessels (degeneration and swelling of the vascular walls, fatty changes^ calcification, intimal thickening, thrombosis, etc.) circumscribed areas situated in the large blood vessels or in the various organs 486 OEDEMA AND NEPHRITIS of the body are deprived of a proper blood supply. In conse- quence of this deprivation destructive lesions result which in the kidney give rise to the spots of dying kidney tissue seen scattered through an otherwise healthy-looking kidney. It will be recalled that the chronic interstitial types of nehpritis asso- ciated with vascular disease show as a rule the greatest variety of morphological changes. While certain regions are entirely FiGUEB 142. normal in appearance, others show characteristic " degenera- tive " changes, as evidenced by the presence of cells that are swollen and granular, or perhaps have lost their nuclei and are disintegrated (localized parenchymatous nephritis). To take the place of the dead cells we may find new parenchyma cells forming, or there may be evidences of connective tissue proliferation, indicating the ultimate formation of a scar.^ ' The morphological changes that characterize chronic interstitial nephri- tis are in no sense specific. Entirely similar pictures are obtainable in any gland in which the blood supply is cut down directly, or indirectly through ligation of the secretory duct, as Dudley Tait has shown. NEPHRITIS 487 This patchy appearance, resulting from a mixture of nor- mal, degenerating and regenerating cellular elements in the kidney, stands in marked contrast to the uniformity of appear- ance presented by a kidney that has been poisoned, say, with the toxins of an acute infectious disease. Here in a certain sense, all parts of the kidney are affected and to about the same degree. The appearances correspond with the fact that in the first case small patches of the kidneys are successively affected by local disturbances in the circulation in the kidney, in the second all the cells are at once subjected to the same FiGTJKE 143. destructive agent. These facts can be interpreted only by rec- ognizing a local cause for the spots of (parenchymatous) nephritis, and this spotty cause resides in the vascular changes. They are the cause of the nephritis and not the other way about. It also becomes intelligible why little albumin, and few casts go with these types of chronic interstitial nephritis and why the output of water remains normal or as some say is even increased. There is plenty of healthy parenchyma left to secrete the normal amount of water and the spots of parenchyma affected path- ologically give rise to but few casts and little albumin. Nor need the urine of such a nephritic be as highly acid as that of the frankly parenchymatous types, for it is the product of 488 (EDEMA AND NEPHRITIS that coming from healthy kidney mixed with that secreted by the nephritic spots. §2 Just as the arterosclerosis associated with kidney disease is not its consequence, but its cause, so the hypertrophy of the heart observed in such cases is not the consequence of the kidney, but of the blood vessel disease. This is clearly proved by the fact that the (physiologically) worst types of nephritis are those least hable to be associated with any hypertrophy of the heart. We do not find hypertrophied hearts in patients with a gen- eralized parenchymatous nephritis, even though they may for years have suffered from this. Even gradual destruction of the kidneys is not followed by heart hypertrophy. I have under observation a man from whom one kidney was removed for infection eight years ago and who has had constantly since then large numbers of casts and much albumin in the urine from the remaining kidney. In spite of the evident destruction of much kidney substance he has no heart hypertrophy and a systolic blood pressure of 126 with a diastolic of 95 mm. of mercury. That, on the other hand, enormous hypertrophies of the heart may be associated with no kidney symptoms what- soever is familiar to everyone. In this subject of heart hypertrophy and chronic interstitial nephritis we seem, as clinicians, all too often to lose sight of the fact that the hjrpertrophy results in this case, as in any case, from the increased demand for work and the increased rate at which a given amount of work must be done. In the hyper- trophy associated with arteriosclerosis these are determined by at least two changes in the circulation: the reduction in the caliber of the blood vessels and the loss of the elasticity of the blood vessel walls. It should be clearly borne in mind that such roughening of the blood vessel walls as is observed has nothing to do with increasing the work of the heart. The friction encountered in driving the blood through the vessels is not that of blood against blood vessel wall, for since the blood " wets " the walls the friction is that of one layer of liquid over another. With a given kind of blood, the blood vessels determine how much work and power is required to force the blood through NEPHRITIS 489 them, only so far as their length (constant in body), diameter, and elasticity are concerned. So far as the effect of xihanges in diameter is concerned (and in arteriosclerosis the diameter of the blood vessels is diminished), it must be borne in mind that the force required to drive a given volume of liquid through a tube increases about as the cube when the cross -section is diminished one-half. The loss of elasticity becomes a factor because, under physiological conditions, in the time of a single contraction of the ventricle an amount of blood, the equivalent of that ejected from the heart, is not at once pushed along the entire arterial and capillary bed out into the veins. Under normal circumstances it is simply thrown into the elastic arterial system, which dilates somewhat, and then, during the period that follows the systole of the heart, the elastic forces resident in the arteries slowly recoil and squeeze the blood out into the veins. When this elasticity is markedly diminished, the heart must in that proportion force its quota of blood during the time of each systole at once through the whole arterial and capil- lary system. In other words, the heart must do an amount of work in the time of the systole which it ordinarily does in the time of a systole plus a diastole plus the pause — roughly, say, in a third the time. To meet such a contract requires in engineer- ing practice a three times larger engine, and the hypertrophied heart of the patient with sclerotic arteries represents the same idea put to work in nature. A third factor tending to increase the demands upon the heart and so inducing its hypertrophy might reside in the blood itself. A liquid moves through a tube with greater and greater difficulty the more viscid it is. Anything that increases the viscosity of the blood, therefore, increases the amount of work demanded of the heart to push the blood forward. The viscosity of such colloid solutions as the blood Is enormously increased by slight traces of acid ^ (P. von Schroeder,^ W. B. Hardy '^ and especially Wolfgang Pauli and Hans Handovsky*) and so this factor which comes into play, not only in nephritis, but ' See page 106. 2 P. VON Scheobdbb: Zeitschr. f. physik. Chem., 45, 106 (1903). »W. B. Habdt: Journal of Physiology, 33, 251 (1905); Proceedings of the Royal Society, 79, 413 (1907). See page 281. 610 (EDEMA AND NEPHRITIS responsible for the normal oxidation processes that occur here (such as the production of carbonic acid). While such an increase in the amount of acid held in the kidney as occurs in nephritis, does not interfere with the absorption of water from the blood, favors it, rather (as evidenced by the swelling of the kidney), it interferes decidedly with the subsequent loss of the absorbed water which constitutes the palpable external evidence of secretion. The loss of the acid to the blood must be ren- dered the more difficult the higher the amount of acid already present here — wherefore the question of the amount of acid contained in the body as a whole becomes an important factor in the problem of nephritis. As a final word let me point out that the nephritic kidney in -swelling (as it possesses a firm capsule) compresses its vascular supply. In consequence of this it not only decreases, through the decrease in the absolute amount of blood going through the kidney, its opportunities for losing such acid as it has already accumulated, but places its component cells in a position where an abnormal acid pro- duction is immensely favored (lack of oxygen). All these facts must be borne in mind, and point the way to be followed when we come to discuss the matter of treatment. 3. The Secretion of Dissolved Substances by the Nephritic Kidney As already noted, the nephritic kidney shows deviations from the normal secretion of dissolved substances by it in two directions. Other conditions remaining the same there is, first, a decrease in the absolute amounts of the various substances secreted, and second, a change in the relative proportions that these bear to each other when compared with the secretion of these same substances as observed in health. The nephritic kidney secretes some substances as well as does the healthy kidney, others decidedly less well, a third group even better} It is our problem to say how such a condition as an acid produc- tion in the kidney brings this state of affairs to pass. In order to do this we must recall some of the facts of normal secre- tion by the kidney. As is familiar to everyone, a secretion of some substances proportionately more easily than others, in other words, a • SeeW. A. Babtjer: Arch. Int. Med. 11, 593 (1913). NEPHRITIS 511 " selective " secretion by the kidney, is not characteristic of the diseased kidney, but of the healthy kidney as well. This is really the rock on which most of the mechanical, or to use a broader and better term, non-vitalistic or physico-chemical conceptions of urinary secretion have foundered — and these founderings have given momentary comfort to those who believe that kidney secretion, as many another physiological phenomenon, is " vital " in character. But such a pessimism would seem to be premature, for we are already familiar in physical chemistry with not a few systems in which differences in the concentration of any substance are easily maintained over indefinitely long periods of time, and, of course, without the assistance of those " peculiar " forces believed by some to inhabit the living cell. Reference is here made to the difference in the distribution of any substance between two phases {the distri- bution coefficient)} Through the work of Hans Meyer and E. Oveeton the differences in the solubility of such substances as alcohol, ether, chloroform, morphin, cocain, etc., in water and in fats and fatlike bodies (lipoids) — their distribution coefficients be- tween two solvents — have been shown to explain very satis- factorily why these substances not only diffuse with greater speed into and through cells, especially rich in the fatlike bodies (the fat cells and the cells of the central nervous system), than into and through such as contain these in smaller amounts (yellow elastic tissue, white fibrous tissue), but why in the end they are found in larger absolute amounts in some tissues than in others. A second property of protoplasm which permits one cell or tissue to take up more of any given substance, and this more speedily than is the case with another cell, is the char- acter of the colloids contained in the cells and their state. This is one of the reasons why certain stains when injected intraven- ously are not taken up with the same speed, or to the same ex- tent, by all the tissues of the body. A third property of protoplasm, which makes for inequal- ities in the distribution of a substance, resides in the chemical differences existing between different kinds of protoplasm. Certain, but by no means all, of the " vital " and " specific " protoplasmic stains are examples of this class. In these a chem- • See page 165. 512 (EDEMA AND NEPHRITIS ical combination results between the dye and the chemical compounds found in some cells. What use caii we make of these facts in the explanation of the alterations observed in the secretion of dissolved substances by the nephritic kidney? In discussing the colloid-chemical theory of urinary secretion/ I tried to show how the " selective " character of secretion may be explained in the following way: All secretion of dissolved material by the kidney is dependent upon a primary secretion of water by the kidney. After the water is secreted all the constituents which characterize it as urine come to be added to it, in its course through the uriniferous tubules, by a process of leaching out of the dissolved substances present in the kidney cells. But in this process of leaching out, not all the constituents present in the protoplasm leave the cells in which they are originally present with the same ease. Depend- ing upon the character of the dissolved substance, and the state of the protoplasm as to lipoid content, colloid state, and chemical composition, the water present in the uriniferous tubule may come to take up the dissolved substance to an extent which allows it ultimately to be found here in a lower concentration than in the kidney cells, in the same concentration, or in a greater one. It is all a matter of equilibrium. But the equilibrium points with different substances are different, and so the relative amounts of these different substances that appear in the urine are. also different. In other words, the (normal) leaching out is " selective," or, to put it biologically, the " secretion " of the dissolved substances is selective. But this leaching out of dissolved substances from the kidney is only one-half of the process of urinary secretion. The other half is the process of the absorption of dissolved substances from the blood by the kidney cells preparatory to their secretion into the lumen of the uriniferous tubules. This is also a selective process, and here the same laws of lipoid solubility, colloid adsorption, and chemical combination, which have already been discussed in the leaching out process, again come into play. All these various processes of absorption and secretion of dissolved substances by the kidney cells are most markedly influenced by the content of acid, of salts, etc., in them, and it is for this reason 1 See page 324. NEPHRITIS 513 that the observed variations from the normal in the secretion of dissolved substances by the nephritic kidney occur. It is easily appreciated why there must be a decrease in the absolute amount of dissolved substance secreted by the nephritic kidney. If the secretion of water is diminished, then not as much dissolved substance can be leached out of the kid- ney parenchyma as when more is secreted. Into this, however, enters the element of time. When much water is being secreted by a kidney its discharge into the pelvis of the kidney is also hastened. The time that a given portion of the urine (secreted as water initially) is in contact with the kidney cells is thereby diminished, and so not all that this water is capable of absorbing is taken up. When the water is secreted more slowly, the ultimate equilibrium point for the distribution of dissolved substances between the kidney and the urine is more nearly approximated. We find daily expression of this in the clinical observation that after the consumption of much water the concentration of the urine falls, while with a diminished intake of water, or when the kidney cannot secrete it (as in nephritis), the concentration of the urine becomes progressively higher. Yet, other things being equal, the absolute amount of any dissolved substance secreted by the kidney must be the greater, the larger the absolute amount of water secreted by the kidney in any unit of time. To illustrate how the increased acid content in the kidney in nephritis leads to variations in the secretion of the dissolved substances, I introduce some simple test-tube experiments and experiments on rabbits, which concern themselves particularly with that part of the selective secretion which deals with the state of the colloids in the kidney cells. This constitutes by far the most important part of the whole problem of selective absorption and secretion, for the state of a colloid in the body is more easily affected by external conditions than is the solvent property of a lipoid, or the chemical character of any part of living protoplasm. As the various dyes betray themselves not only qualitatively, but, in a sense, also quantitatively, to the naked eye, illustrations of the " absorption " and the " secretion " of these, under conditions that interest us in our discussion of nephritis, seemed to me best suited to our needs. I chose, moreover, dyes that have been used physiologically in the study of the kidney. The results of a few experiments 514 (EDEMA AND NEPHRITIS on the staining of fibrin, which are familiar to any worker who has at all touched upon the problem of dyeing, and which might be multiplied indefinitely by using other dyes and different colloids, are shown in Fig. 144. Tube 1 contains an aqueous solution of toluidin blue. If into another tube (2), containing the dye in the same concentra- tion, some powdered fibrin is dropped, this soon absorbs most of the dye and stains intensely blue. The supernatant liquid retains only a faint tinge of the blue, but ^his remains indefinitely. If the supernatant solution is carefully pipetted off, and distilled water is placed over the dyed fibrin, the water now slowly turns blue. In this way, through successive washings, we can again get considerable of the blue out of the fibrin. In other words, the fibrin absorbs the dye until an equilibrium is reached between the concentration of the dye in the fibrin and the concentration of the dye dissolved in the supernatant liquid. If we disturb this equilibrium by removing the blue solu- tion above the fibrin and substituting water for it, some of the dye comes out of the fibrin until equilibrium is once more established. If we will now substitute the words kidney colloids for fibrin we have what happens in the kidney when it secretes any dye. The absorption of the dye by it from the blood is analogous to the first series of changes, the leaching out of the dye by the urine to the second. We see also why the quantity of urine secreted and the time that this remains in contact with the kidney cells are of such importance. This corresponds with the renewal of the distilled water above the dyed fibrin and the time this is allowed to remain there before being pipetted off. What happens if we introduce into this whole system a trace of acid? The result is shown in tube 3. The fibrin swells somewhat, but the toluidin blue is now scarcely taken up. The supernatant liquid remains practically as blue as the control tube 1. What would this mean when applied to the kidney affected with nephritis, for which we have maintained that an abnormal acid content is responsible? That the kidney would swell as does the fibrin, we already know. But such a kidney would now not absorb the toluidin blue preparatory for secre- tion as does the healthy kidney. Yet, we must not hastily conclude herefrom that under such circumstances the kidney CD 10 CO i ) CM i 1 - ^^^■H g^ NEPHRITIS 515 would necessarily secrete this dye badly. Once any dye was in the kidney colloids it would rapidly diffuse into the urine, not only because the kidney colloids are not holding on to the dye particularly firmly, but because the acid liable to be in such urine as is secreted from the nephritic kidney would further favor the passage of the dye into it. In tubes 4, 5, and 6 are shown a parallel series of experi- ments carried out with sodium indigosulphonate. It is clear that with this dye conditions are exactly the reverse of those obtaining in the case of toluidin blue. The very circumstances which favor absorption before, hinder it here, and those which hindered it before now favor it. In tubes 7, 8, and 9 are shown the results obtainable with neutral red, which, it will be observed, behaves like toluidin blue. But the kidney is not thus offered one substance at a time to secrete into the urine. The blood that passes through this organ brings it many at once. What must be the behavior of the tissue colloids under such circumstances? As tubes 10, 11, and 12, and tubes 13, 14, and 15 clearly show, a colloid under such circumstances behaves toward each of the substances offered it as though the others were not present. In tube 10 is shown the effect of mixing sodium indigosulphonate and neutral red. If some fibrin is introduced into this mixture it absorbs the red (chiefiy) and leaves behind (almost) all the blue. This would correspond with the kidney function in health. If now an abnormal amount of acid were present in the kidney (nephritis) just the reverse would result — the red would now be left behind in the blood, while the blue would be absorbed. In tubes 13, 14, and 15 are shown the results on the stain- ing of fibrin when toluidin blue and neutral red are mixed. The resulting color is shown in tube 13. In the presence of fibrin alone both of the dyes are absorbed as shown in tube 14, but if a little acid is present, or is subsequently added, the fibrin fails to stain. Fig. 144 was painted from the results obtained in Experiment 75, after the tubes had stood some eighteen hours. Marked differences in the degree of staining are readily visible, however, after ten minutes. 516 OEDEMA AND NEPHRITIS EXPEBIMENT 75. 1. 15 cc. .01% toluidin blue+15 cc. water. 2. 15 cc. .01% toluidin blue+15 cc. water+0.4 gram fibrin. 3. 15 cc. .01% toluidin blue+15 cc. n/20 acetic acid+0.4 gram fibrin. 4. 15 CO. .02% sodium indigosulphonate+15 cc. water. 5. 15 cc. .02% sodium indigosulphonate+15 cc. water+0.4 gram fibrin. 6. 15 cc. .02% sodium indigosulphonate+15 cc. n/20 acetic acid+0.4 gram fibrin. 7. 15 cc. .02% neutral red+15 cc. water. 8. 15 cc. .02% neutral red+15 cc. water+0.4 gram fibrin. 9. 15 cc. .02% neutral red+15 cc. n/20 acetic acid+0.4 gram fibrin. 10. 15 cc. .02% sodium indigosulphonate+15 cc. .02% neutral red. 11. 15 cc. .02% sodium indigosulphonate+15 cc. .02% neutral red +0.4 gram fibrin. 12. 7| cc. .04% sodium indigosulphonate+7i cc. .04% neutral red + 15 cc. n/20 acetic acid+0.4 gram fibrin. 13. 15 cc. .01% toluidin blue+15 cc. .02% neutral red. 14. 15 cc. .01% toluidin blue+15 cc. .02% neutral red+0.4 gram fibrin. 15. 7| cc. .02% toluidin blue+7| cc. .04% neutral red+15 cc. n/20 acetic acid+0.4 gram fibrin. The details of this experiment have already been discussed in the text. It follows from all this that the presence of a little acid in such colloid material as composes the kidney must be followed by profound changes in the character of the secretion of dis- solved substances by it as compared with the normal secretion of these same substances. But depending upon the way in which the acid displaces the equilibrium point, it is clear that, with otherwise constant conditions, the secretion of any substance may. not only be decreased or simply remain unaffected, but it may actually be increased. Before closing this chapter it is well to refer to a few animal experiments which show that what has been said above regard- ing the staining of fibrin actually holds in the case of the living animal. As pointed out in our discussion of the experiments of Heidenhain, Deeser, and Nussbaum, these authors found the kidneys of their experimental animals stained most deeply, and most generally with sodium indigosulphonate or acid fuchsin (which stains fibrin just as does sodium indigo- sulphonate) when conditions favoring the accumulation of acid NEPHRITIS 517 in the kidney were most clearly at hand. This corresponds with the improved tendency of fibrin to stain with these dyes when an acid is present. When in a rabbit under morphin anes- thesia the artery to one kidney is clamped for an hour or two, and the clamp is then removed while sodium indigosulphonate or acid fuchsin is injected intravenously, it is found that the clamped kidney not only stains sooner than the undamped one, but more intensely. When frozen sections are made of the two kidneys the dye in the healthy kidney is found only in the lumina of the uriniferous tubules, while in the ligated one it is found in the cells themselves. And yet a kidney so clamped for an hour or two may not jdeld any urine for hours afterwards. Mere staining of the kidney, as already noted above, can not at once be taken as an index of secretion. The reverse of this experiment can be done with neutral red. Here the normal kidney stains well and rapidly, while the clamped one remains without color, owing to the acid developed in it in the absence of a circulation. After what has been said it must be self-evident that so many factors enter into the picture of the secretion of any dissolved substance by the kidney — so many at which we can to-day but guess in a clinical case — that conclusions regard- ing the functional activity of the kidney, as derived from a study of the elimination of some one compound swallowed by or injected into the patient and sought for in his urine, must only be drawn with the greatest care. Even though we ignore all other elements of error, the state of the blood, the state of the kidney colloids, and the state of the urine all influence the rapidity and perfection of the elimination of the substance in so marked and (for us) uncontrollable a way, that trust- worthy conclusions are hardly possible, and when we take the liberty, as is so often done, of applying without modification what we may have learned from the elimination of one sub- stance to some other or all other constituents found in the urine, then we are on dangerous ground indeed. Until we have learned far more regarding the laws that govern the secretion of dissolved substances by the kidney than we know to-day, we had best accept as the most reliable test for the functional activity of this organ its ability to eliminate water} 1 See page 621. 518 (EDEMA AND NEPHRITIS IX SOME EXPERIMENTAL FOUNDATIONS FOR THE TREAT- MENT OF NEPHRITIS. FALLACY OF SALT RESTRIC- TION IN NEPHRITIS AND (EDEMA 1. Introduction Before use is made in clinical practice of the rather obvious conclusions to which the considerations of the previous pages compel us, conservatism demands that we apply a further test to them. We have labored thus far to show how a parallelism exists between the changes that various protein colloids undergo in the presence of acid and similarly acting substances, and the changes that are observed in the kidney when this becomes the seat of a nephritis. We have noted how the swelling of the kidney in nephritis is like the swelling of fibrin or gelatin in water when a little acid is added to this; how vmder the same circum- stances some of the colloids go into solution, and so the develop- ment of an albuminuria is simulated; how when a colloid of the nature of casein is mixed with these, it is precipitated under conditions which make the others swell, thus behaving like cer- tain granules observed to arise in the cells of the kidney under conditions associated with a nephritis. In studying the behavior of the pure colloids we learned more than this. We learned that the swelling of the colloids could be reduced, not only by neutralizing the acid, but by add- ing to the acid any neutral salt. So far as the precipitation of casein was concerned the salts divided themselves into two groups — the one added itself to the effect of the acid and favored precipitation, the other counteracted such an effect. If now our contention is correct that the series of changes observed in these simple colloids and in the kidney are identical in character, then it is to be expected that the administration of 'properly selected salts should relieve the various signs characteristic of a nephritis. That such is the case is not only a long accepted clinical fact, but can easily be proved experimentally. NEPHRITIS 519 Of the many salts which might be tested one can, of course, foresee that those will give best results which have no " specific " poisonous action, which have the power of neutralizing acid, and which combine a maximum of those effects which tend on the whole to help the nephritis (reduction of protein solubility and swelling of the kidney) with a minimum of those which aggravate such a condition (augmentation of protein precipita- tion in the cells). But neutral salts are also effective in reduc- ing the solution and the swelling of such colloids as fibrin, gelatin and serum albumin. Of the long list of such, one has in recent years been particularly signaled out for attack. It has been widely taught that sodium chlorid is not only not good for a nephritic, but distinctly bad, in that it is held to increase not only the signs of a nephritis, but is supposed to be responsible for a retention of water and so far the aggravation of the oedemas so often seen accompanying such. Neither in the observations on pure proteins nor in the experiments on urinary secretion can a single fact be found to support such a view. As the following experiments show, various neutral salts decrease the signs and symptoms of nephritis, and sodium chlorid is no exception to this rule. In order to show that salts inhibit the development of the signs of a nephritis, it was first necessary to decide upon satis- factory methods of producing a nephritis experimentally in animals, upon which might then be tried the action of various salts. Three different ones were employed : interference with the respiration of the animal, the intravenous injection of acid, and direct clamping of the renal blood vessels. Of all these the last named is probably grossest in its effects upon the kidney. For the sake of comparison the protocols of the experiments on the animals which served as controls have been inserted in each case. 2. Asphyxia! Nephritis Let us first consider the nephritis that develops in rabbits, when these are tied into the animal holder sufficiently tightly to interfere with their respiration. One always gets an albumi- nuria after such a procedure, as Experiments 76, 77, 78, and 79 show. 520 (EDEMA AND NEPHRITIS ExPEEiMENT 76. — White rabbit; and grass. Snugly tied into holder, catheter. weight 898 grams. Urine obtained with Fed wheat soft rubber Time. Urine in cc. Remarks. 3 00 3.30 3.4S 4.00 4.15 4.30 4. 45 5.00 5.15 5 16 3.7 5.0 3.0 1.8 1 0.9 1.1 1.5 1.8 ' Alkaline to litmus paper. No albumin. Clearer urine. Neutral to litmus paper. Trace of albumin. Clear urine. Neutral to litmus. Albumin present. Clear. Neutral to litmus. Albumin present in every sample, and increasing in amount. Animal seems entirely well. Returned to hutch. Experiment 77. — Belgian hare; weight 1226 grams. Fed wheat and grass. Snugly tied into holder. Urine obtained with a soft rubber catheter. Time. Urine in cc. Remarks. 2.45 Few drops \ 3.00 3.15 1.0 1 0.5 1 Alkaline, thick, chrome yellow. No albumin. 3.30 Few drops > 3.45 Few drops Neutral and clearer. Trace of albumin. 4.00 Few drops ~| 4.15 Few drops 4.45 2.0 > Neutral and clearer. Albumin present in every sample. 5.00 Few drops 5.15 Few drops -^ 5 16 Animal released and returned to hutch Experiment 78. — Belgian hare; weight 1020 grams. Fed wheat and grass. Snugly tied into animal holder. Urine obtained with a soft rubber catheter. Time. Urine in cc. Remarks. 3.30 3.0 1 3.45 0.7 Tied down. Turbid, dark yellow. Alkaline to litmus. No 4.00 0.5 albumin. 4.15 Few drops J 4.30 2.4 Turbid, dark yellow, alkaline to litmus. No albumin. 4.45 8.6 Clearer, pale yellow. Acid to phenolphthalein. Albumin present. 5.00 4.0 1 5.15 5.30 2.0 Few drops \ Clear, acid to phenolphthalein. Albumin present in every 5.45 Few drops sample. 6.00 3.0 > NEPHRITIS 521 Experiment 79.— Belgian hare; weight 1100 grams. Fed wheat and grass. Snugly tied into holder. Urine obtained with a soft rubber catheter. Time. Urine in cc. Remarks. 4.00 4.15 4.30 4.45 5.00 5.15 5.30 5.45 6.46 1.0 1 2.0 1 Few drops | 1.5 J 2.0 1 1.5 1 2.5 f 3.0 I Tied down. Allcaline, turbid, thick. No albumin. Urine clear, acid to litmus. Albumin present in every sample. Liberated. Returned to cage. When, now, rabbits are treated from an experimental stand- point in an identical way but have a concentrated salt solution injected intravenously, the albuminuria does not develop. This is shown by Experiments 80 and 81. Experiment 80. — Black rabbit; weight 917 grams. Fed wheat and grass. Snugly tied into animal holder. Urine obtained with a soft rubber catheter. 105 cc. m/2 (2.918%) NaCl solution are given intravenously in the course of the experiment at the rate of 5 cc. every five minutes. Time. Urine in cc. Remarks. 4.00 7.5 Thick chrome yellow, alkaline. No albumin. Injection begun. 4.15 2.5 Thick, chrome yellow, alkaline. No albumin. 4.30 7.5 Clearer. No albumin. 4.45 12.0 C?)l 5.00 23.0 Clear as water. Neutral to litmus. No albumin in any speci- 5.15 36.5 } men. 5.30 32.0 5.45 27.5 J 5 46 Animal well. Released and killed by blow on head. Nothing abnormal noted on autopsy. Experiment 81. — Belgian hare; weight 919 grams. Fed wheat and grass. Snugly tied into holder. Urine obtained with a soft rubber cathe- ter. 105 cc. m/2 (2.918%) NaCl solution are injected intravenously in the course of the experiment at the rate of 5 cc. every five minutes. Time. Urine in cc. Remarks. 3.15 3.30 3.45 4.00 4.15 4.30 4.45 5.00 4.0 2.5 20.0 57.5 40.0 39.0 29.0 37.0 ' Alkaline to litmus, thick, yellow. No albumin. Injection begun. Somewhat clearer. No albumin. Clear, colorless. Neutral to litmus. No albumin in any specimen. 522 CEDEMA AND NEPHKITIS The following Experiments 82, 83, and 84 show that a mixture of different neutral salts, as represented by a Ringer solution, yields entirely similar results. ExpEBiMENT 82. — Belgian hare; weight 901 grams. Fed wheat and grass. Snugly tied into holder. Urine obtained with a soft rubber catheter. In the course of the experiment there are injected intravenously 135 cc. of a Ringer solution X 4,' at the rate of 5 cc. every five minutes. Time. Urine in cc. Remarlcs. 1.40 4.0 Turbid, alkaline to litmus. No albumin. 1 45 2.00 4.0 Turbid, alkaline to litmus. No albumin. 2.15 16.0 1 2.30 38.0 2.45 47.0 } Clear, alkaline. No albumin in any specimen. 3.00 40.5 3.15 32,0 J 3.30 13.0 3.45 7.0 Clear, neutral. No albumin in any specimen. 4.00 6.0 1 4.05 No uriae Animal well. Returned to cage. Experiment 83. — Belgian hare; weight 823 grams. Fed wheat and grass. Tied tightly into holder. Urine obtained with a soft rubber catheter. In the course of the experiment there are injected intravenously 125 cc. of a Ringer solution X4, at the rate of 5 cc. every five minutes. Time. Urine in cc. Remarks. 1 50 1.55 2.10 2.26 2.40 2.55 3.10 3.26 3.40 3.55 4 00 3.0 18.0 13.0 -, 17.0 20.0 \ 8,0 4.0 J Turbid, alkaline. No albumin Clear, alkaline. No albumin. Clear, neutral to litmus. No albumin in any specimen. Dies. Nothing abnormal noted at autopsy. ^ The sodium, potassium, calcium chlorid mixtures that are known as Ringer solutions have a different composition with different authors. I used the following: NaCl 0.7, CaClj 0.0026, KCl 0.035, and H2O enough to make 100 cc. Ringer solution X4 means four times this amount of salts in each 100 cc, a solution which has then about the same osmotic concen- tration as m/2 NaCl, as used in the previous experiments. NEPHRITIS 523 ■ Experiment 84.— Belgian hare; weight 855 grams. Fed wheat and grass. Tied tightly into holder. Urine obtained with a soft rubber catheter. In the course of the experiment there are injected intravenously 150 cc. of a Ringeh solution X4, at the rate of 5 cc. every five minutes. Time. Urine in cc. Remarks. 2.30 2.45 3.00 3.15 3.30 3.45 4.00 4.15 4.30 4.45 5.00 5.15 5.20 1.0 1.7 , 3.6 11.0 12.5 21.0 25.0 23.0 25.0 (?) 1 9.5 12.5 10.0 Turbid, alkaline. No albumin. Tied down and intravenoua injections into ear begun. Urine clears until it looks like water. No albumin at any time. Clear, acid. No albumin at any time. Killed. On autopsy nothing abnormal except that 25 cc. fluid are obtained from the peritoneal cavity! Let us now retrace our steps and see what has happened so far as urinary secretion is concerned, for we were rather par- ticular to emphasize that the ability of a kidney to secrete water was the best index of its functional activity. The experiments detailed above already suffice to show that the secretion of urine 10 8 - ,^^ 6 X \(a\ 4 / /x^^^ 2 cc. ^^L==4 ^>^Cr^ Hours 2 Figure 145. from a nephritic kidney, or one threatened with a nephritis, tends to be maintained at a normal level or to he increased by the giving of various neutral salts, including sodium chlorid. We need but compare with each other the secretion curves of Figs. 146, 147 and 148 made by plotting time on the horizontal and the number of cubic centimeters secreted every fifteen minutes on the vertical and Experiments 76 to 84, upon which the curves are based. All the figures are drawn to the same scale. 524 (EDEMA AND NEPHRITIS U Iloiir^ FiaxiEB 146. Fig. 145 is introduced for comparison and shows normal urinary secretion in three rabbits loosely tied into an animal holder. The curves a, h, c, and d, of Fig. 146 (based respectively on Experiments 76, 77, 78, and 79) show, when compared with the curves of Fig. 145, how the secretion of urine is diminished when instead of being loosely tied into the animal holder 60 ■ the rabbits are so snugly tied down as to embarrass their respiration. The di- minished secretion gives way to an enormously heightened one if animals similarly treated are injected with a concentrated sodium chlorid solution. Fig. 147 shows this. Curve a is taken from Experiment 81, curve b from Experiment 80. These experiments (as others to be described di- rectly) show clearly that administration of sodium chlorid does not lead to a retention of water by the living animal. Fig. 148 shows the curves obtained by inject- ing concentrated Ringer solution. Evidently all neutral salts (that have not specific poisonous effects) when injected in sufficient concentration increase the output of urine. The Hours 1 2 curves a, b and c are con- Figure 147. NEPHRITIS 525 structed respectively from Experiments 82, 83, and 84. These rabbits were again snugly tied into animal holders, but not only did none of them develop an albuminuria, but in conse- Hours 1 2 Figure 148. quence of the injection of concentrated Ringer solution the urinary output was greatly increased in all. 3. Nephritis Produced by Injecting Acid As the following experiment shows, sodium chlorid when injected intravenously, in concentrated solution, simultaneously with a hydro- chloric acid solution of a concentration found in Experiments 60 and 61 {pages 417 and 418) to lead to the symptoms of a most intense 526 CEDEMA AND NEPHRITIS nephritis, practically suppresses this. The albuminuria scarcely appears, and there are no casts, no red Mood corpuscles, no hemo- globinuria, no decrease in the amount of urinary secretion, and no general osdema. ExPEEiMENT 85. — Belgian hare; weight 2136 grams. Has been fed liay, oats, corn, and greens. In the course of the experiment there are injected intravenously at a uniform rate 140 cc. of the following mixture: 150 cc. n/10 HCl + 4.666 grams sodium chlorid and enough water to make the whole up to 160 cc. This yields a final solution, that is m/2 (2.918%) so far as the sodium chlorid is con- cerned. Urine obtained with a catheter. Time. Urine in cc. Remarlss. 3.30 3 45 Injection into ear begun. Sliglitly turbid, neutral to litmua paper. No albumin. No 4.00 0.3 easts. 4.15 17.0 Clear as water, barely reddens blue litmus paper. No albu- min. No casts. 4.30 61.0 Clear, barely affects blue litmus paper. Faint shimmer of albumin! No casts! 4.45 64.5 Urine clear, barely affects blue litmus paper. Faint trace of 5.00 58.0 albumin. No casts. No hemoglobinuria at any time. 5.15 38.0 No red blood corpuscles. 5.18 1.0 Dies. Total amount of urine secreted since beginning injection 239.8 cc. Autopsy. — Weight 2035 grams! Nothing abnormal is noted. The body cavities contain no fluid. The blood seems to coagulate abnormally rapidly. It might be insisted in criticism of this experiment, that while sodium chlorid is thus able to counteract the effects of an acid in producing a nephritis, it cannot relieve such after once being established. This criticism is met in Experiment 86, in which a nephritis is first induced by injecting (practically) pure acid, after which its relief is brought about by injecting m/2 (2.918 per cent) sodium chlorid. Experiment 86.— Belgian hare; weight 2343 grams. Fed hay, oats, corn and greens. Urine obtained with a soft rubber catheter. In the course of the first li hours of the experunent there are injected at a uniform rate 125 cc. of the following mixture: 120 cc. n/10 HCl plus 8 cc. 2/m NaCl, in consequence of which all the signs of a nephritis develop. For the acid mixture is then substituted a pure m/2 (2.918%) NaCl solution of which, up to the end of the experunent, there are injected 125 cc. With the change in the character of the injec- tion fluid the signs of the nephritis are seen to disappear. NEPHRITIS 527 Time. Urine in cc. Remarka. 2.45 5.0 Catheterized. Turbid, light yellow, faintly alkaline to litmus paper. No albumin. No casta. 3.00 1.5 Weighed. Tied to animal holder. Intravenous injection of acid mixture into ear begun. Urine turbid, light yellow, faintly alkaUne to litmus paper. No albumin. No casta. 3. IS 3.30 3.45 0.7 Urine neutral to litmua paper. No albumin. No casta. 2.1 Urine neutral to litmus paper. No albumin. No casts. 4.00 10.0 Urine faintly acid. Albumin. Isolated casts. Epithelial cells and red blood corpuscles. 4.15 7.0 Urine has a pink tinge. More albumin. Numerous casta and a larger number of red blood corpuscles. Injection of acid mixture stopped. Injection of m/2 NaCl begun. 4.30 20.0 Urine decidedly red (hemoglobinuria). Albumin content still rising. Fewer casts and red blood corpuscles. 4.45 42.0 Pink color to urine. Albumin decreasing. No casts can be found after long search of sedimented urine. 5.00 60.0 Pale pink. Albumin decreasing. No oasts or red blood corpuscles. 5.15 53.0 Like water. Barely visible trace of albumin. No casts or blood corpuscles. 5.30 32.0 Like water and neutral to litmus paper. No albumin. No casts. No blood cells. Injection stopped, as animal has embarrassed respiration. 5.35 11.0 a) Some urine accidentally lost as animal dies. No albumin. No casts. No blood cells. Autopsy. — Weight 2342 grams. Nothing abnormal in any of the organs. The peritoneal cavity is wetter than normal. The pericar- dial and pleural cavities are empty. A number of interesting facts come to light in the two experiments just detailed. Let us first ask about the out- put of urine. In Fig. 149 we find in the curves a and b (Experi- ments 60 and 61) a graphic representation of the amount of urine secreted when n/10 hydrochloric acid (in m/8, that is, 0.729 per cent sodium chlorid, added to reduce somewhat the hemolytic action of the acid) is injected intravenously. When we compare these curves with those of Fig. 145 (normal secretion in rabbits), we notice that in spite of the great injection of water, the urinary output lies below the normal. The presence of the acid along with the water brings it to pass that the water is retained in the body; in other words, an oedema develops. The same factor, therefore, which we are holding responsible for certain of the kidney changes in neph- Hours Figure 149. 528 (EDEMA AND NEPHRITIS 60 50 40 ritis, is responsible for one of the most prominent symptoms of such kidney disease, namely, the oedema. How enormously the urinary output is increased if a con- centrated sodium chlorid solution is injected along with the hydrochloric acid is apparent when Fig. 150, drawn to the same scale, is compared with Fig. 149. And when we look through the protocols we find that this increased urinary output is asso- ciated with a loss of weight by the animal, in other words, a failure to develop an oedema, or the reduction or total disappearance of such as may be existing. Clearly, therefore, salts, including sodium chlorid, all tend to reduce oedema, as I have previously insisted. Another point of interest in these two experiments is the fact that when enough sodium chlorid is injected along with the acid, the hemoglobi- nuria fails to develop. As is well known, a pure acid solution when injected intravenously leads to a rapid and extensive destruction of the red blood corpuscles (hemolysis) and the escape of hemoglobin in the urine. The only reason why some sodium chlorid was given along with the acid injections in the various ex- periments described in this volume, in which the effects of the pure acid on the kidney were particu- ^ larly sought, was to escape in part this so great dissolution of the red blood corpuscles. When enough sodium chlorid is added the hemolytic action of the acid is avoided altogether, as Experiment 85 shows. This fact is of interest and importance, not only because it teaches 30 Hours Figure 150. NEPHRITIS 529 us that by increasing the salts in the diet we can relieve the signs and symptoms of paroxysmal hemoglobinuria/ but because it was a result to be expected if a theory of hemolysis which I have previously advanced ^ should be correct. Incidentally, the last two experiments in conjunction with Experiments 59, 60 and 61 (pages 416 to 418) serve to meet a criticism that might have been raised against the experiments on asphyxial nephritis, in which it might have been said that the great urinary secretion obtained after injecting salt solu- tions was due merely to the injection of so much water. 4. Nephritis Due to Temporary Closure of the Renal Vessels As Max Herrmann first showed, direct interference with the blood supply to the kidney leads to very destructive changes in this organ in an incredibly short space of time — the output of urine falls or may be stopped entirely and albumin, casts, and blood are found in such as is secreted. If the kidneys are examined they are found swollen, maybe grayish, and to present varying degrees of hemorrhage into the kidney substance. The following experiment illustrates this: Experiment 87. — Belgian hare; weight 2335 grams. Fed hay oats, corn, and greens. Urine obtained with a catheter. The right renal artery and vein, and the left renal artery are clamped for one- half hour. Time. Urine in cc. Remarks. 2.05 0.008 gram morphin hydrochlorid given subcutaneously. Clear, brownish-yellow, faintly acid to litmus. No albumin. 2.15 ' isio ' No casts. After catheterizing the animal is weighed. 2.50 Tied into holder. 3.00 0^5 Right renal artery and vein and left renal artery are clamped. 3.15 • 3.30 Clamps removed. 3.45 4.00 4.15 1.0 Much albumin. Hyaline casts and red blood corpuscles. 4.30 4.45 0.5 Thick, turbid, acid to litmus. Full of albumin and casts. 5. (JO 0.8 Same 5.15 5.30 0.8 Thick, turbid, acid to litmus. Full of albumin and casts. 5.31 Animal appears well, is killed. Autopsy. — Kidneys are swollen and deep red, but otherwise show nothing strikingly abnormal to the naked eye. 1 Oscar Berghausbn: Unpublished paper. 2 Martin H. Fischer: KoUoid Zeitschr., 5, 146 (1900) and page 364 of this volume. 530 (EDEMA AND NEPHRITIS The urinary output in this experiment is illustrated in curve c of Fig. 151. Let us now see how an animal similarly treated fares if it receives an in- travenous injection of a "physiological" m/8 (0.729 per cent) sodium chlorid solution. Such an experi- ment serves to answer two important questions. First, is the giving of water to a case of " acute nephritis " dangerous because it "throws work on the kidney " and we need to " protect " this organ against doing any work; and second, does the administration of sodium chlorid aggravate such a nephritis because, as some say, it " further increases ijie work of the kidney " or because it " irritates " this organ? A no incon- siderable portion of the therapeutic world to-day in- sists on both restriction of water and of sodium chlorid in cases of acute nephritis. That both may be given and that the sodium chlo- rid, far from adding itself as a factor of evil to the water, really counteracts the only bad effects this has (through favoring the swell- ing of the kidney and wash- ing out salts) is shown by the t results of Ex- FiGUEE 151. periment 88. NEPHRITIS 531 ExPEBiMENT 88.— Belgian hare; weight 2184 grams. Fed hay, oats, corn, and greens. Urine obtained with a catheter. The right renal artery and vein, and the left renal artery are clamped for one- half hour. Thereafter, 215 cc. of m/8 NaCl solution are injected at a uniform rate intravenously. Time. Urine in cc. Remarks. 11.20 0,016 gram morphin hydrochlorid given subcutaneously. 11.40 2,0 Thick, yellow, turbid, acid to rosolic acid. No albumin. No casts, Catheterized and weighed. 12.10 2,0 Same, Right renal artery and vein and left renal artery clamped. 12.25 12,40 Clamps removed. 12.50 Injection of m/8 NaCl into ear begun. Accident to needle interrupts injection for ten minutes. 1.05 2,0 Full of albumin, epithelial, finely granular and hyaline casts. 1.20 4,7 Acid to rosolic acid. Full of albumin, epithelial, finely gran- ular and hyaline casts. 1.35 12,0 Clear as water. Albumin going down. It ia noted that there ia decidedly more albumin in this experiment than when m/2 NaCl is used, volume of urine duly considered. 1.50 12,5 Decided drop in albumin. Still some caata. 2.03 15,0 Clear as water. Albumin present in tracea only. Occa- sional cast only. 2,20 18,5 Same, Trace of albumin viaible after standing. 2,35 19,0 r 2.50 20,0 No albumin. No casta. 3.05 21,0 I 3.06 Killed. Autopsy. — Weight 2269 grams. 6 cc. fluid in peritoneum. Pleural and pericardial cavities dry. Nothing abnormal noted in any organ. The increased urinary output, in consequence of the injec- tion of the m/8 sodium chlorid solution, is clearly evident when curve 6 of Fig. 151, based on this experiment, is compared with curve c obtained in the previously described Experiment 87. We note, moreover, that with the concentration of sodium chlorid employed in Experiment 88 a not inconsiderable amount of the injected water is retained, in other words, the animal develops an oedema. Many clinicians who believe that sodium chlorid " leads to oedema " might be inclined to say that this experiment supports their contention. That it does not is shown by Experiment 89, in which an animal, again rendered neph- ritic by clamping the renal vessels, is again injected with the same amount of water at the same rate, but the concentration of the sodium chlorid is further increased (to m/2 NaCl, that is, 2.918 per cent). As the protocol and curve a of Fig. 151 show, the urinary output under such circumstances is still 632 (EDEMA AND NEPHRITIS further, really enormously, increased, and not only does no CBdema develop, but the animal actually loses in weight; To the interpretation of these various findings, which it was entirely possible to predict, we shall come immediately. Experiment 89.— Black rabbit; weight 2778 grams. Fed hay, oats, corn, and greens. Urine obtained with a catheter. The right renal artery and vein, and the left renal artery were clamped for one- half hour. Thereafter, the animal received at a uniform rate an intravenous injection of 170 cc. m/2 (2.918%) sodium chlorid solu- tion. Time. Urine in ec. Remarks. 1 55 0.016 gram morphin hydrochlorid are given subcutaneously. 2.20 2.45 3.00 3.15 3.25 3.40 3.55 4.10 4.26 4.40 4.55 5.10 5.25 5.26 14.0 2.3 3.6 32.5 68.0 1 73.0 69.0 46.5 37.0 39.5 ' ? Yellow, turbid, alkaline to litmus. Ng albumin. No casts. After catheterizing, the animal is weighed. Yellow, clear, alkaline. No albumin. No casts. Right renal artery and vein and left renal artery are clamped. Clamps removed. Intravenous injection of m/2 NaCl begun. Filled with casts and red blood corpuscles. Fairly sets with albumin. Last portions clear as water. Clear as water. Acid to phenolphthalein, alkaline to rosolic acid. Faintest trace of albumin only. No casts. Occa- sional red blood corpuscles. Animal killed. Approximately 10 cc. urine lost in interval between stopping injection and making autopsy. Autopsy. — Weight 2554 grams! 19 cc. fluid cavity, grayish. Pleural and pericardial cavities are dry. found in peritoneal Kidneys are slightly Experiment 90 shows for what a long period the blood supply to the kidney may be cut off and yet the dangers ordinarily incident to such a procedure (partial to complete suppression of urine) be reduced by giving a concentrated salt solution. The experiment was really undertaken to indicate how the so-feared consequences of temporary occlusion of the blood vessels in operations on the kidney may be largely avoided or over- come — a discussion to which we shall return later. Experiment 90. — White and blue rabbit; weight 2344 grams. Fed hay, oats, corn, and greens. Urine obtained with, a catheter. The right renal artery and vein and the left renal artery and vein are NEPHRITIS 533 clamped for I5 hours. After an interval, 90 cc. m/2 NaCl are injected at a uniform rate intravenously. Time. Urine in cc. Remarks. 9.20 0.016 gram morphin hydrochlorid are given subcutaneously. Tied down. Catheterized. 9.50 10.05 1.3 Deep brownish-yellow. No albumin. No casts. Renal blood vessels are clamped. 10.20 10.33 10.50 11.05 11.20 11.35 Clamps removed. 11.50 12.05 12.20 12.35 12.50 1.05 1.20 1 35 1.55 Injection of m/2 NaCl into ear begun. 2.10 2.25 0.3 ] 2.40 2.55 0.4 1 0.4 Filled with albumin and hyaline casta (exclusively). 3.10 0.8 J Injection stopped. 3.25 2.6 Urine clearer. Filled with hyaline and granular casts. 3.40 2.8 Casts fewer. 3.55 1.5 No casts. 3.55 to 5.25 } "-3 } No casts. Red blood corpuscles found (traumatic). 5.40 1.5 No casts. 5.41 Killed. Autopsy. — Weight 2300 grams. 10 cc. fluid in peritoneal cavity. 1.2 cc. in right pleural cavity. Left unusually moist. Kidneys soft and somewhat gray. The secretion of urine in this experiment is represented graph- ically in Fig. 152. The first arrow indicates the point in the experiment when the clamps were removed. Up to the point of the second arrow no urine was obtained. At this time the sodium chlorid injection was started. The secretion of urine began less than half an hour afterwards. 5. Interpretation of Experimental Findings It behooves us now to study the experiments just described in order to discover the principles that underlie the results obtained, for only by knowing these can we hope to put them to any intel- ligent therapeutic use. In the light of the facts developed in 534 CEDEMA AND NEPHRITIS the earlier pages of this volume it follows that the living organ- ism represents in the resting state a series of colloids saturated with water. The slight (normal) secretion of urine' observed in an animal which is quietly tied into an animal holder represents a certain amount of " free " water still available for secretion. When we tie such a normal rabbit into an animal holder sufficiently snugly to interfere with its easy respiration the urinary output falls, as shown in Fig. 146. This is because by such means the accumulation of carbonic and other acids in its body is favored, which increases the capacity of the body colloids for holding water. None is therefore left over to be secreted as urine. The animal is, in other words, at once put into a condition similar to that attained after several hours if it is simply kept off food and water. Parenthetically we may add that a similar asphyxial state, induced here by tying the animal into a holder, is obtained by chemical means when we give a dose of morphin, „ cocain, atropin, or arsenic, an anesthetic like chloroform or ether, an excessive dose of alco- hol, or a nitrite. The same increased hydration capacity of « the kidney colloids and the body colloids generally is produced if we inject acid in- travenously. The acid acts upon them and they therefore absorb and hold on to any "^ water that may be given to them in these experiments along with the acid. Again none i remains over to be secreted by the kidney. * The animal retains the water and increases O in weight, m other words, it develops an " oedema." To the relief of all these conditions comes the administration of salt (along with the water). The salts — including sodium chlorid — reduce the amount of water that can be held by the body colloids generally, and so this freed water now becomes ■a !> o s P o (Wuo NEPHRITIS 535 available for urine. The kidney itself shares in this process and by shrinking admits a better circulation to be once more estab- hshed through it. The body now loses water, and so the animal loses weight, in other words, the oedema disappears. The kidney proteins become less soluble and so the albuminuria goes. Following a sudden apparent increase in number, the casts go. The apparent increase is due to the shrinkage, under the influence of the increased salt concentration, of the casts as they lie in the kidney tubules, followed by their easier and sudden washing out by the increased urinary flow. They tend, more- over, to change from the hyaline type to the granular, the latter representing reconversions from the hyaline under the influence of the salt. 6. Inhibitive Effects of Alkaline Salts on the Albuminuria of Hard Work We shall conclude this section by giving a concrete illus- tration of the fact that, by increasing the alkali-salt content of the body, the opportunities for the development of the signs of a nephritis are greatly reduced. For such a test the albumi- nuria that develops in athletes after hard work was used, and with the following results: In Experiment 63 on page 419 were detailed the quantitative findings regarding the excretion of albumin during an ordinary match basket-ball game, as determined by collecting the urine over the period of an hour and a half, in which time the game was played. In the following two experiments the urine was similarly collected, every precaution being taken to have the conditions for collection, regarding time, etc., as nearly the same as in the control. The athletes were under no restrictions regard- ing diet, the only difference being that in the two experiments now to be detailed, they took in addition to their ordinary food the juice of six sweet oranges in the first, aiid twelve in the second. The six oranges were consumed in the course of three hours pre- ceding the game; the twelve in the twelve hours preceding the game. Oranges were chosen not alone because they are palatable and so offer no difficulty in having the men take them, but because the salts contained in them have not only a decided capacity for combining with stronger acids but because the 536 OEDEMA AND NEPHKITIS citrates, malates, etc., are the very salts which act most power- fully in reducing the solubility of proteins in acids (the swelling of organs, etc.) . The game played in Experiment 91 was decidedly harder than that detailed for control purposes in Experiment 63, that of Experiment 92 fully as hard. The first five players were the same in all these three games, though the order in which they are numbered is not the same. Experiment 91. — Juice of six oranges fed the players. Urine collected for period of H hours, during which time the play occurred. Phosphotungstic-hydrochloric acid-alcohol reagent used in the Esbach albuminometer. Before the Game Player. Urine in cc. Nitric acid teat. Heat test. 1 2 3 4 5 232 72 30 85 280 Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative After the Game Player. Urine in cc. HNOs test. Heat teat. Esbach reading. Albumin excreted, in grams. 1 2 3 4 8 62 17 152 42 228 Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive 1.25 1.5 0.75 0.75 less than . 2 .078 .025 .114 .132 .046 Av. .079 Experiment 92. — Juice of twelve oranges fed each of the players. Urine collected for period of 1| hours, during which time the play occurred. Phosphotungstic-hydrochloric acid-alcohol reagent used in Esbach albuminometer. Before the Game Player. Urine in cc. Nitric acid teat. Heat teat. 1 2 3 4 5 6 73 170 187 6 23 62 Negative Negative Negative Negative Negative Negative Negative Negative ■Negative Negative Negative Negative NEPHRITIS After the Game 537 Player. Urine in cc. HNOi test. Heat test. ESBACH reading. Total albu- min excreted, in grams. 1 2 3 4 97 66 11 44 44 45 Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive 0.75 1.25 1.6 1.3 0.6 0.25 .073 .070 .018 .057 5 6 Av. .054 .028 .011 Player numbered 5 played first half only; number 6, second half only. Even when we count out the player in the control game who started with an albuminuria, and those in the succeeding games who did not play through, we still find that the albumin excretion, both so far as average concentration and average absolute amount is concerned, is decidedly lower after feeding citrus fruit than without such feeding. X ON THE TREATMENT OF NEPHRITIS 1. Introductory Remarks For the treatment of nephritis everything has been suggested from prayer to hot irons down the back. A discussion of the subject can therefore scarcely bring the mention of anything new that might be tried. It can-hope to be of interest or impor- tance only as, on the basis of the views entertained by an author regarding the nature and the cause of nephritis, he will assign to certain practices grades of importance different from those assigned to these same practices by another — a difference of opinion that may perhaps be carried to the point where the one will find virtue in procedures that another regards only as evil, and vice versa. Were we to formulate a general rule for the prophylaxis and for the treatment of nephritis we should evidently have to say that this lies in an avoidance and removal as far as possible, of every condition that favors the abnormal pro- 538 (EDEMA AND NEPHRITIS duction or accumulation of acids in the kidney, or of such other substances which in their effects on the colloids behave like acid. It therefore becomes necessary in approaching any case of nephritis first to get as clear a conception as possible of all the factors conspiring toward the production or maintenance of the abnormal content of acid and like substances in the kidney. Rarely shall we find but one factor active. To the toxic cause of a scarlatinal nephritis may be added that of an inadequate lung ventilation if the patient develops a bronchopneumonia. To the toxic nephritis of a pneumococcus infection already aggravated by a decrease in available lung for respiratory pur- poses, may be added an additional item if the patient develops a convulsion and so (through muscular work) produces suddenly an enormous additional amount of acid. The eclamptic patient who has just managed to drag through the last weeks of pregnancy faces death when the muscular efforts of labor or of a convulsion add more acid to that resulting from the intoxication of pregnancy. The patient with subacute or chronic nephritis, who, with per- sistent CBdema, some casts, albumin and a deficient output of urine, is permitted or insists on working about the ward, may be breaking the camel's back with this added straw of acid production by even such light muscular work. After we have, through restriction of muscular and mental effort, through rest in bed, through an adequate supply of fresh air, etc., removed as many of these conditions as possible, we turn our attention to combating those which we cannot remove. The rule to be followed then may be summarized in these words: Give alkali, salts, and water. To this may be added a fourth bidding: Give dextrose (glucose) or, if the conditions for its proper utihzation are present in the body, some other sugar or starch. The reasons for all this are, of course, apparent. The alkali is given to neutralize the acid present in abnormal amount in the kidney (and the other oedematous organs of the body) . The salts are indicated (and sodium -chlorid is no exception) because the various changes induced in such colloids as constitute the kidney by the action of acid upon them, are counteracted by adding to such acid any salt, even a neutral salt. We need to give water in order to have this present in the body over and above the amount necessary to saturate all the body colloids; other- wise we shall have no " free " water left over out of which to make NEPHRITIS 539 urine. Dextrose or other carbohydrates are necessary, not alone from a chemical point of view, in that an abnormal production and accumulation of acid in the body is frequently the conse- quence of carbohydrate starvation, but because the sugars are peculiarly powerful in reducing certain types of increased hydra- tion in proteins not produced by acids. The exact methods to be adopted, the aggressiveness and persistence with which this simple rule is followed, and the results obtained by so doing must evidently depend upon the condition or combination of pathological conditions that our patient may have developed, and which we hold responsible for the abnormal production or accumulation of acid and like substances in his kidneys. Evidently an anesthesia nephritis with suppression of urine will call for a more aggressive therapy than a nephri- tis secondary to a slowly progressing arteriosclerosis. On the other hand, if we succeed in getting our first nephritic over his immediate kidney symptoms we may make a hopeful prognosis, for when he has exhaled his anesthetic he has rid himself of the condition that was responsible for the abnormal acid content in his kidneys. But in our second nephritic so hopeful a prognosis cannot be made, for while we may also benefit him, he continues to carry the original condition that brought him to us — his arteriosclerosis — even after we have treated him. 2. Diet in Nephritis We have long paid attention to the diet in nephritis. Clearly, the direct consumption of acid as such by the nephritic is con- traindicated. The mineral acids which would be worst in this regard do not enter into our foods to any appreciable extent, though in the forms of fruits, sour wines, etc., not inconsiderable amounts of organic acids are swallowed. With the exception of benzoic, oxalic and tartaric acids, most of these undergo oxidation in the body rather easily, being converted into carbonic acid, which is readily excreted. The organic acids are therefore less poisonous than might at first appear. But from this is not to be concluded that they are of no importance at all. Not only are certain " weak " organic acids (notably, tartaric, acetic, and lactic) quite as active physiologically as the " stronger " acids, but consumed in excessive amounts they are not without 540 (EDEMA AND NEPHRITIS effect even In " normal " individuals, as witness the oedemas observed in children that are fed buttermilk/ the urticarias following the consumption of excessive amounts of grapes,^ etc. In nephritics these organic acids assimie a yet more important role, for these individuals have, for various reasons, a decidedly decreased capacity for properly oxidizing them. Still it is not to be concluded that every article of diet which yields some of these organic acids is therefore at once to be excluded. Such measures all too often lead to the absurd food restrictions with which patients are constantly tortured. Their importance need only be recognized to the end that an adequate amount of alkaU, for example an alkaline water, be fed along with the food containing the organic acids. In the metabohsm of proteins not inconsiderable amounts of acid are produced.^ Herein is to be sought at least part of the explanation of why a restriction of the proteins in nephritis is of use. If im,properly used in the body, the proteins may, moreover, yield other than the normal amins, in other words, toxins which like acids are capable of increasing the hydration capacity of the body proteins and otherwise influencing their state. In the absence of sufficient carbohydrate, the fats may also yield abnormal or abnormally great amounts of such acids as diacetic, beta-oxybutyric, etc. But before one proceeds to a too drastic revision of the dietary, a process in which we are particularly liable to eliminate the proteins too vigorously, the absolute amounts of acid formed by the various constituents of the food should be considered. When this is done it will be found that the evil consequences expressed in terms of a direct acid yield from the proteins of our food, for example, are small compared with those that may be calculated from a bottle of dry wine. Consideration of the acid content of alcoholic beverages 1 Ernst Schloss: Deut. med. "Wochenschr., No. 22 (1910). Schloss does not, however, consider this an oedema due to feeding acid, but as " idiopathic." He found it to disappear on administering calcium salts. * Personal observation. The urticaria disappears as soon as calcium salts are given, or fails to appear if such are consumed with the grapes. 'G. VON BtTNGE: Zeitschr. f. Biol., 10, 111 (1874); N. Lunin: Zeitschr. f. physiol. Chem., 5, 31 (1881); Emil Abdebhalden: Biochem. Zentralbl., 2, 257 (1904). See also the important studies on the balance of acid-form- ing and base-forming elements in food by H. C. Sherman and A. 0. Gettler: Proc. Soc. Exp. Biol, and Med., 8, 119 (1911); N. R. Blatherwick: Arch Int. Med., 14, 409 (1914). NEPHEITIS 541 helps us, moreover, to understand why some of these exercise a worse effect in nephritis than others. We have long recognized that the alcohol content is not alone responsible for the effects of alcohoUc beverages in kidney disease, for while it is true that in large doses alcohol allies itself with the general anesthetics, small doses do not interfere with the oxidative reactions in the hving cells, but rather favor these (and so kidney function). But whatever is fed, it is clear that all the acid effects of the food need not appear if we will take the precaution of seeing that the diet contains sufficient alkali to neutralize the acid. Just as certain features of a dietary thus favor the develop- ment of a nephritis, so also do others counteract it. Most notable here are the long-observed beneficent effects that follow the use of alkalies and the substitution of a more strictly fruit and vegetable diet for our ordinary mixed diet. The bases present in fruits and vegetables are from the start combined with weak organic acids which in the body are largely oxidized to car- bonic acid. How successfully a diet rich in fruits and vegetables counteracts even the normal tendency of the body to run toward the acid side is a matter of common knowledge to any physician who has watched the urine of his patients turn from its normal acidity to an alkahnity, when ordered from an ordinary mixed diet upon one richer in the vegetables and fruits. But this neutralization capacity of the fruit and vegetable diet for acids is not the only factor which accounts for its beneficent effect. We learned earlier that the solubility of protein in any acid is markedly reduced by salts. Not only are the vegetables rich in salts, but they are rich in the very ones which act most powerfully in reducing the solubility of the protein. So we may find in this fact, along with what has already been said regarding the capacity of the vegetable diet to neutralize acids, a satisfactory scientific foundation for the reduction of the albuminuria in nephritis, when a diet rich in vegetables follows one in which these were not so abundant. Such salts also serve to reduce the size (swelling) of the kidney, and as we have seen, they practically prevent those precipitation effects in the kidney cells (granule formation) that are characteristic of the early changes of nephritis from a morphological point of view. Speaking generally, a diet high in fruits and vegetables also means that the individual is. consuming more water. The 542 (EDEMA AND NEPHRITIS effect of this will be discussed later, but even now it may be pointed out that such will aid our nephritic, because it brings to his kidney the " free " water from which alone he can make urine. Practically expressed, I let the nephritic eat pretty much as he pleases. In the ambulatory subacute or chronic cases associated with blood vessel disease or secondary to infections of the kidney I permit a moderate meat ration on condition that the patient will eat his vegetables first. This trick in- creases his alkali intake at the expense of the acid side of his dietary. The soups are not forbidden, because the salts and water thus consumed offset largely any bad effects which the accompanying meat extractives are imagined to have. If the proteins are thought to be giving rise to " autointoxication " products in the alimentary tract, I use small doses of calomel or occasional large enemas of salt solution, sodium bicarbo- nate or soap. Enemas of sodium bicarbonate are especially to be recommended. Calomel in small doses does not injure the kidneys, but exercises only the dehydrating effect upon all the body colloids, which increases secretion from the ali- mentary tract and the kidneys. Since even large amounts of fruits and vegetables may not prove sufficient to keep the patient well supphed with alkaU, I urge the daily use of suffi- cient natural or artificial alkaline waters, either carbonated or still, to accomplish this purpose. A patient is getting enough alkali when his urine is kept persistently neutral to litmus. He should be given neither more nor less than this amount. I teach the ambulatory nephritic to test his own urine and to increase or decrease his alkali intake as necessary.^ The bedridden cases are treated from a dietary point of view in the same fashion. More harm is done these patients by underfeeding them than by overfeeding. The caloric needs of the individual must be covered daily. A " starvation acidosis " is as bad as any other kind. Orange juice, grape fruit juice, lemonade, cereals, milk and cream, vegetables, fruits, and ' To escape the effect of alkali upon the stomach and obtain it in the small intestine E. G. Ballenger and O. F. Elder (Jour. Am. Med. Assoc, 62, 197 (1914) ) suggest its administration in admixture with mutton suet and paraffin. They further make use of the clever expedient of adding a little phenolsulphonephthalein. The patient then knows he is consuming enough alkali as soon as he voids a constantly pink urine. NEPHRITIS 543 various meats and fish should in the order named be urged upon the nephritic. Such foods as will carry it should have sugar added to them, for the first items in which our diets are likely to be short are the carbohydrates. These patients also are given an alkaline water of some sort. The natural or artificial alkaline spring waters are perhaps most easily borne. If the patient will tolerate it, 0.5 to 1.0 gram of sodium carbonate or sodium bicarbonate may be added to each glass of such alkaline water or to plain water. Some patients who caimot tolerate the carbonates will take sodium citrate, sodium tartrate, 1 sodium acetate or other salts of a strong base with a weak acid in half-gram or larger doses every hour, either when dissolved in water or in capsules followed by water. Cal- cium hydroxid can easily be given by mixing lime-water with milk. As the salts of the bivalent metals are particularly active in decreasing the hydration-capacity of the body-colloids, the administration of magnesium oxid or milk of magnesia up to the point where two or three easy movements of the bowels are obtained daily, gives good results. The soluble salts of calcium and strontium, as the iodids, acetates, lactates or citrates are to be highly recommended. They are, unfortunately, absorbed rather slowly, but once in the body they keep the hydration capacity of the body colloids low. The same is true for the salts of iron which the older doctors used so effectively. For reasons discussed before, I do not think that table salt should be restricted in a threatened or developed nephritic, but on the contrary should be urged upon the patient. Food serves as a natural carrier for this. In the form of salt meats and salt fish we can easily get considerable quantities of sodium 1 Tartaric acid is little, if at all, oxidized in the body. Its appearance in the body therefore draws upon the alkali contained therein and empha- sizes the necessity of seeing to it that foods rich in this acid (as grapes) are accompanied by sufficient fixed alkali to do away with the acid effect. Underhill, Wells, and others have reported deleterious effects upon the kidneys from feeding tartrates, but W. E. Post (Jour. Am. Med. Assoc, 62, 592 (1914)) noticed none when up to 24 grams were given to nephritics; in fact he noticed nothing but good results to follow the use of sodium and potassium tartrates — as the older doctors have so long known. William Salant (Jour. Am. Med. Assoc, 63, 1076 (1914)) noticed poisonous efTects in rabbits on an oat diet but not on a diet of carrots. But an oat diet as Weiskb has shown is an "acid" diet and one insufficient in calcium. As Salant also observed, calcium administration does away with the poisonous tartrate effect. 544 (EDEMA AND NEPHRITIS chlorid into our patient, and by placing a salt shaker at hand, he can liberally increase his intake by dusting it on his vegetables, his fats, and such proteins as we allow him. Through the diet alone we can, therefore, do a great deal to keep the intake of alkali and salts high. 3. Water Consumption in Nephritis The question of water consumption resolves itself into two parts — ^into the use of water in cases where a nephritis is Hkely to arise, and into its use in an estabhshed case. The reasons are obvious why water needs to be given and why we wish to give it at a uniform rate in the largest possible amounts that will not injure the patient. Not only must the patient get water in order to have the wherewithal to make urine, but whatever the conditions lying behind the nephritis, the pathological state amounts in the end to an intoxication. This is strikingly true of course in the infectious diseases or in an eclampsia case. Here the whole organism is suffering from the effects of a poison. The effect of that poison depends not alone upon the length of time that it acts upon the organism as a whole or any individual part of it, but upon its concentration at any one time. If now our interest centers upon a toxic effect that such a poison may have upon the kidney, and we are anxious to protect this organ, it is clear that the concentration of the poison must be kept as low as possible in it. To do this only two possibilities are open, and when we cannot control the factor of poison produc- tion, we can hope to cut down the effect of the poison only by keeping what is produced as dilute as possible. This calls for the giving of water. In this connection the practical point should be remembered that in ordinary practice an ever so patient administration of water through the day is likely to be neglected in the night. As toxin production does not cease with nightfall, it is clear that water administration also should not, otherwise we are hkely to lose in a few hours at night what we cannot subsequently regain in days, if at all. The administra- tion of water cannot therefore be left to the haphazard desire of the patient. It must be insisted upon in a regular manner. A good rule is to give a glass every hour day and night. At this point we are likely to be met by the argument that NEPHRITIS 545 while such a water therapy is accepted as advisable in the toxic nephritides, those associated with heart lesions, etc., are not to be similarly treated. Let us first point out that these too are toxic nephritides — the patient with a broken heart com- pensation, or a compressed lung due to a carcinomatous pleurisy, and albmnin in his urine shows this (according to our views), because the acid content in his kidneys is abnormally high. The more the concentration of this can be reduced the less will be its effect on the kidneys. Thus far, therefore, he needs water quite as much as the nephritic who is such in consequence of an infectious disease. But it will be argued that the giving of water increases the work of the heart in these cases, and so is bad. Let us consider this problem dispassionately. The belief that the giving of water increases the work of the heart is based upon the notions of urinary secretion which imagine that water is pushed through the kidney cells by gross mechanical means. And so it is reasoned that the more water that is given, the more push is required, that is to say, the more blood pressure, and so the more work from the heart. Actually such a belief lacks every experi- mental support. If one fact regarding urinary secretion stands out as well established, it is that the forces active in producing the urinary secretion lie within the kidney itself. The only thing, therefore, that we might call upon as responsible for increasing the work of the heart would be some product of the work of the kidney in separating urine from the blood. We have a right to consider here the effect on the heart of the extra carbonic acid produced whenever the kidney functions. But the effect of this cannot be greater than that of an equal amount produced normally, or in a given case of kidney disease than an equal amount produced in some other organ. And when com- pared with the total amount of carbonic acid produced from all other sources, the amount produced in the kidney because of the consumption of a few extra liters of water is small indeed. Even if the carbonic acid production by the kidney did mark- edly increase the work of the heart — a view for which we have not a single unequivocal clinical observation — it would still have to be proved that such an effect is worse than that result- ing from An accumulation of acid in the kidney (and in the body generally) because of its insufficient elimination for lack 546 (EDEMA AND NEPHRITIS of water. As a matter of fact, we know that kidney lesions and the administration or non-administration of water to nephritics have nothing to do with the heart changes so often observed in patients with kidney disease, for in the frankly parenchymatous types of nephritis, in which water elimination is most decidedly insufficient, and in which it would be expected in consequence that the heart would be working hardest in order to rid the body of any consumed water, high blood pressure, and heart comphca- tions (hypertrophy) are most conspicuously absent. Let us now ask if in the so-called chronic interstitial types of nephritis water should also be freely urged. As already pointed out, the work done by the heart in pumping blood through the blood vessels depends upon the length, diameter, and elas- ticity of the blood vessels, and upon the viscosity of the blood. The giving of water certainly does not affect the first three factors. So far as viscosity is concerned, it can only decrease this and so diminish the work of the heart, as diluting a syrup with water makes it easier to draw through a straw. The only objections that may be raised against a too vigorous administration of water, it seems to me, are two. The first is associated with the fact that when the hydration capacity of any tissue has been increased (as that of the kidney in nephritis) the giving of water permits it to swell; the second with the fact that pure water in washing through the kidney washes out not only poisonous substances of which we would be rid, but good salts of various kinds in addition. Thus, it might be reasoned that to give water in the acuter forms of nephritis would be to aid this swelling. Such swelling of the cells, so far as the cells themselves are concerned, can hardly be considered serious, any more than a moderate oedema of any tissue is in itself particularly destructive to the tissue. But in the case of the kidney a complicating circumstance arises which does make such a swelling dangerous. This resides in the fact that the capsule of the kidney is not as expansile as the rest of the kidney substance. As the kidney substance swells, this tends, therefore, to press upon the blood vessels and retard the circulation of the blood through the kidney. This condition actually comes to pass. Thus, a kidney already nephritic, say from the toxin of an infectious disease or an anes- thetic, aggravates its state by hampering its own blood supply. The washing out of salts from the kidney acts in the same NEPHRITIS 547 direction, for, as already noted, these tend to counteract the swelling. Our problem might, therefore, seem to become that of balancing the good effects of water against certain bad ones. Actually it is much simpler. We give the kidney the benefit of the virtues of water, while we protect it at the same time from the dangers associated therewith by giving along with the vjater properly chosen salts in sufficient amount, 4. The Role of Salts in Nephritis In our every-day diet we never seriously consider whether we drink distilled water, tap water, or a table water. And from this point of view we might be inclined to ignore the exact composition of the liquids consumed by a nephritic. But let us look at the problem a httle more critically. The normal individual does not really with impunity ignore the matter. It only seems so. In his food he consumes large quantities of various salts, so what he really obtains in any longer interval of time is a salt solution. The proper regulation of this — that is to say, the continuous consumption of a proper salt solu- tion which in its turn maintains a proper salt concentration about the various cells in the body — is accomplished through the " taste " of the individual. If his salt consumption has been too high, he craves fresh water, and so washes out the excess. If, on the other hand, he has lost too great an amount of salt he consumes more and makes up the deficit. The truth of these statements is attested by the most varied scientific and social facts. The animal treated with a strong salt solution makes desperate efforts to get at water, and the salt-starved animal Ucks the sides of its cage and laps up its urine. The American retailer of beer feeds his customers salt meat and fish gratis. This gives them a thirst which they satisfy with beer, and over- drinking, they turn about and demand more salt food. If these facts are borne in mind, it becomes easy not only to devise a therapy which from present evidence promises most in the relief of nephritis and allied conditions but to recognize the merits of long recognized therapeutic procedures. The milk diet has, not without reason, been popular. By giving milk we give a patient a useful balanced ration of fat, carbohydrate, and protein. But we do more than this — we give 548 (EDEMA AND NEPHRITIS water and salts. The water helps to wash out poisons and the salts contained in the milk have a concentration which just suffices to do away with the effects of giving an equal amount of water pure. Similar reasoning explains the beneficent effects of giving " physiological " salt solution in large amounts by rectum, intravenously or subcutaneously, in various acute infections It is again the combined effects of much water to wash out poi- sons, and enough salt to counteract that accidentally lost ^ by the same process that washes out the poison. When in spite of such procedures the signs of a nephritis develop we need to press more salt (alkali and sugar). In the nephritic, because of the accumulation of substances in parts or all of his kidneys or other organs, which increase their hydration capacity, these are abnormally swelled. To reduce their swelling a more than " physiological " concen- tration is demanded. To accomplish this the patient must consume a proportionately larger amount of salt. This is a matter which for various reasons we cannot leave to his taste alone. We need in consequence to give him specific instruc- tions as to how to increase his salt intake. As previously noted, sodium chlorid forms no exception to this rule. There are, however, many other and more powerfully effective salts, and herein lies the scientific reason for the so-long established custom of giving these patients the acetates, tartrates, cit- rates, etc., of sodium and potassium, as well as magnesium sulphate, magnesium citrate, Basham's mixture, etc. These are all salts which in low concentration produce great dehy- drating effects in all the tissues of the body. Such salts there- fore permit us by keeping the hydration capacity of the body colloids low, to get the beneficent washing-out effects of water without its deleterious consequences. ' We have become all too inclined to consider everything that comes out in the urine as something that the intelligence of the kidney has found harmful to the body. It is scarcely as wise as this. It is rather hard to see, for example, why in a salt-starved animal that is being given water, the animal continues to eliminate some salt in the urine up to the moment of death, when it is this very elimination that is killing the animal. NEPHRITIS 549 5. More Aggressive Methods of Alkali and Salt Administration _ Let us now imagine that in spite of these procedures the nephritic's symptoms do not improve. Many patients cannot long keep up a high consumption of alkali and salt by mouth. They may begin to vomit; or a " uremia " with vomiting may develop so that we may fail in our therapy. Or let us imagine that the symptoms are rather severe from the start, so that alkali and salt by mouth seem inadequate. What are we then to do? We continue to get as much use out of the gastric route as we can, but we use the rectum also in order to get an absorption of alkali, salt and water. Here again a more than " physiological " concentration of salt is demanded. A " hypertonic " solution is necessary. To reduce the oedema in the kidney or in the brain, in the optic nerve or in the tissues generally, we need to try to increase, at least temporarily, the absolute concentration of salt in the whole body. We therefore use a hypertonic salt solution to which has been added an alkali. Obviously, when using such a hypertonic solution by rectum we do not allow pure water or any solution of low concentration to be taken by mouth. The patient may wet his mouth to relieve his sense of thirst, but no more, otherwise we only reduce the concentration and so the therapeutic value of the solution we are administering by rectum. The following formula,^ or an even stronger one, works very well: Sodium carbonate (Na2C03- IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. Simple as is this formula, care must be taken in its prepara- tion if good effects are expected. While we need not for rectal use insist upon the same grade of care that is necessary when such a solution is to be injected intravenously, it is well to consider everything even here. Sodium carbonate is used, not bicarbonate. The carbonate is physiologically more effective than the bicarbonate, for one of the acids which are produced in the body is carbonic acid, and sodium bicarbonate is already saturated with this. In con- sequence, it cannot act as a carrier for it. '■ This is the solution frequently called by my name. 550 CEDEMA AND NEPHRITIS The chemically pure, crystallized sodium carbonate (Na2C03 • IOH2O) or the monohydrated form (Na2C03H20) is to be insisted upon. Two other foVms of sodium carbonate are found in the market, the dry (Na2C03), and the so-called dry or " dried." The "dried " salt found on the ordinary drug shelf contains approximately two molecules of water of crys- stalhzation (Na2C03-2H20). We are inclined to advise against all except the large crystalhzed form or the monohydrated form; but whatever salt is used its content of water of crystalliza- tion must be remembered, otherwise a solution of a different strength from that which has been found most useful will be obtained. The proportionate amounts of these four salts that may be used are to each other as their molecular weights, or, in definite terms : 10.00 grams Na2C03- IOH2O (molecular weight 286) crystallized "sal-soda " = 4 . 95 grams NazCOa ■ 2H2O (molecular weight 142) " dried " 4.33 grams NazCOs -1120 (molecular weight 124) " monohydrated " = 3.71 grams Na2C03 (molecular weight 108) really " dry." The sodium chlorid-sodium carbonate solution should be made up in distilled water and filtered. If the salts used are pure and the whole is properly prepared the resulting solution is perfectly clear. Rectal Injection of the Solution. Unless the patient is mentally incapable of comprehending what we say, it is well to explain to him before an injection is made just what we desire to accomplish and so secure his cooperation. As the solution is hypertonic and contains alkali in addition, it not only irritates the rectum somewhat, but leads temporarily to a secretion of water into the rectum while the salt and alkali are being absorbed.^ The patient, in consequence, has a desire to go to stool, which after we have once permitted him to satisfy, we wish him to overcome. By obtaining his cooperation, the solution is retained for longer periods of time, or entirely, and correspond- ingly we get a more perfect absorption of the alkali and salt. To inject the solution we may make use either of a con- tinuous drip method, or inject larger quantities at varying intervals of time. It is not necessary first to cleanse the rectum locally, and especially are we not to try to accomplish this end • No solution is absorbed or secreted " as such." See page 274. NEPHRITIS 551 by a previous administration of cathartics. These methods only increase the irritabihty of the rectum. It is well, of course, if the lower bowel is empty (as immediately after the patient has had a stool). But if no movement has occurred recently we make our injection just the same. Perhaps the patient will then have a stool from the alkali-salt solution injected. This cleans the rectum, and we begin again. For administration of the alkali-salt solution the patient should first be comfortably arranged in bed. He should lie on his left side on a rather hard bed and with no pillow, or only a small one, under his head. His hips may be elevated slightly to give still greater pitch to the rectum, but this should not be at- tempted unless the patient can be made 'perfectly comfortable. We are now ready to inject the solution, and may choose either the slow drip method or the fractional instillation method. The choice from the patient's point of view is about evenly divided between the two, and he should be consulted and his wish followed. The fractional instillation method will be de- scribed first. For this the apparatus shown in Fig. 153 may be used. A is a funnel that has a capacity of not less than 250 cc. 5 is a soft rubber tube at the lower end of which is the pinch clamp G. 2) is a short glass tube that connects B with the soft rubber catheter or rectal tube E. The solution to be injected is heated to 40° C. (110° F.) and the patient being in position, the catheter is lubricated with petrolatum. Into the funnel are now poured 250 cc. of Figure 153. the solution, and the stop-cock is temporarily opened so as to drive the air out of the tube and catheter. The catheter is then gently inserted well into the rectum and the funnel is emptied by again opening the pinch-cock. The short glass tube will inform the operator when the last portions of the mixture are flowing into the rectum, when the hold on the pinch-cock is released. In this way no air will be allowed to enter the rectum and balloon it, and extra irritation from this source is avoided. If the patient's lower bowel has been empty, the injection is 552 CEDEMA AND NEPHRITIS usually easily retained, and may be repeated in half or three- quarters of an hour. If the bowel was filled with fecal matter the first injection serves to bring on a movement, and the cleared bowel will subsequently absorb better. The injections may be repeated as often as the symptoms of the patient demand it, or until the patient finds them difficult or impossible to retain. A period of rest should then be given. If our case is so urgent that this cannot be allowed with safety, then the solution must be given intravenously (see below). For the continuous drip method the arrangement shown in Fig. 154 works well. A is an ordinary half liter or liter graduated irrigating vessel with a side tubulation. It connects through D » B Figure 154. the rubber tube B, carrying a pinch-cock C, with the glass insert D, in which lies a thermometer. The insert connects with the soft rubber catheter or rectal tube E. F is a glass insert which permits observation of the rate at which the solution is dripping into the rectum. From 1 to 4 drops a second should enter. This is about as high a rate as the patient can stand without rejecting the fluid. Roughly, this corresponds to an injection of from 240 cc. to 960 cc. per hour. The injection fluid is retained best if it is delivered into the patient at not less than body temperature, and 40° C. (110° F.) is better. For this reason the thermometer in D, located as near the rectum as possible, is of great convenience. As the solu- tion slowly passes out of A through the tube, it falls in temper- ature. The vessel A is therefore conveniently filled with the solution at a temperature somewhat above 40° C. Or one can NEPHRITIS 553 set this vessel into a second one containing warm water, or place the tube B in warm water, or cover it with a blanket by way of maintaining the solution at a proper temperature as con- venience and the ingenuity of the medical man dictate. In hospital practice, thermostatic devices heated by electricity or gas may be conveniently installed. Amount and Time Interval. How much of the solution may be given by rectum and how long do we continue with it? The answer to this is found in the condition of the patient. So far as I have been able to observe, no harm can be done by indefinite use of the solution. A full •physiological effect is obtained when the patient is kept free from the various signs and symptoms of nephritis, and the urine is per- sistently neutral in reaction toward litmus. Such a full effect may be obtained in a day or two, or it may require several. The symptoms may have cleared entirely even before the saturation of the body with alkali has been carried to the point where the patient secretes a persistently neutral urine. His improvement will by itself suggest a reduction in the number of injections or their entire abandonment. Individuals differ greatly in their behavior toward these injec- tions. I have seen patients bear them for weeks at a time with- out complaining, and without ever rejecting them. Others will insist from the first that they cannot hold them. In this con- nection it is well to point out that if these hypertonic salt-alkali mixtures are retained for any time at all, say even for an hour, they do much good. When at the end of such a period the patient rejects some fiuid from the bowel, this is not the same as that which was introduced, simply minus a certain quantity that has been absorbed. Hypertonic sodium chlorid-sodium carbonate mixtures are not absorbed as such. The salt and alkali are absorbed out of the solutions while water is being secreted into the bowel. Therefore, if the solution is retained even for a short time, the patient will have increased his body-content of alkali and salt, which is the whole purpose of the therapy. It is well, if possible, in the first twenty-four hours of treat- ment to have two liters (about two quarts) of the solution retained by the patient. For a few hours after starting the injection 554 (EDEMA AND NEPHRITIS it is well to give him but little water by mouth, for we wish to obtain as great a shrinking effect upon all the tissues of the body (including the kidney) as possible. But as soon as we see the urinary output begin to increase and rise toward a more normal point, we urge the patient to drink water, so as to have free water available for urine. I find good practitioners overlooking this point constantly. While some water is introduced with the hypertonic salt solution, most of the " free " water that is lost through the kidney after absorption of the hypertonic solu- tion has been taken from the tissues (which may be oedematous). But it is self-evident that this robbing of the tissues can only go on for a limited time. Then we have to supply water from the outside to keep up the urinary secretion. The thirst developed by the patient when treated as here suggested with alkaline hypertonic salt solution helps us in this regard, and we are glad to satisfy it as soon as we find that his kidneys are better able to secrete water. If for any reason (as when the patient is comatose or vomit- ing) we cannot feed enough water by mouth, then we can use the rectum. After the kidneys are functioning in a more normal way we may substitute for the hypertonic sodium chlorid solu- tion one more nearly isotonic with the body fluids. A solution of sodiimi bicarbonate containing 12 to 14 grams to the liter of distilled water does very well. This is about equal in concen- tration to a " physiological " sodixmi chlorid solution and pref- erable to it, for, while not as powerful as the carbonate, sodium bicarbonate neutralizes acids stronger than carbonic and so helps to maintain the neutral reaction of the urine. Sodium bicar- bonate may be injected in indefinite amounts without giving rise to rectal irritation. I use it much on this account in children and in protracted nephritides in adults, where I frequently raise the concentration to 18 or 20 grams to the liter. The desire to introduce into the nephritic the bivalent metals which dehydrate the body colloids far more than do the uni- valent metals cannot be easily satisfied. The reason for this is obvious. Their carbonates and hydroxids are largely insoluble, and so the administration of bivalent metals such as calcium or magnesium along with carbonates or hydroxids is impossible. The only schemes that I have found of service are limited to patients with mild nephritis and to those recovering from the NEPHRITIS 555 severer types, where lime water may be added to the milk con- sumed by the individual, or he be given magnesium oxid, milk of magnesia, and soluble calcium or strontium salts by mouth. For rectal injection a " physiological," 0.85 per cent sodium chlorid solution, to which 0.1 per cent calcium chlorid is added, also works well if alkaline solutions have not been used for some hours previously. It is well here to consider the effects of the sweating so com- monly practiced in nephritis. It has been used so long and with such good results that its usefulness cannot be doubted, and yet I question whether what it accomplishes is really correctly understood. For the most part sweating (as purging) is sup- posed to act as a partial or complete substitute for kidney function, it being held that with the sweat, poisonous substances which should be eliminated through the urine are carried off through the skin. Since the sweat chemically contains much the same sort of material as urine, this belief is, of course, partly justified, but even with copious sweating we get in the aggregate but little through the skin. Sweating is an effective method of dehydrating the swollen body colloids and the relief of coma, of headache, of high blood pressure due to cerebral oedema, of vomiting and Chetnb-Stokes respiration and of a generaUzed oedema, together with evidence later of a better urinary secre- tion (following dehydration of the swollen kidney) a better general circulation, etc., are more logically explained through this dehydration than on the older basis which held that " nephri- tic toxins " assumed to be responsible for these various signs are thus lost vicariously through the skin. Sweating dehydrates all the organs of the body and permits a better circulation to take place through them. It is self-evident how sweating may relieve a swollen kidney and how in the end it therefore accomplishes what alkalinization and increase in salt concentration are trying to do. Only in sweating two things must be kept in mind. While it is true that by this means we dehydrate the body tissues, sweating does nothing to meet the causes for the increased colloid swelling. Second, when we carry off water through the skin (or bowel) we must not expect the same water to be able to come through the kidney. Diuresis follows sweating only secondarily. Only when after the shrinking consequent upon the sweating a better 556 (EDEMA AND NEPHRITIS blood supply has been assured the kidney may this recover sufficiently to be able to put out later water brought to it in a " free " state. 6. The Treatment of Severe Cases of Nephritis Under this heading we shall consider those patients in whom we encounter such especially alarming signs and symptoms as great or complete suppression of urine, rapidly progressing optic- nerve changes, persistent headache and nausea, vomiting, con- vulsions, stupor and coma, great quantities of albumin in the urme, etc. While we realize that a complete explanation of the nature and of the cause of these various clinical signs cannot be summed up in any brief statement, we beUeve that as previously em- phasized, an essential, if not the essential element in all of them is an oedema of the affected part. This oedema is represented physico-chemically by an increased colloid swelling of the tis- sues involved, and as responsible for such, we hold the. abnor- mal production or accumulation of acid in the part, either alone or in conjunction with such other substances as are also capable of increasing the hydration capacity of the tissue colloids. Or, to repeat, what we call the serious complications of nephritis are not really complications secondary to this pathological entity, but are manifestations in other organs of the body, of the thing which in the kidney we call nephritis. Just as the nephritis is in large measure an oedema of the kidney, so the optic nerve swelling and the " retinitis " of nephritis are cedemas of the optic nerve and of the retina; the headache, convulsions, and coma are manifestations from an oedema of the brain, the persistent nausea and vomiting of central origin, manifestations from an oedema of the medulla; the generalized oedema, an expression in the body tissues of what in the kidney we call nephritis. The same intoxication or the same vascular disease, underlies all these changes, and it is a mere accident that one nephritic will show particularly prominent eye symp- toms, another a great generalized oedema, while a third will call us to his side by a convulsion. For any one of a number of reasons, the oedema may become particularly prominent in his optic apparatus, or in his body tissues generally, or in NEPHRITIS 557 his brain. If we bear these facts in mind, it will serve to indi- cate why T believe that any or all of these signs demand the same general treatment, and why, when we succeed in com- bating a particularly prominent one, we find that we have suc- ceeded in combating all the rest as well. Whether the particularly alarming symptoms spring from the kidney itself (a suppression or urine, a great albuminuria, etc.), or whether they spring from the brain (convulsions, stupor, coma), the eye (" papilHtis," "retinitis," partial blindness), or the medulla (nausea and vomiting), the purpose of our therapy is the same — we wish to stop and reduce the swelling of the involved tissues. Naturally the best and quickest way to do this is to inject something into the blood, and what this something must be we have already discussed. Preparation of an Alkaline Hypertonic Salt Solution for Intravenous Injection The following formula, already recommended above, does very well : Sodium carbonate (NajCOs • IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. It is well to have this solution ready for immediate use, for its preparation in sterile form takes time, and need for it when it arises is urgent. What we have said regarding chemicals holds here also. Only chemically pure salts and freshly distilled water are to be used. The crystallized sodium carbonate con- taining ten molecules of water of crystallization is recom- mended. If some other form of the salt is used, less than 10 grams to the liter are to be used, as discussed on page 550. The finished solution as injected into the patient must be per- fectly clear and sterile. Simple as it would seem to be to obtain this result, it is not always secured even when its preparation is left to trained helpers. Therefore, I may be pardoned for detailing the following rules for its preparation in containers that make it available for immediate use. Trouble arises from the fact that alkaline solutions cannot long be kept in contact with ordinary glass containers without 558 (EDEMA AND NEPHRITIS reacting with the glass and so leading to a separation of insoluble silicates. Jena glass flasks resist better than other material, and so the finished solution may be sterilized and kept in these. For this purpose, it is only necessary to dissolve the sodium chlorid and the sodium carbonate in the necessary amount of freshly distilled water and filter the solution through moistened filter paper (in order to get no shreds into the solution) into thoroughly cleaned Jena flasks which have been rinsed in dis- tilled water. It is convenient to have two liters of solution in each flask. The flasks are stoppered with gauze-wrapped cot- ton stoppers, and may be sterilized in the ordinary way by boil- ing. This scheme works well in hospitals or anywhere where storage room is plentiful. When needed for injection, this solution may then be poured into any one of the properly sterilized intravenous injection apparatuses that abound in the market. If the solutions show a precipitate in spite of the use of Jena glass containers, the clear solution may be decanted or the whole may be fil- FiGUEB 165. tered through a sterilized funnel into the neck of which has been forced a little sterile glass wool. Or a sterilized glass bulb insert of the type shown in Fig. 155, into which has been forced some glass wool may be used in the delivery tube of the injection apparatus. As the carbonates and hydroxids of the polyvalent metals are all insoluble, every piece of any injection apparatus must be rinsed and sterilized only in distilled water. When ordinary tap water is used calcium and magnesium precipitates cloud the injection mixture. In my own experience I have found it convenient to make up the sodium chlorid-sodium carbonate solution in concen- trated form in ampoules, and then mix this with enough freshly distilled water to yield the proper injection mixture at the time it is needed. One proceeds as follows : Any desired number of multiples of 10 grams of sodium car- bonate (Na2C03 ■ IOH2O) and 14 grams of sodium chlorid are dissolved in enough water to make 60 cc. of finished solution. This solution is filtered and then sterilized by boiling.^ One or 1 In my former book on nephritis and in a paper or two I cautioned, against the use of excessive heat in sterilizing these carbonate solutions. The cau- NEPHRITIS 559 Uc \ more ampoules of the type shown in Fig. 156, A, and of about •60 cc. capacity (if these are not available, bottles will do as well) are thoroughly cleaned, rinsed in distilled water, and then boiled in distilled water to sterilize them. Into each of these is then filtered through a small sterihzed funnel plugged with glass wool, 60 cc. of the concentrated sodium car- bonate-sodium chlorid mixture. When the ampoules have been filled, they are sealed in a flame, as in Fig. 156, B. If bottles are used they are stoppered J with sterile rubber or paraffined corks, and over these is fastened a sterilized paper hood to protect the necks from contamination. When an intravenous injection is to be given, an ampoule is taken, its neck is nicked with a file, and this and the lateral bead are cleaned with alcohol, and broken off. The con- tents are then poured into 940 cc. of freshly distilled water, care being taken to mix the whole so that the specifically heavier salt solution may not simply settle to the bottom. After many trials I have found these the best ways to proceed, and as dispensing pharmacists in any community are willing to carry ampoules as here described in stock, one can easily obtain fresh and clean solutions at all times. If a precipitate of silicates should be found in an ampoule, one can readily avoid pouring this into the injection apparatus, or one may filter the contents of the ampoule through a little glass wool. A B FiGUHE 156. Technic of the Intravenous Injection of the Solution. My experience has convinced me that it is best in making these intravenous injections to cut through the skin and expose clearly to view the vein to be used. The results are bad if one fails in his attempt to enter a vein with a hypodermic needle through the skin. The salt-alkali mixture produces a great destruction of tion I find was scarcely necessary, for at the ordinary temperatures and pressures at which such sterilization is carried out, the carbonate is not decomposed. 560 (EDEMA AND NEPHRITIS the tissues if it is by accident injected into them. Under no circumstances must such salt-alkali mixtures ever he given sub- cutaneously, under the breast, or intramuscularly} A vein that is deemed sufficiently large is sought in the arm, in the leg, or, if necessary, in the neck. As we may wish to make several injections it is advisable to pick for the first injections the prominent veins most distant from the heart. Unfortunately in many of the conditions in which we wish to use the sodium chlorid-sodium carbonate solution, not much choice is allowed, for the blood vessels are so much contracted (toxemic shock?) that it is often impossible to find any usable vein below the bend of the elbow. Even here surgeons have been unable to find the median basilic in such cases. This will explain to the reader why in extreme cases such a vein as the jugular needs to be and has been used. To expose the vein painlessly a few drops of a cocain or novo- cain solution may be injected into the skin. If the patient is stuporous or in coma this is, of course, needless. The vein is freed from its surroundings and a ligature is tied about its distal end. A second ligature is thrown about the vein and after a cut has been made into the vein and the cannula connected with the injecting apparatus has been inserted, this second ligature is tied. Of course, care is taken to have no air enter the vein. The best type of cannula to use is shown in Fig. 157. The two openings at the tip and laterally make it well-nigh im- possible to shut off the infusion stream by crowding the cannula against the vein wall. The taper- FiGUBE 157. ing character of the cannula allows one to push it into even a small-cahbered vein, and the corrugation holds the cannula in place when the second ligature is tied. Sometimes it is better to use a large hypodermic needle in place of the cannula. The first ligature about the exposed vein then serves to steady the vein when the needle is pushed into it. When the hypodermic needle is used it is simply held in place ' Men who have failed to heed this oft-repeated caution of mine have been severest in attacking my teachings. Is it too much to ask critics at least to read what one has written? NEPHRITIS 561 until the injection is completed. As can easily be imagined, the use of the needle is especially convenient when one works with such a vein as the jugular. The disadvantages in its use arise from the fact that one is likely at any time to injure the blood vessel if the patient moves, and from the further fact that the carbonate solution affects the coat of the vein and so tends to leak out about the needle after the injection has been kept up for some time. The medical attendant should choose for intravenous injec- tion the apparatus with which he is most familiar. Perhaps that shown in Fig. 154, with a cannula replacing the catheter, and minus the thermometer and the insert F, is the simplest available form. The side tubulation with the rounded bottom, or a glass bulb insert filled with glass wool (Fig. 155) will make it almost impossible, if care is used, to inject any sediment that may accidentally appear in the injection fluid. The pressure bottle arrangement shown in Fig. 158 possesses some advantages over the apparatus just referred to. No special comment is necessary re- garding its use, and we need not in this day em- phasize the necessity of having all rubber tubes, etc., perfectly sterilized by boiling in distilled water. The faults of the ap- paratus shown in Fig. 158 are that it possesses no arrangement for keeping the solution at body tem- perature, and that we do not know in as accurate a manner as we desire, the exact rate and the exact pressure at which the sodium chlorid-sodium carbonate solution is entering our patient at any moment. To meet these difficulties, the useful apparatus shown in Fig. 159 was devised by Edmund M. Baehr. We have here again the glass pressure bottle shown in Fig. 158, but it is now surrounded by a copper water jacket by means of which the FiGUEE 158. 562 (EDEMA AND NEPHRITIS injected fluid may be kept at body temperature or a little above. The thermometer registers the temperature existing in the jacket, and as the temperature of the injection fluid falls on its way into the patient, it is advantageously kept a little above that at which we wish to deliver the solution into the patient. The rubber bulb in Fig. 158 is advantageously replaced in the appa- ratus shown in Fig. 159, by a metallic pimip. A mercury manometer, inserted as indicated in the drawing, allows one to know at all times the exact pressure obtaining in the pressure bottle. One needs at all times to inject the solution slowly into the circulation so that it may mix with the blood, and at as even a rate as possible. Not over 30 to 40 cc. should be in- jected per minute. By testing out the apparatus before making the injection one can easily note just how much pressure is necessary to accom- plish this. As the pressure in the larger veins is almost nil, 30 to 40 mm. of mercury pressure usually suffice, and one need never run above 50 mm. if a cannula or needle of proper diameter is chosen. If we give the solution to a patient who for any reason has to maintain an upright position, it is well to remember that in such a case the arm must be comfortably supported in as horizontal a posi- tion as possible in order not to have to work against a consider- able hydrostatic pressure in the veins. The pressure and the oscillations of the mercury column tell us every moment whether our solution is flowing in properly or not. Figure 159. The Quantity and Time Interval of the Salt-Alkali Injections It is necessary to say now how much of the solution may be injected at one time and how often the injection may be repeated. In any suppression case or in a case with convulsions, per- sistent vomiting or other alarming symptoms, 1800 to 2000 cc. NEPHRITIS 563 of the solution should be given for the first dose. In the case of a child we give a proportionate dose obtained by dividing the child's weight by that of a small adult. A 30-kilo (66-pound) child gets half the dose of a 60-kilo (132-pound) man, etc. The repetition of the injection and the amount given subsequently must then be determined by the condition of the patient. If, within two or three hours, urine begins to come and the convul- sions stop, or if the sensorium clears or the headache and eye symptoms improve, then we know that we have given enough for the time being. If the patient is awake, he is likely to complain of thirst during the hour that we are making the injection. It is best not to let him satisfy this immediately, for we wish to get as great a shrinking effect of the salt and alkali upon his various organs as possible. But there is no objection to his moistening his mouth, and at the end of four to eight hours, if his alarm- ing symptoms seem to be under control, we are only too glad to have him drink. Only we must always remember that as soon as water is given we decrease the patient's salt concentra- tion, and if his kidneys are functioning, we are actively wash- ing salt out of the body. So, to carry along our therapy, we give a natural or artificial alkaline water instead of plain water by mouth, and with this we may give various salts. By thus giving alkah and salts by mouth or by using alkali and salt by rectum, we may now be able to keep our patient growing pro- gressively better. But if this is not the case, or if the redevel- opment of some prominent sign or symptom informs us that our patient is relapsing into his previous state, then we may give a second injection of one or even two liters of solution, six, twelve, or twenty-four hours after the first injection. Closer rules than this can hardly be given. If our patient improves for a number of hours after the first injection, and then goes down again, we repeat the injection at this time; 500 to 1500 cc. in any twenty- four hours after the first injection is certainly safe. If the suppression of urine is not absolute, then it is a useful guide to the amount of alkali and salt that may be urged upon the patient. It is safe to give alkali and it should be given until the urine is persistently neutral to litmus. For self-evident reasons, it is possible for the urine to be alkaline immediately after an intravenous injection, even when the acid content of the 564 CEDEMA AND NEPHRITIS body generally is still abnormally high. Not until the urine is persistently neutral and is held there are we really succeeding in getting an adequate amount of alkali into the patient. Various observers have commented on the large quantity of' fluid that is injected intravenously in clinical cases of neph- ritis and allied conditions. I have injected as high as 6 liters (6 quarts) in twenty-four hours. Some clinicians have for various reasons remonstrated that this is dangerous, chiefly because they hold such injections to " increase the blood pressure." Observation is better than guessing in settling such points. Measurement made just before and just after a two-liter intra- venous injection shows either no change whatsoever in the blood pressure, or, if it has previously been high, a fall. Others hold such injections to " throw work on the heart." In so far as the elimination of water from the body costs energy, this is to a limited extent true, but only to a very limited extent, as I have previously insisted. Physiologists know that the volume of the circulating blood can be more than doubled without appreciable effect. Counting the blood in the human being as one-thirteenth the body-weight, it is therefore entirely safe to inject a liter of fluid for every 13 kg. (about 28| pounds) of body- weight; but as pointed out previously, this is a safe figure if blood, in other words, water in combination with a colloid, is injected. Only such remains in the blood vessels. When the water is injected " free," as in a salt solution, this rapidly leaves the blood vessels, and so the amount of this that may safely be injected lies still higher. What I have said here is still largely true even when we deal with sclerosed blood vessels, though to allow for the diminished elasticity, a slower injection or injection of a less amount at more frequent inter- vals, as the judgment of the operator may dictate, may advan- tageously replace the single large injections. About one in every three patients injected with alkaline hypertonic salt solution develops some reaction. At times he develops a chill which, however, does not last long, or he has a slight rise in temperature. Once I found sugar in the urine. It will be noticed that these findings are similar to those observed after intravenous injections of salvarsan and other medicaments. European authors have laid stress on the distilled water employed in making up the solution, maintaining that bacterial products NEPHRITIS 565 are present in old distilled water. For this reason I have always urged the use of freshly distilled water, but even then I have seen these reactions. I next attributed the effect to the action of the alkali on the red blood corpuscles, and a resulting hemolysis. This, however, can only be a small part of it, for one of the worst reactions I ever saw occurred in a woman to whom I gave a concentrated solution of neutral salt only. To explain the findings — which are now being studied on laboratory animals — I have come to the tentative conclusion that two things are active: first, a shrinking of the red blood corpuscles which makes these less able to carry oxygen; second, a more important direct action of the salt and alkali on the medulla. The latter effect I think, leads to the vasomotor disturbance in the skin which we call a chill, to the consequent retention of heat that accounts for the rise in temperature, and to the appearance of sugar in the urine, as C. Bock and F. A. Hoffmann '■ found many years ago when rabbits were perfused with sodium chlorid solutions. How to avoid these effects is not yet entirely clear. The rules I have formulated for my own guidance call for the use of freshly distilled water and as slow an injection as is conveniently possible. It is undoubtedly better to give in place of single large intravenous injections several smaller ones separated by intervals of time which allow the salt and alkali to diffuse into the body-tissues, but unfortunately the necessity of opening into more than one vein and the critical condition which we are usually asked to meet does not always allow of this. Before, throughout and after the injection the patient is carefully pro- tected from muscular exertion, and the possibility of a chill is guarded against by extra blankets, hot-water bags, etc. If complete muscular and mental relaxation is not easily obtained, a small dose of codein, heroin or morphin may be used. Perhaps the way to proceed is best illustrated by abstracting a few clinical experiences and commenting upon them directly. iC. Bock and F. A. Hoffmann: Arch. f. (Anat. u.) Physiol. (1871). Martin H. Fischer: Univ. of California Publications in Physiology, 1, 77, (1903); 1, 87 (1904); Pfluger's Archiv. 106, 80 (1904); 109, 1 (1905). 566 (EDEMA AND NEPHRITIS 7. Clinical Abstracts and Comment §1 The credit of having been the first to utilize upon patients the principles outlined in this volume belongs to James J. Hogan. Since then others of my friends and colleagues have used alkalies, salts, and water for the relief particularly of the acuter neph- ritides, and their accompanying manifestations, and with favorable results. My thanks are due them all for permitting me not only to see many of their patients with them but to use the facts contained in the brief illustrative histories that follow. Case VII. — Mr. G. B., a laborer, aged twenty-four years, and previ- ously in good general health, was operated upon under ether at 9 p.m., February 15, 1912, for a right inguinal hernia of several years' standing. The operation was a long one. Examination of the urine before opera- tion had been negative. Nothing abnormal was noted except that vomiting was rather severe after the operation, continuing until late in the afternoon. At this time, in response to inquiry the patient said he had no desire to urinate. The same condition existed late that night, even though by this time the patient was swallowing and retaining con- siderable quantities of water. The following morning the patient was catheterized, and 20 cc. of brownish, viscid liquid heavily charged with albumin were obtained. Through the day he was giv§n hot drinks and hot fomentations over the kidneys, but no spontaneous voiding occurred. At 9 p.m., that is to say, eighteen hours after the operation, he was again catheterized, and 15 cc. of urine of the previously described character were found. At this time liquid by mouth was stopped and administration of the following solution by slow drip into the rectum was started : Sodium carbonate (Na2C0r lOHaO) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. The patient retained the solution well, and by 10.30 had taken up the whole liter. At 11.45 he asked for a urinal, and passed 180 cc. of highly albuminous urine filled with casts and red blood cells. At 1 a.m. he passed another 160 cc, and at 3 a.m. 205 cc. The urine by this time was almost as clear as water. As he felt thirsty, a glass of water was now permitted him every hour. The urinary output continued so that by 9 p.m. of February 16, that is to say, in the first twenty-four hours after the drip was started, 2350 cc. were voided. By the evening of this day the albumin had dwindled to a trace, and only occasional granular casts could be found. NEPHRITIS 567 In the following two days, during which 2470 and 2385 ec. respectively of urine were obtained, these signs disappeared entirely. The urine of this patient, who is entirely well, has been examined repeatedly since, with negative results. Case VIII. — Mrs. M. L. T., aged forty-seven years, a laundress, was operated upon under ether for an extensively broken pelvic floor with prolapse of the uterus, at 8.30 a.m., July 1, 1912. She had not been in good general health previously, though she complained of nothing specifically excepted her uterine condition. Urinary examination on admittance to the hospital the day previously had shown a trace of albumin and occasional casts, but the general condition of the patient did not seem sufficiently bad to contraindicate an operation which was much needed. Her condition after the operation was fair, but she did not urinate. Catheterization twelve, twenty, and twenty-eight hours after her return from the operating room was dry. At 9 p.m., July 2, when catheteriziation was again found to be dry, she was slowly injected intravenously with 1600 cc. of the following mixture: Sodium carbonate (Na2C03 ■ IOH2O) 20 grams Sodium chlorid 28 grams Distilled water enough to make 2000 cc. She was catheterized at 3 a.m., when 90 cc. of a viscid, brownish urine, filled with albtunin and casts, were obtained. The urine was highly acid to methyl red.^ At 7 a.m. another 90 cc. of a similar looking urine were obtained. Because of the perineal operation, rectal injec- tion of alkali and salt was not urged, but half a glass of Vichy water, with a powder of half a gram each of sodium bicarbonate and magnesium oxid, was given every hour. On this regime 420 cc. in all of brownish, highly albuminous urine filled with casts were obtained in the first twenty-four hours after the intravenous injection. This alkali therapy was continued. The urine became neutral to litmus and remained so after July 5. In her second twenty-four hour period she secreted 670 cc. of urine, and in the third, 1140. The urine became clearer, the casts fewer, and the albumin content diminished steadily (from 9 grams to the liter in the first specimens to 1.5 grams on July 5 and 6). Her further history is summarized below : For 24-hour period of Amount of urine. Grams of albumin per liter (Esbach). July 7. 1790 2100 1873 1840 1.0 July 8 0.8 July 9 0.9 July 10 0.5 1 The advantages of using methyl red and paranitrophenol as indicators instead of the ordinary litmus are explained on page 634. 568 OEDEMA AND NEPHRITIS Numerous hyaline and granular casts were found in all these specimens. These continued with a urinary output of 1700 to 2200 cc. per twenty- four hours, containing albumin that varied little from half a gram per liter until she was discharged from the hospital July 21. A more detailed physical examination after her operation than had previously been possible, showed this patient to have easily palpable peripheral blood vessels with an evidently enlarged but regularly beating heart. The systolic blood pressure was 165 mm. of mercury, the diastolic 140. Tender bones with bogginess of the tibial periosteum, a tender nasal bridge, vague night pains, and three miscarriages for which no cause was assigned, together with the casts and albumin found on admis- sion to the hospital, led to the diagnosis of syphilitic vascular disease with cardiac hypertrophy and involvement of the kidney (chronic interstitial nephritis) . Since leaving the hospital this patient has been seen occasionally. Some albumin and casts are constantly found in the urine, and her blood pressure continues at approximately 160 mm. The patient herself complains of nothing and feels herself improved by her operation. In the light of our considerations it is not surprising that an anesthesia nephritis as illustrated in these two cases is easily relieved by alkali and salt. During an anesthesia we introduce into the body a poison which interferes with the normal oxida- tion chemistry of the cells and that we have an abnormal pro- duction and accumulation of acid following this, is attested not only by the thirst of which the patient complains, but by the accelerated breathing and heart beat, the abnormally high acid (hydrogen ion) content of post-anesthetic urine, and the appear- ance in it of such " acidosis " products as acetone, diacetic acid, lactic acid, etc. But as soon as we stop administering the anesthetic the patient begins to exhale it, and so in a compara- tively short time the intoxication responsible for the abnormal production and accumulation of acid disappears. The patient also usually succeeds in oxidizing the acid products resulting from his intoxication, and so it is the usual thing to see him bear his anesthesia without bad after-effects. But some- times he is not so successful and then if his intoxication evidences itself chiefly from the side of his kidneys, we say he has a post- operative nephritis. (If it should happen to involve his liver particularly, we say he has a " postoperative jaundice," a " chloro- form liver," etc.). There is nothing surprising about the fact that an administration of alkali and salt relieves this condition. The alkali neutralizes the abnormal acids present, and the NEPHRITIS 569 increased salt concentration reduces the hydration capacity of his swollen kidney (and other body) colloids. If the injury to them has not been too great (or technically put, if the colloid changes characteristic of nephritis have not become "irreversible"), these measures quickly restore his kidneys (and other involved organs) to a more normal state. As the kidney shrinks, a better blood supply to the organ is obtained, and so the kidney cells are once more enabled to resume their work. Once this result has been obtained a recurrence need scarcely be feared, for the original intoxicant (the anesthetic) has by this time disappeared from the body, and so the relief obtained is permanent. What has just been said applies in my judgment to Case VII. In Case VIII the problem is essentially the same, only the anesthesia intoxication is this time added to the effects of a blood- vessel disease which in itself has already led to the signs of a nephritis. The effects of the anesthesia intoxication could, as in Case VII, be overcome, but an injection of salt and alkali does not remove an endomesarteritis with its effects upon parts or all of the kidney, and so casts, albumin, etc., continue to be found in the urine even after the suppression following the anesthesia has been reUeved. §2 If we will write the name of any other intoxicant in the place of anesthetic in these considerations, we have what happens, to my mind, in the nephritides that we encounter in any of the acute infections, and in the various other acute intoxications which we know to be associated with the development of neph- ritis. Here again nature herself takes care of the great majority of cases, but when she does not, we may again be able to relieve the urinary condition by alkali, salts, and water. Such a situa- tion is illustrated in Case IX, where an unknown intoxicant led to suppression of urine and in cases X and XI, where the suppression followed scarlet fever. Case IX. — (Db. Elizabeth Campbell, Cincinnati, Ohio.) A. H., a girl, aged three and one-haK years, had never been ill previously. When first seen she was extremely nauseated and vomiting. There was a slight general csdema, and her urinary output was low. The urine contained much albumin and casts. Enemas of 0.85 per cent sodium chlorid solution and calomel by mouth improved the child's condition, and brought up the urinary output. Five days later the 570 (EDEMA AND NEPHRITIS mother reported that since two in the afternoon of the day before the child had passed no urine. The child was given hot baths and sweated several times. Several 0.3 gram doses of sodium bicarbonate were given by mouth, and water was urged. The child had vomited once. She was extremely quiet. Temperature and respiration were normal. The pulse was 76. 0.2 gram calomel was given at night, and through the night hot alkaUne drinks, blanket sweats, and enemas of 0.85 per cent sodium chlorid solution were continued. The anuria persisted. This scheme of treatment was kept up through the following day also. The general oedema had in the meantime increased, the patient was vomiting, and had grown stuporous, and the pupils reacted slowly. In consultation in the early evening of this the third day of the anuria it was decided to try the administration of an alkaline hypertonic sodium chlorid solution. As it was thought there might be some urine in the bladder, the child was catheterized. A teaspoonful of bloody urine was obtained. At 9 p.m. 320 cc. of the following solution were injected into the rectum by slow drip, the child retaining aU of it. Sodium carbonate (Na2C03- IOH2O) 15 grams Sodium chlorid 14 grams Water, enough to make 1000 ec. The child had a restless night. At 5 a.m. she passed 96 cc. of urine that looked like pure blood. This was sixty-three hours after the suppression was first noted. At 6.15 a.m. another 130 cc. of bloody urine were passed; at 7.30, 256 cc, with only a slight amount of blood. At 10 A.M. a large voiding was lost with a watery stool. An hour later 250 cc. of the above alkaline hypertonic sodium chlorid solution were slowly injected into the rectum. Through the afternoon the nurse reported that " the urine came in an almost steady stream." The albumin content of the urine fell rapidly, so that by the following day only a trace could be found, and on the fourth day later it disappeared entirely. The general oedema disappeared rapidly, and was about gone on the third day after the urinary secretion became reestablished. Recovery has been complete. Case X. — (Db. James J. Hogan, San Francisco, California. Mrs. W., twenty-two years old, had passed through a, scarlet fever. Dr. Hogan was called in consultation on the evening of March 11, 1911, and found the patient unconscious with practically a complete suppres- sion of urine that had lasted for twenty-four hours. The unconscious- ness had lasted for twelve. The following formula was given by the continuous drip method into the rectum : Sodium carbonate (NazCOs • IOH2O) 20 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. The urinary flow recommenced after four hours; on the following day her mind had cleared, and the patient made a subsequent uneventful recovery. NEPHRITTS 571 Case XI. — (Dr. H. Kennon Dunham, Cincinnati, Ohio.) Master M., seven years old, was seen on April 30, 1911, by Dr. Wm. C. Schmidter in a rather mild attack of scarlet fever. The temperature at no time ran above 101° F. In spite of the apparent mildness of the attack, the child developed urinary symptoms. On May 7, when Dr. Dunham was first summoned in consultation, a complete suppression of urine had lasted for fifty-one hours, the child was conscious, but very stupid, presenting a grave picture of intoxication. The eyelids and ankles were swollen, the pulse 105, respiration 24. At 4.00 a.m. the following mixture was prepared and its injection into the rectum begun: Sodium carbonate (NasOs ■ IOH2O) 20 grams Sodium ohiorid 30 grams Distilled water enough to make 1000 cc. The injection required one and a half hours. About 180 cc. were rejected, the remainder of the above solution was retained. Three and a half hours after the injection was completed the patient passed invol- untarily a large watery stool. Ten hours after the completion of the injection he passed a small amount of highly colored urine. Following this at short intervals came large voidings of urine which were lost into the bed, as the child could not control himself sufficiently to use a bed- pan. Not until this secretion had lasted for four hours could the urine be collected. Not counting that which was lost there were collected 2272 cc. of urine in the first twenty-four hours after urinary secretion commenced. The first specimens of urine obtained were so filled with albumin as to set into a solid mass on boiling. The amount of albumin rapidly decreased, so that during the second day after the injection only a moderate reaction for albumin was obtained, and on the ninth day it disappeared entirely. The intense stupor left the child within the first twenty-four hours after injection, and on the third day he was actively interested in his surroundings and free from CEdema. His urinary secretion after being started was readily maintained by the milk diet on which he had been from the first and to which alkaline mineral water was added ad libitum. §3 For self-evident reasons our prognosis grows worse and our efforts need to be greater as the intoxication grows in length or becomes of a less removable type. We encounter this situation in protracted infections, in acute infections that leave behind toxins that stick particularly firmly to the kidney cells (scarlet fever?) and in poisonings of this type (phosphorus and the metals). Not alone do some of these produce irreversible col- loid changes (necrosis) in the cells of the kidney from the start — changes, therefore, which can never be " cured " by any thera- 572 (EDEMA AND NEPHRITIS peutic procedure — but even when such is not the case, in these lasting intoxications the interference with the normal oxidation chemistry of the kidney cells is of a more lasting character, and so our therapy must also be more persistent. A single dose of alkali and salt may then do no more than give temporary relief. We must meet the intoxication as long as it persists. Cases XII, XIII, and XIV may serve in illustration of these remarks. Case XII. — (Drs. Otto P. Geieb and J. L. Tuechtbe, Cincinnati.) G. L., a 34-year-old attorney, developed a severe tonsillitis involving both tonsils on May 17, 1911. His temperature was 103.5° F., pulse 120. The urine was very scanty, highly colored, and contained albumin and casts. The next day the patient had intense headache, and in the evening became delirious. During these second twenty-four hours of his illness he passed but 90 cc. of urine, very smoky in color and fiUed with albumin, red and white corpuscles and casts of all sorts. On the third day of his illness he passed no urine at all. His delirium con- tinued and his temperature remained at 103° F., his pulse at 124. I^ate at night he was given the following mixture per rectum: Sodium carbonate (NazCOs- IOH2O) 20 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. In his delirium most of the first injection was rejected. At 3.00 A.M., May 20, the injection was therefore repeated. About 500 cc. were retained. At 6.00 a.m. 150 cc. of dark, thick urine were obtained which on heating fairly set into a jelly. The urinary secretion became more profuse as the day wore on, and in the first twenty-four hours after the successful injection 1184 cc. of urine were obtained. As the urinary secretion increased, the drowsy deUrium passed away, the headache disappeared, and the patient volunteered that he felt well. The temperature fell to 101° F., the pulse rate to 100. The later speci- mens of urine voided in these twenty-four hours after the successful injection were clear and amber in color and contained only a little albumin, and few casts and blood cells. The rectal injections of 500 cc. of the above formula were repeated May 21 (temperature 99.5° F., pulse 90) and May 22 (temperature normal, pulse 70). Between the injections the patient was urged to take as much Vichy water by mouth as he could. The urine secreted May 21 measured 1376 cc, that secreted May 22, 1408 cc. Some albumin and casts were found in the former, only a trace together with some red blood corpuscles but no casts in the latter. On May 23 all urinary signs had disappeared, and the patient made an uneventful recovery. Case XIII.— (Dr. William E. Kiely, Cincinnati.) Three weeks before entering the hospital S. C. W., thirty-eight years old, and a mod- NEPHRITIS 573 erate beer drinker, became short of breath, suffered from headaches, and noticed a swelling of his legs and abdomen. Physical examination showed no disease of the heart or lungs, but fluid in the pleural and peritoneal cavities, with a general cedema of the subcutaneous tissues. The urine was low in amount, of high specific gravity, and contained much albumin, some blood cells, and hyaline casts. On this a diagnosis of (chronic) parenchymatous nephritis was made. After twenty-five days of rest in bed, a milk diet, a daily hot bath, saline cathartics and digitalis, no improvement in his general condition was noted. There was now added to his diet a liter of water daily containing 25 grams of sodium chlorid. Improvement in his general signs and symptoms began immediately, the urinary output rose, the blood disappeared, and the casts and albumin progressively diminished in amount. After ten days of this treatment he was much better, and at the end of twenty- five all signs of his oedema and the effusions into his serous cavities had disappeared. At his own request he got out of bed and began to work about the ward, and shortly thereafter left the hospital free of all signs and symptoms, except a faint trace of albumin in his urine. In this state he has continued up to the present time (that is, for two months since leaving the hospital) . Case XIV. — (Dr. Julius H. Eichbbrg, Cincinnati.) A. B., a 40-year-old lawyer, entered the hospital in April, 1911, with a history of kidney disease of eight years' standing. At various times during these years he had had a diagnosis of chronic parenchymatous nephritis made upon him. He had no enlargement of the heart and no increased blood pressure. The original cause of the nephritis could not be made out. When first seen the patient was passing about 400 cc. of urine per twenty-four hours, containing 4 grams of albumin per liter and filled with all varieties of casts. On a milk and vegetable diet, sweat baths, and sahne cathartics his urinary secretion increased somewhat, but his general condition did not improve, the number of grams of albumin lost each twenty-four hours did not decrease, and his csdema, ascites etc., increased. After two weeks in the hospital he had a well-marked cedema of his legs, back, chest-wall, scalp, and face. The fluid in his abdomen extended to the umbilicus when sitting up. While his general hospital regime and diet were kept as before, he now had added to his drinking water and consumed each twenty-four hours 7 grams of dried sodium carbonate. After ten days of the carbonate administration his cedema and ascites disappeared completely, his urine increased to approximately 800 cc. per twenty-four hours, though the quantity of albumin lost per twenty-four hours did not change perceptibly. The patient at this point refused to continue taking the carbonate. In five days his weight went up 2^ kUos. The patient was persuaded to resume the carbonate, and at the end of another seven days his original weight had again been attained, and the visible signs of oedema which had developed when the carbonate was discontinued had once more disappeared. The urinary output amounted at this time to 800 cc. daily, and the albumin dropped to 2.5 grams per Uter. 574 OEDEMA AND NEPHRITIS At this point the patient refused a second time to take the sodium carbonate, and again the swelling of his legs and back developed, while his weight rose as before, 2^ kilos in less than a week. Following this period he returned a third time to the carbonate, and in six days had again lost his 2| kilos and the obvious signs of an oedema. This is his state at the present writing when for four months he has been passing 1280 cc. or more of urine daily, containing some casts and 0.75 gram of albumin per liter. He has left the hospital in fair condition, has a good appetite, sleeps well, and has resumed the practice of his profession. §4 The nephritides occurring in the course of pregnancy ^ constitute a common and important group. How alkali, salt and water help in their management is illustrated in Cases XV, XVI, XVII, XVIII, XIX, XX, XXI, XXII, XXIII, and XXIV. '■ The origin and nature of the poison which gives rise to the symptoms of the pregnancy intoxications form interesting food for speculation. The beginning of the intoxication with pregnancy and its prompt cessation with birth of the child, together with the fact that an organism is immune to its own proteins, makes me believe that the foreign protein of the male brought in with the spermatozoan marks the starting point of the intoxi- cation. In this sense the morning sickness, the nausea, etc., occurring early in the pregnancy mark the beginning of the intoxication, but as immunity is usually established they are likely to pass away. When immunity is not established the severer signs of the later months of preg- nancy supervene. A woman who has once been pregnant is less likely to be a second time the victim of an intoxication, because the immunity developed in a first pregnancy protects her against the intoxication conse- quences of a second. Moreover, a woman married more than once may show intoxication with one and not with another man (because the foreign pro- tein is different in the two). Termination of the pregnancy (removal of the foreign protein as contained in the developing embryo) cuts short the intoxication. Many circumstances, moreover, serve to aggravate the nephritis encountered in pregnancy when it has once become established. Such are the " acidosis," for example, secondary to the starvation conse- quent upon the vomiting, the nausea, and the absurd dietary restrictions to which these patients are so often subjected, a state betrayed only too clearly by the high ammonia excretion and the presence of acetone, diacetic acid, etc., in the urine. I should further like to emphasize that not every nephritis observed in a pregnant woman is at once to be attributed to the pregnancy, as is so gen- erally done. Vascular disease and infections of the kidney are very com- monly overlooked. It has also been my experience that many patients are said to have a pregnancy intoxication resulting in the death of the fetus when as a matter of fact syphilis or some other condition first killed the fetus and the products of its decomposition then served to affect the kidney. NEPHRITIS 575 Case XV. — (Dr. James J. Hogan, San Francisco, California.) Mrs. R., pregnant and practically at term, entered the hospital March 7, 1911, at 5.30 P.M., complaining of continuous uterine pain. She had a general oedema. Signs and symptoms indicating that a nephritis had existed for at least some days past were evident, but no proper examina- tion of the urine had been made. The os on examination was found rigid. Because of the intense pain 0.015 gram morphin was given hypodermically at 9.00 p.m. She went to sleep but awoke at 11.00 in a severe convulsion. The patient was catheterized and 60 cc. of bloody urine of a syrupy consistency were obtained. On testing this for albumin it fairly set. Casts, cellular detritus, red blood cells, etc., were found microscopically. 600 cc. of an 0.85 per cent sodium chlorid solution were given by rectum and immediate emptying of the uterus was deemed necessary. This was done under ether anethesia and as the os was very rigid required a half hour. A second convulsion occurred on the operating table. Immediately after the operation another 500 cc. of an 0.85 per cent sodium chlorid solution were given by rectum. Between this time (11.30 p.m., March 7) and 4.50 p.m., March 11, in other words, for practically four days, no urine could be obtained by catheter. During this time no convulsions occurred and the patient's mind remained clear. A continuous salt drip was used in the rectum and water and magnesium sulphate were given by mouth, but no evidence of a return of urinary function was obtainable. It was now decided to use a more concentrated sodium chlorid solution and alkali. The following mixture was therefore prepared : Sodium carbonate (NaiCOs- IOH2O) 20 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. This was injected into the rectum at body temperature by a con- tinuous drip method. In an hour and ten minutes 30 cc. of bloody urine were obtained, and an hour later 80 cc. more. From now on the urine fairly streamed out. The secretion continued and the albumin and casts entirely disappeared from the urine by the fourth day. The patient made an uninterrupted recovery. Case XVI. — (Dr. Lemuel P. Adams, Oakland, California.) Mrs. E., twenty-six years old and a primipara, began to feel below par, became pale, and developed a generalized oedema when pregnant seven and a half months. The secretion of urine was low, and this contained much albumin and various casts. Her condition gradually grew worse, so that it was deemed wise to put her to bed in the hospital. For ten days here, on a milk diet, and cared for in the approved ways, she showed no improvement, passing between 240 and 360 cc. of urine per twenty-four hours, filled with albumin, casts, and red and white blood corpuscles. As she now began to develop twitchings, was extremely oedematous and nearly blind, and as the onset of convulsions was feared, premature labor (at 8 months) was induced through gradual dilatation of the uterine os by means of water bags. Complete suppression of 576 ■ (EDEMA AND NEPHRITIS urine followed delivery. After this had lasted for thirty-one hours and no urine had come consequent upon hot packs, cupping, digitalis, etc., a slow injection of the following mixture into the rectum was begun: Sodium carbonate (NazCOs • IOH2O) 20 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. Urine began to come four hours after the injection was commenced and amounted to 1536 cc. in the first twenty-four hours. Two injec- tions daily of 500 cc. each of the above formula were continued for three days, together with water, milk, and cereals by mouth. On the second day 2176 cc. of urine were obtained, on the third 2140, on the fourth 2180, and on the fifth 1856. On the fifth day casts and blood cells had entirely disappeared from the urine and only the faintest trace of albumin remained. The oedema had diminished greatly, eyesight was returning, and the patient was actively interested in her surround- ings. On the following day the last of the albumin was gone and the patient went on to an uneventful recovery. Case XVII. — (Dr. Dudley Smith, Oakland, California.) Mrs. W., aged thirty, and seven months pregnant, presented herself for examination in May, 1911, with a history of nephritis and threatened eclampsia in her first pregnancy, ten years before. The second preg- nancy, three years before, had been uneventful. Urinary examination when the patient first presented herself was negative. On June 7, she began to show albumin in her urine and marked signs of general intoxication. (Edema of the face and feet developed. She was put to bed and placed on a milk diet, and saline cathartics were administered. Under this treatment she got no better. About the first of July, active administration of alkalies was begun in the form of 1 to 1 5 grams of sodium carbonate dissolved in a glass of plain water, or Vichy water, every two hours. Marked and positive improvement occurred in all her general symptoms and the oedema disappeared entirely. She was permitted to get out of bed again, but the alkali therapy was continued. On this regime she was carried to full term with no further general symp- toms of consequence. Her urinary output lay between 1800 and 2800 cc. daily and some albumin and casts continued in the urine. On July 24 she complained of severe continuous uterine pain, and with this came a marked reduction in the urinary output, extreme nervous- ness, and severe headache with nausea and vomiting. On the morn- ing of July 25 the urinary secretion had stopped entirely. She was sent to the hospital at noon and the following formula was slowly injected into the rectum: Sodium carbonate (NajCOs-lOHjO) 15 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. At 3 P.M. the uterine pain, the headache, and the nausea had disappeared and the patient went to sleep. At 4 p.m. the rectal NEPHRITIS 577 infusion was given a second time and almost a liter was absorbed. At 11 P.M., 258 CO. of urine were voided and the patient passed a good night, sleeping soundly. The following morning 500 cc. of urine, very high in albumin, casts, and blood, were passed. At three o'clock of this day, she again developed severe headache, nausea, and vomiting, and was unable to retain the rectal infusions, or anything by mouth. At 10 p. M. all the symptoms had so increased in severity, that 300 cc. of the above solution were given intravenously. In fifteen minutes the patient volunteered the information that her headache and nausea were gone. She was comfortable until the next afternoon, when periodic uterine pains developed, and the headache and vomiting returned. The patient was taken to the operating room, and the cervix was dilated slowly by hand. Delivery of the living child was accomplished in an hour and a half. This was followed by another intravenous injection of 645 cc. of a solution containing 7^ grams sodium carbonate (Na2C03- IOH2O) and 14 grams sodium chlorid to the liter. In the following twenty-four hours 2200 cc. of urine were voided, and as the nausea, vomiting, etc., had disappeared it was an easy matter to main- tain such a urinary output by giving water and alkalies by mouth. Albumin and casts disappeared from the urine on the fourth day and the patient had an uneventful convalescence. Case XVIIL— (Dr. W. A. Claek, Oakland, California.) Mrs. C. H., aged thirty-five, and pregnant for the second time, presented herself for examination in March, 1911. She had menstruated slightly, and for the last time, January 22. A year previously she had given birth to a healthy child at term, though in the later months of her preg- nancy her limbs and face had swelled, she had much headache, and her eyes had troubled her. At the time of her first visit, and repeatedly afterward, physical examination and examination of the urine showed nothing abnormal. On August 11, she showed a well-marked gen- eralized oedema, and complained of headache, extreme restlessness, sleeplessness, dimness of vision, and constant nausea. Her urinary secretion had fallen to 500 cc. per twenty-foyr hours, was highly acid, and high in albumin and casts. She was immediately sent to the hospital and kept in bed on a diet rich in water, alkalies, vegetables, and milk. Epsom salts were administered by mouth, and 0.85 per cent sodium chlorid solution was repeatedly injected slowly into the rectum. On this regime all of her symptoms and signs, including the albumin and casts, disappeared, and the urinary output rose, so that 2200 to 2674 cc. were voided every twenty-four hours. August 26 the patient felt so well that she insisted on getting out of bed and busying herself about her room. On the second day following this renewed activity, her headaches again showed themselves, and her nervousness and sleepless- ness returned. On August 29 her nausea and vomiting became severe, and her vision very dim. The oedema of the legs and face returned, and her urinary output fell slightly, to 1984 cc. When the heat test was applied to the urine, the whole became solid. This condition continued 578 CEDEMA AND NEPHRITIS until 11.30 P.M. of August 30, when the headache, nausea, vomiting, etc., were so severe that it was decided to give alkali and salt intraven- ously. The following formula was given: Sodium carbonate (NajCOs ■ lOHjO) 10 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. In an hour the patient volunteered the information that her head- ache and nausea were better, and that she felt brighter. She slept well, and passed the next morning comfortably. Examination of the urine passed in the night and early morning showed a decided drop in the amount of albumin excreted. Even though the subjective symptoms of the patient continued well, the albumin content of the urine again rose so that on the morning of September 1 this was sufficient to make the contents of the test-tube again set in a solid mass when boiled. The amount of urine obtained continued good, being 1984 and 2048 cc. respectively, for the last two twenty-four-hour periods. It was deemed best to empty the uterus, and at 10 a.m. of September 1, dilatation of the uterine os by means of rubber bags was begun. Rhythmic pains began two hours later and as these increased in number and severity, the patient's headache and nausea increased, and the urinary secretion fell. At 4 P.M. the patient vomited and developed a twitching of the face and arms. This continued at intervals until 11 p.m. when two liters of the alkali-salt mixture of the composition previously used in this case, were injected intravenously. Shortly after this, the sub- jective symptoms of the patient became better and she fell asleep, passing a fairly good night, and examination of the urine again showed a decided drop in the amount of albumin present. The general condi- tion of the patient continued good, and on the evening of September 2, she was delivered under chloroform anesthesia of a 1760-gram, living, female child (left shoulder presentation with version). On the operat- ing table the patient received 1000 cc. of 0.85 per cent sodium chlorid solution under the skin, and for subsequent treatment the patient was given this same salt solution by rectum. Alkaline water (a gram of sodium carbonate in a glass of water every hour) was given by mouth. The urinary secretion on this regime never fell below 2200 cc. On September 4 the albumin in the urine had dwindled to a trace, and on the next day it disappeared entirely. Examination of the urine twice daily from this time on invariably showed an alkaline reaction to litmus paper and no albumin. The general oedema disappeared on the third day after delivery. On September 17 the patient was fully convalescent. Case XIX.— (Dr. N. A. Hamilton, Franklin, Ohio.) Mrs. C, twenty-seven years old, and a primipara in the seventh month, showed nothing abnormal on examination, September 8. On September 20 some albumin was found in the urine, and on September 27 it was present in abundance. Her general condition was good. At 10 P.M., October 2, she. was seized with sudden nausea and vomiting which continued through the night. At 3.30 a.m., October NEPHRITIS 579 3, she had short lapses of consciousness. Headache was severe; there was some oedema of the face and legs; the pulse was 100 and hard. Veratrum was given by hypodermic injection. At 8.30 a.m. her pulse had fallen to 52; her temperature was nornal. No urine had been passed through the night, but at this time she passed 30 cc. The patient was dizzy, still vomiting, had pain in her neck, and her sight was blurred. She was now given 800 cc. of a strong (hypertonic) sodium chlorid solu- tion (1.5 per cent) by rectum. This was all retained. At 11.30 a.m. 90 cc. of dark-colored urine filled with casts and containing so much albumin that on boiling it fairly set was passed. Another 800 cc. of the sodium chlorid solution were now given and at 2 p.m. an unknown amount of urine was lost with a stool. Twenty minutes later a con- vulsion lasting a minute occurred, and this was repeated a half hour later. The patient was vomiting, and could not distinguish colors. There was a general twitching of the muscles. A general anesthetic was given at 3.30 and an attempt made to dilate the very rigid uterine 05 instrumentally. At 5.00 p.m. the membranes ruptured, and at the same time 30 cc. of dark-brown urine were obtained by catheter. At 6 P.M. the temperature of the patient was 100.2° F. by axilla. Another injection of 800 cc. of the strong saline solution was given by rectum at this time and repeated at 8 p.m., but neither was retained well. At 10 P.M. a little urine (estimated as 30 cc.) was passed with a stool. At midnight the patient's temperature was 100.2° F., she was dizzy, could not distinguish between men and women, and was unable to differentiate white from black. At this time she was given the following formula intravenously : Sodium carbonate (crystallized) 20 grams Sodium chlorid 28 grams Water, enough to make 2000 cc. The injection required an hour. While giving the injection the patient volunteered the information that her nausea had left her,andthat her headache was disappearing. At 2.30 a.m., October 4, she passed 75 cc. of dark-brown urine filled with casts and fairly solid with albumin on boiUng. At 4 a.m. she passed another 75 cc. and at 6.45 a.m. 95 cc. During these hours she slept at intervals. When she awakened her headache and nausea were gone, and she could distinguish between gross objects, and recognize colors. From now on and through the day she was plied with water by mouth and five injections of 400 cc. each of the above sodium carbonate-sodium chlorid mixture were given by rectum. These were well retained. Urine was voided about every three hours, and in increasing quantity. By midnight, that is to say in the first twenty-four hours after the intravenous injection, she had voided 572 cc. not counting two " large " voidings that were lost. The later portions of this urine were clearer in color and contained much less albumin than the specimens already described. In the night of October 5, the patient went into labor, and at 8 a.m. forceps were introduced and she was delivered of a macerated fetus. 580 (EDEMA AND NEPHRITIS In spite of the exertions of labor she passed 320 cc. of urine between mid- night and the time of the delivery of the placenta. Through the night the alkali-salt enemas could not be retained, but through the day she took and retained four enemas of 400 cc. each. In this second period of twenty-four hours she passed 734 cc. of urine. After delivery, her temperature, which on the night before had risen to 103.6° F. (by mouth), fell to normal. In the twenty-four hours of October 6, she received and retained four enemas of 500 cc. each of the alkali-salt mixture, and drank freely of water (a glass every hour). She passed in this period 1840 cc. of urine, not counting two voidings that were lost with the stools. The later portions of this urine contained only a little albumin. The patient was sleeping well, and rehshing her toast, gruel, eggs, milk, and broth. In the next two days the alkali-salt enemas were reduced to two daily, one night and morning, and then stopped entirely. She was given a Uberal diet, and water was insistently given by mouth. Lemonade and orangeade were urged. When the alkali was no longer given by rectum, sodium carbonate (0.5 gram) was given in a glass of water as often as the patient would take it both day and night, and she was asked to salt her food Uberally. Her urinary output on this regime was as follows : October 7 3616 cc. Oc-tober 14. . .2400-1- cc. October 8 3264 cc. October 15. . . .4096-t- cc. October 9 3520 cc. October 16. ..3808 cc. October 10 2528 cc. October 17. ...3200 cc. October 11 2108 cc. October 18 . ... 1920 cc. October 12 2396-1- cc. October 19 1915 cc. October 13 2432 -f- cc. The great rise in urinary output on October 15 followed an increase in the amount of alkali and salt given by mouth; the fall on October 18 a reduction of this. The cedema had disappeared and the albumin dwindled to a trace by October 7. This trace persisted up to October 19. The patient developed a slight temperature (100.8° F.) on the fourth day after delivery, but following intrauterine douches with bichlorid of mercury and iodin this fell so that only a temperature of 99° or 99.2° was registered in the afternoons up to October 17. From October 16 she was given an unrestricted diet, and on October 17 she sat up for the first time. On October 24 she " is downstairs, voiding an abundance of urine and happy." Recovery was complete. Case XX.— (Dr. E. A. Majors, Oakland, California.). Mrs. A. B., pregnant for the second time and at term was found in labor and delivered, of a healthy living child, in an entirely normal way at 1 a.m. No previous history was obtainable. Following labor she fell into a deep sleep and at 5 a.m. it was impossible to arouse her. As there was no evidence of urinary secretion, she was catheterized at 6 a.m. NEPHRITIS 581 No urine was obtained. At 7 a.m. she had two severe convulsions. Following this she lay in a deep stupor with rapid breathing. At 10 she was again catheterized, but no urine was obtained. She now received by slow injection into the rectum the following: Sodium carbonate (NaBCOs ■ IOH2O) 15 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. Sixty cc. of urine were obtained an hour after the beginning of the 'injection, and half an hour later another 130 cc. filled with albumin and casts. At the same time the patient began to clear mentally. Three hours after beginning the injection she would respond to ques- tions. From this time on she was plied with water by mouth. Later in the afternoon 500 cc. of the above formula were again given by rectum and this was repeated next day. In the first twenty-four hours 1525 .cc. of urine were obtained, and in the second 2240 cc. At the same time the albumin and casts diminished and on the third day the urine cleared entirely. Uneventful convalescence followed. Case XXI.— (Dr. C. C. Fihe, Cincinnati, Ohio.) Mrs. E. J. H., aged twenty-six years, showed much albumin and many casts in her urine, and developed a generalized oedema in the seventh month of this, her first pregnancy. Her urinary output was scanty and highly colored, and she felt herself below par. Her physical activities were much restricted, and she was placed on an active alkali therapy. Vegetables and sweet fruits were urged upon her, sodium phosphate and citrate were frequently given, and at regular intervals sodium carbonate and sodium chlorid were administered in capsules followed by water. Her symptoms cleared markedly, and she was carried to term, voiding 700 to 1500 cc. of urine daily. At midnight, February 1, she went into labor, and at 8.30 a.m. was delivered of a living child. She had passed no urine the day previously, and none was passed during these hours. A gradually increasing blurred vision in her right eye, of which the patient had complained for several days past, had increased. She complained of headache. At 11 a.m. a convulsion lasting ten minutes occurred, and at 2.30 p.m. another last- ing twenty minutes. At both times chloroform was administered. She was catheterized and 30 cc. of dark brown urine filled with albumin were obtained. At 6 p. m. catheterization was dry. The pulse ranged be- tween 148 and 92. At this time 150 cc. of the following solution were given intravenously : Sodium carbonate (NaaCOs • IOH2O) 10 grams Sodium chlorid 14 grams Water, enough to make 1000 cc. A severe convulsion occurred while the intravenous injection was being made, and so this had to be discontinued. 0.015 gram morphin 582 CEDEMA AND NEPHRITIS was given hypodermically, and another 250 cc. of the solution were injected into the rectum. At 8.15 p.m. 1200 cc. of urine were obtained by catheter from the bladder. At 8.45, 1200 cc. of the above solu- tion were given intravenously. An unmeasured amount of urine was passed with a stool at 10.30. At 11, 1022 cc; at 12.30, 96 cc; at 2.40, 16 cc; at 6.40, 512 cc. were passed voluntarily. The pulse gradually feU. during these hours to 76. Sodium phosphate, sodium carbonate, and sodium chlorid were given in small doses at regular intervals by mouth, and after the first signs of a freer urinary output, water and milk were urged at hourly intervals. In ,the first twenty-four hours after the intravenous use of the alkaline hypertonic salt solution, 4814 cc. of urine were passed, not counting two voidings that were lost. The patient's uneventful subsequent history is summarized below. Five grams each, per twenty-four hours, of sodium chlorid and sodium carbonate were given in divided doses in capsules followed by water, at regular intervals day and night. In addition, 15 grams of disodium phosphate were given once or twice daily, and for the patient's anemia, 0.5 cc. tincture of iron chlorid three times daily was prescribed. Albumin in Date. Temperature. Pulse. Urine in 24 hours. grams per liter (Esbach). Remarks. Feb. 2-3 98.0° to 98.6° 78 to 80 4128 + 1.0 __Milk diet. Feb. 3-4 98.2° to 98.6° 68 to 78 5578 0.5 Milk diet. Feb. 4-5 98.2° to 98.6° 64 to 74 4127 0.5 + Light mixed diet. Feb. 6-6 98.2° to 99.2° 72 to 86 5382 0.5 Light mixed diet. Feb. 6-7 99.0° to 100.2° 74 to 86 4288 0.5- Light mixed diet. Feb. 7-8 98.6° to 99.6° 74 to 80 3904 0.25 Light mixed diet. Feb. 8-9 98.0° to 98.6° 68 to 75 5272 Trace Light mixed diet. Feb. 9-10 98.6° to 99.0° 68 to 72 4032 Trace Light mixed diet. Feb. 10-11 99.8° to 98.0° 70 to 72 2200 + Trace? Light mixed diet. Thereafter until No dietary March 10 Normal Normal 1700 to 2700 No albumin restrictions. Case XXII.— (Dr. W. A. Clark, Oaldand, California.) The his- tory of Mrs. H., aged twenty-three years, as well as it could be obtained, brought out the fact that she had been markedly cedematoils and had had headaches and a scanty urinary output for several weeks past. At 5 P.M. of November 14, 1911, she gave birth to her first and living child. At 9 she had a convulsion that lasted fifteen minutes, and four more occurred in the night. She became unconscious. At 7 the next morning she was removed to the hospital. A severe convulsion occurred in the ambulance, and five more before 7.30 that evening. The patient vomited several times, the pulse ranged between 140 and 160, the respira- tion was 36 and the unconsciousness continued. In these twenty- seven and one-half hours she was effectively sweated several times, two magnesium sulphate and glycerin enemas were given, and two liters of 0.85 per cent sodium chlorid solution were given subcutaneously; 210 cc. of bloody urine filled with albumin were obtained by catheter during these hours. NEPHRITIS 583 At this time two liters of the following solution were given intravenously: Sodium carbonate (NajCOs' IOH2O) 10 grams Sodium chlorid I4 grams Water, enough to make 1000 cc. An hour later on involuntary urination occurred, of which only 320 cc. were caught; at 10.30 occurred a second, and at 11.30 a third. At 2.15 in the night the patient was rational for a few minutes, and at 8.30 A.M. she awakened completely. Through the day she was given 1000 cc. of the above formula by slow drip into the rectum, and water by mouth ad libitum. The total urinary output in this twenty-four- hour period was 3986 cc, and except for the first specimens which were bloody, the urine was fairly clear, intensely acid, and filled with albumin and casts of all kinds. The pulse which at the time of the intravenous injection was 140, rose to 150 after the injection, to fall gradually to 102 by the next morning. In the next twenty-four-hour period the injection by rectum of the alkahne hypertonic sodium chlorid solution was continued, about 300 cc. being injected and retained every six hours. Alkahne water by mouth was freely urged day and night. At 8 a.m. of November 17 the urine was neutral for the first time, and of a clear amber color. The total urinary output for this twenty-four-hour period was 3300 cc, with considerable albumin still present. The history of the next five days is indicated in the following summary: Date. Tempera- ture. Pulse. Respi- ration. Urine in 24 hours. Character. Remarks. Nov. 18 99.4°-97.8° 104-84 24-20 3840 + Dark amber to clear. Neutral, no albu- min. 600 cc. alkaline hypertonic salt solution by rectum. Glass of alkaline water every half hour, day and night, by mouth. Nov. 19 99.8°-97.8° 108-92 20 5216 Clear, neu- tral, no albumin. (Edema of limbs subsiding. SoUd food allowed. Al- kali by mouth only. Nov. 20 99.8°-97.4'' 108-88 20 3552 + Clear, neu- tral, no albumin. OEdema of limbs subsiding. Solid food allowed. Al- kaU by mouth only. Nov. 21 99.8°-97.8° 112-88 20 3840 Clear, neu- tral, no albumin. CEdema noticeable in flanks only. Gone from rest of body. Alkali by mouth only. On light general diet. Nov. 22 98.4°-98.0° 100-80 20 3600 Clear, neu- tral, no albumin. (Edema entirely gone. From this time on until her discharge from the hospital, December 5, her history was uneventful. Her diet was unrestricted except that alkali was added to her drinking water, given both day and night, and she was urged to salt her food. 584 (EDEMA AND NEPHRITIS Case XXIII.— (Dr. N. A. Hamilton, Franklin, Ohio.) During the last week of her pregnancy, Mrs. M., aged twenty-nine years, had oedema of the legs, and Suffered from impairment of vision to the extent of not being able to recognize her friends on the street. Her urine during this period was not brought to her physician for examination, though previous examinations had been negative. At term, on March 24, she went into labor, and was delivered normally of a living child by Dr. S. S. Stahl. The delivery occurred rapidly. The patient made normal progress until the fourth day (March 27), when she developed a temperature of 102° F., with a pulse of 110. This was attributed to an infection of the parturient canal. At the same time she became markedly nervous, complained of headache, and had great roaring in the ears. The urine became very scanty, and heavily charged with albu- min and casts. This condition continued until March 31. On this day the pain in the head became very severe, nausea and vomiting occurred, and a marked numbness of the right arm and leg developed. The patient was pale and generally oedematous. There was a twitch- ing of the muscles in various parts of the body. At noon on this day an active administration of sodium carbonate and sodium chlorid by mouth (0.33 gram each, every hour, with half glass of water) was started, and on the succeeding day this was given in the following form by rectum: Sodium carbonate (Na2C03 • IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. "^In the twenty-four hours during which the sodium chlorid and sodium carbonate were given by mouth, the patient's condition improved only slightly. On the following day, when administration by rectum was commenced, a rapid clearing began. The patient's history is summarized in the following table. 24-hour High- Highest period est temper- Urine. Medication. Remarks. ending pulse. ature. April 1 110 102.0° 720 Salt and alkali mouth. by Signs and symptoms as above. General twitching. Much albumin. April 2 115 102.2° 3420 Salt and alkali rectum. by General symptoms better. Albumin less. April 3 100 101.0° 2730 Salt and alkali rectum. by Patient comfortable; no head- ache, nausea or vomiting. (Edema less. Albumin very much less in amount. April 4 80 101.0° 3060 Rectal administration stopped. Alkali and salt by mouth only. Same. CEdema lessening. Albumin in traces only. April 5 78 101.3° 2700 Salt and alkali mouth only. by Traces of albumin only. April 6 84 101.1° 1680 Salt and alkali mouth only. by No albumin or casts. April 7 84 100.4° 1110 Salt and alkali mouth only. by No albumin or casts. April 8 80 100.4° 1530 Salt and alkali mouth only. by No albumin or casta. April 9 78 100.1° 1500 Salt and alkali mouth only. by All oedema gone; patient feels entirely well. No albumin or casts. NEPHRITIS 585 The temperature continued a few days longer, but otherwise the patient went on to an uninterrupted recovery. Case XXIV. — (Dr. Carl E. Curdts, Oakland, California.) Mrs. A. B. J., aged twenty-three years, entered the hospital, practically at term, in the early morning of July 14, 1912. She had always been well, and up to the eighth month of this, her first pregnancy had noted nothing abnormal. At this time her feet began to swell and her face became puffy, and some albumin and casts were found in her urine. In the ten days before entering the hospital she had been on a milk diet, and largely in bed. Nevertheless, she had grown progressively worse, her oedema becoming general and marked, and her urinary secretion dropping to " less than a pint measure " a day. Headache in the last two days had been constant. Her eyesight had been failing for a week past, so that on admission she could only distinguish light from darkness with her left eye, and with her right she could only recognize gross objects at a distance. In the last day before coming to the hospital, and during the morning at the hospital she complained of " twitching and nervousness." At noon of her first day in the hospital she passed 90 cc. of urine — the first since the night before, according to her statement. This urine was deep amber, rather syrupy, and acid to paranitrophenol. On boiling with acetic acid the urine set into a solid jelly. An Esbach determination, using citric-picric acid as the reagent, showed 14 grams of albumin to the liter. The urine was filled with casts, mainly hyaline. The patient's pulse was 102, her temperature normal. While immediate delivery by vaginal Csesarean section had been recommended, it was felt, for reasons to be discussed later, that no such immediate haste was necessary. She was, in consequence, given the following formula by slow drip into the rectum : Sodium carbonate (NajCOs- IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 ce. She took the drip well, and in the course of the afternoon absorbed and retained 720 cc. At 3.30 p.m. she voided 378 cc. of urine; at 6.45, 300 cc; at 8.55, 90 cc; and at midnight 315 cc. The patient's pulse gradually fell to 82 by 5.30, and remained here. At the same time she volunteered that her headache was much improved, and that she could see more clearly. In the night of July 15 she was given 420 cc. more of the solution by rectum. At 2.30 a.m. she voided 150 cc. of urine; at 4, 60 cc; at 8.30, 120 cc; at 10.20, 60 cc, and at noon, 63 cc. The total urine for these first twenty-four hours in which the alkaline hypertonic sodium chlorid solution was given was, therefore, 1536 cc In spite of the disturbance incident to giving her the solution, the patient insisted this morning that her headache was almost gone, that her eyesight had greatly improved, and that she felt more comfort- able than for several days past. The urine had gradually decreased in acidity, so that the second specimen obtained after starting the drip 586 CEDEMA AND NEPHRITIS was neutral to litmus. The Esbach determination showed a drop to 7.5 and 6.5 grams of albumin to the liter. The later specimens of urine clearly indicated that a higher salt concentration was prevailing in the body because the albumin on boiling with a drop of acetic acid no longer jellied, but was precipitated in flocculent masses. In the early afternoon of July 15 the patient experienced some uterine pains, which gradually became more severe. Because of them the rectal injections were stopped. The patient was given as a substitute 8 grams of sodium citrate daily, by mouth in several small doses. At 7.30 p.m. the pains became very severe and frequent, and at 11.30 she delivered herself of a healthy, living, male child. Ether was administered, and forceps were used to accelerate labor after the head appeared at the vulva. In the hours while the pains were on, the patient's headache increased, but otherwise she continued to feel well. The pulse rose from 82, when the first labor pains were felt, to 102 between the pains, when delivery was in actual progress. Up to the time of delivery she passed 270 cc. of urine, and at 10.30 the next morning 340 cc. were obtained by catheter. Both these specimens were dark amber, intensely acid to methyl red, and barely alkaline to paranitrophenol, gave a heavy curdy precipitate on boiling with acetic acid, and showed 15 grams of albumin to the liter (Esbach). Every field of a centrifuged specimen contained dozens of hyaline and finely granular casts. At this time her headache was slight, her eyesight more blurred than before labor, her temperature normal, and her pulse 88. At noon of this day, July 16, she was again started on the alkaline hypertonic sodium chlorid solution, and in the course of the afternoon absorbed 570 cc. At 4.15 she was catheterized and 690 cc. of urine were found, and at 11 p.m. another 780 cc. were obtained. The first of these specimens was paler than previous specimens, was still decidedly acid to litmus and methyl red, and contained 5 grams of albumin to the liter. Many granular and hyaline casts and red and white blood cor- puscles could be found in the sedimented urine. The second specimen was alkaline to methyl red and neutral to litmus and the Esbach deter- mination showed 2.25 grams of albumin to the liter. Only isolated casts could be found in the centrifuged specimen. The patient was now given 2 grams of sodium citrate by mouth every four hours, and milk and- lime water every two hours. Through the night and up to noon of July 17 she passed 1770 cc, making a total for this twenty- four-hour period of 3240 cc. This night and morning urine was clear, neutral to litmus, alkaline to methyl red, showed a few granular and hyaline casts, and contained 1.75 grams albumin to the liter. The patient slept fairly well, said her headache was almost gone, and that she could again recognize the details of her surroundings. At noon she had no headache, could read the smaller letters of a newspaper, and showed a decided decrease in her general cedema. She was now given an unrestricted diet, to which she was urged to add plenty of table salt. Vichy water was urged upon her, to which several grams of sodium citrate were added. To the milk which she NEPHRITIS 587 consumed lime water was added. Her history is summarized in the following table : For Grama 24-hour period ending Total urine. albumin per liter (ESBACH) . Remarks. at noon. July 18 3280 2.0 Unrestricted diet, Vichy water, milk with lime water and 8 grams of sodium citrate per 24 hours. Some casts and white blood cells. July 19 2520 + 2.0 Unrestricted diet, Vichy water, milk with lime water and 8 grams of sodium citrate per 24 hours. No general cedema visible. Eyes slightly puffy. Good appetite. Feels well. July 20 2366 + 2.5 Unrestricted diet, Vichy water, milk with lime water and 8 grams of sodium citrate per 24 hours. contaminated One speci- All signs of oedema gone. Only occasional cast in with lochia. men only. centrifuged specimen. Some red and white blood corpuscles. Urine neutral. The albumin does not precipitate easily on boiling. July 21 3000 + 1.7 No casts. Diet, Vichy water, etc., as before. Sodium citrate stopped and 0.7 gram of calcium chlorid every two hours day and night substituted for it. Urine faintly alkahne. Albumin comes down readily on boiling with acetic acid. Pulse 78; temperature normal. July 22 2300 + 1.0 No casts. Diet, Vichy water, etc., as yesterday. July 23 2250 + Trace No casts. July 24 1980 + Trace No casts. Calcium discontinued. July 25 2400 + Urine negative. July 26 2136 + Urine negative. July 29 2346 Urine negative. Left hospital. At the present writing this patient is entirely well. §5 We move from the protracted case of nephritis which is the result of a lasting intoxication of some kind by imperceptible steps over into the chronic nephritides. But as soon as we discuss the chronic nephritides we find that we have to distin- guish between those which represent a mere continuation of what was once a more acute process (the chronic parenchymatous nephritides, the secondarily contracted kidneys) and those which are chronic from the start, as in the type generally known as chronic interstitial nephritis (primarily contracted kidney) associated with changes in the vascular system. Very evidently, if our views are accepted, the nephritis which continues because of a protracted intoxication needs to be treated with alkali, salt and water in an equally protracted way. It has seemed to me that such a procedure yields good, and at times unexpect- 588 (EDEMA AND NEPHRITIS edly good, results, but I refrain from a detailed recital of such cases, for it is impossible, except as one works these things out for himself, to meet adequately the eternal argument that what has happened in such cases would have happened anyway. There is at hand, as a matter of fact, no dearth of the most objective sort of clinical evidence indicating the great value of alkali administration in such nephritides. F. T. Fre- EiCHS 1 emphasized it in 1851 and the older generation of physi- cians bore him out in this.^ A particularly careful cUnical study of the effect of alkali on the signs and ssonptoms of nephritis was made more recently by Eudolf von Hoesslin.^ He con- cludes that it is of great help in many cases, but fails in others, a view largely concurred in by such later studies as those of Glaesgen,* Ernst Romberg,^ Fraenkel,^ F. Conzen^ and M. W. SCHELTEMA.^ The use of alkah by many of these authors was purely empirical, and even the more recent studies do little to interpret the positive or negative findings beyond asking whether a relationship exists between the " degree of acidity " of the urine and the intensity of the albuminuria, the number of casts, the urinary output and the general symptoms of the patient. It would take us too far afield to give our own interpretation of the findings of each of the authors, but if in reviewing their con- tributions there is kept in mind what has been written in these pages apparent contradictions will quickly pass. As might be expected, the best results are always recorded when nephritides essentially toxic and evanescent in type are treated, while a persistence of casts, albumin, etc., is most definite when vascular and irremedial heart disease lie behind the urinary findings. In the former instance the " acidity " of the urine is an index ' F. T. Frbrichs: Die Brightsohe Nierenkrankheit, Braunschweig (1851). ^ See, for example, the standard texts of Senator, von Leube, Rosen- stein, OsLER, and Dieulafoy. ' Rudolf von Hoesslin: Miinch. med. Wochenschr., 56, 1673 (1909); Deut. Arch. f. klin. Med., 105, 147 (1912). * Glaesgen: Miinch. med. Wochenschr., 58, 1125 (1911). 'Ernst Romberg: Deut. med. Wochenschr., 38, 1073 (1912). ^ Fkaenkel: Deut. med. Wochenschr. 38 (1912). ' F. Conzbn: Deut. Arch. f. klin. Med., 108, 353 (1912). ' M. W. Scheltema: Toedienung van Alkalien bij Albuminurie, Delft (1914). NEPHRITIS 589 of what is happening in the whole kidney and a reduction in it is certain to be paralleled by improvement in urinary findings. When only pieces of the kidney are involved in consequence of blood vessel disease or localized infections, then the mixed urine coming from diseased and well kidney substance together may easily be neutral or even alkaline, and yet no impression be made upon the albumin output, etc. And since the so-called consequences of kidney disease are usually nothing of the sort they may, of course, appear with any kind of kidney findings. Especially is it difficult to m.eet the argument that whatever improvement is noted in a nephritic would have occurred any- way when we deal with the chronic interstitial tjrpe associated with vascular disease. One can from the start foresee that such offers the least possible chance of being markedly benefited by an alkali-salt-water therapy, and in its final stages none at all. I have emphasized this repeatedly, and were it not for the fact that it is upon this very type of case that some of my critics have based their arguments, it would scarcely be necessary to refer again to some self-evident facts. How much and what can we do for such cases? The primary change in chronic interstitial nephritis associated with vascular disease is not nephritis, but vascular disease. Every experimental fact and all physiological reasoning bears this out.^ In consequence of the vascular disease one piece after another of the kidney suffers destruction. And as this blood vessel disease cannot be and is not materially influenced by the injection of alkali, salt, and water, so also can this therapeutic procedure be of little or no use in this type of disease. The only cases in which it can be of service are those in which the blood vessel disease is in itself not wholly responsible for the observed changes, but where other temporarily active factors have been or are also responsible in bring- ing about our clinical picture. An illustration of this is offered in Case VIII, outlined above. Here to the picture of an established chronic interstitial nephritis associated with vascular disease, was added an intoxication with an anesthetic. The exacerbation is represented by the effects of the anesthetic upon the kidney, which effects are added to those already produced in this organ by the irremovable blood vessel disease. Cold, hard muscular work, an infection, or an 1 See page 482. 590 (EDEMA AND NEPHRITIS alcoholic spree might have done what the ether did. And the alkaU, salt, and water would have relieved the consequences of such added factors equally well; but blood vessel^changes that permanently interfere with the blood supply to a portion or all of an organ, especially when we deal with end-arteries, are not relievable by any such schemes. If this simple argument is borne in mind it will help to a better understanding of what may be, and what cannot be expected from the use of alkali, salt, and water in the chronic types of nephritis. Incidentally, we can also see what may be accomplished for the oedema, whether involving an individual organ or the whole body, in any case of heart disease. The final picture of a chronic interstitial nephritis is half the time not that of a pure nephritis, but one of this plus a failing heart. Only too often is the last insult administered to a remaining nubbin of kidney that for years, maybe, has served to keep a patient alive, by the cardiac muscle giving way (with a resulting generalized lack of oxygen, abnormal acid production and accumulation in all the tissues of the body, and so an oedema). When a heart, from any cause whatsoever, drops below the lowest level of an efficiency necessary to maintain a proper circulation, and has no remnants of recuperative powers left in it, alkali and salt can- not supply them.i 1 This is the type of case chosen by Joseph L. Miller (Amer. Jour. Med. Sci., 144, 8 (1912), Jour. Amer. Med. Assoc, 58, 1972 (1912)), upon which to test out the value of a salt-alkaU therapy. According to his own statement, the majority of his oases were nephritides with permanently decompensated hearts. Naturally, alkali and salt could not produce a diuresis where the mechanism for water secretion was about gone. Only heart tonics such as caffein and its derivatives, drugs, in other words, which through their action on the heart and respiration assured a temporarily better oxygen supply to the kidney and body tissues generally, gave a tem- porary " diuresis." It was not necessary to be a believer in any of the. colloid notions of water absorption to foresee all this, for, as we have known since 1860, an inadequate circulation will not allow even a normal kidney to secrete urine. More recently L. H. Newburgh (Boston Medical and Surgical Journal, 169, 40 (1913)), also concludes that the administration of alkali and salt is valueless or does actual harm in patients suffering from heart disease with broken compensation. To get at the real value of Newburgh's evidence one must center attention not on the apparently convincing argument presented in his main text, but upon the protocols attached thereto. While Newbxjrgh claims to have kept his patients on a fixed dietary and medical NEPHRITIS 591 §6 Cases XXV and XXVI will illustrate the application of an alkali and salt therapy to clinical conditions which are, in general, regarded as consequences of an intercurrent " nephritis." To us they seem rather to illustrate the essential sameness that exists between that which in the kidney is called nephritis and that which in other organs goes by such special pathological or clinical names as cloudy swelHng, stupor, coma, etc.; and as we found alkali and salts of service in the former, it will not surprise us to find them of service in the latter also.^ Case XXV. — (Drs. Charles G. Pibck and E. M. Baehb, Cincinnati, Ohio.) The patient, E. E., aged six years, had scarlet fever on May 19, 1912. The attack was characterized by an intense eruption, but relatively little fever, and no evident throat complications. He was up and around in less than a week. Twelve days later his mother called his physician because the boy had begun to complain of pain and distress in the throat. There was found an enlargement of the cervical glands, but nothing in the mouth or pharynx. The child grew worse during the week, becoming dull and listless, with no desire to eat, and sleepy and feverish. The boy's mother stated she knew he had not been passing a normal quantity of urine during this period. This condition persisted for two weeks, the child growing more and more listless until he was in a continuous state of lethargy. He was asleep most of the time and had to be aroused to eat. Only upon becom- ing aware that his feet had become swollen did the mother call the phy- sician a second time. At the time of his visit he found the child in a deep regime and then tested out the value of alkali and salt administration by adding this or taking it away, he actually did not do so. His patients received daily, in addition to a standard diet, digitalis and half an ounce (15 grams) of magnesium sulphate. On his test days he drops the digitalis and mag- nesium sulphate and substitutes a few grams of sodium bicarbonate. Of course, the urinary output had to fall and the oedema to increase, for what Newbtjbgh did was to substitute for the large dose of that most powerful protein dehydrant, magnesium sulphate, a small one of the weakly acting sodium salt, while rerdoving entirely the cardiac stimulant which alone was whipping up the heart to a point where enough oxygen was getting into the kidney to allow it to secrete any free water that might be brought it. His clinical results, aside from being fraught with experimental errors which vitiate his conclusions, could all be foretold. ' In connection with the idea that coma is an csdema of the brain occasioned by an accumulation of acid in it, it is an interesting fact that one of the harshest opponents of such a conception, namely, F. Marchand, has himself reported (Munch. Med. Wochenschr., No. 4 (1912)) the case of a patient comatose from poisoning with sulphuric acid who roused almost immediately after sodium carbonate was injected intravenously. 592 (EDEMA AND NEPHRITIS stupor with marked swelling of the face and feet. Slight convulsive manifestations were apparent. On the morning of June 2.5, almost one month, therefore, after the onset of the condition, the child was given an intravenous injection of two liters of the following solution: Sodium carbonate (NajCOs ■ IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. No anesthetic was necessary, for the child was comatose. The veins were not collapsed and the injection was given more rapidly than usual, requiring but fifteen minutes. The child's mental condition cleared quickly, so that three hours after the injection he was able to recognize his surroundings and take water when asked to do so. In the twenty- four hours following the injection he was made to drink two liters of pure water. His improvement continued steadily. In three days the oedema had completely subsided, and the urine was flowing freely. Samples of the urine obtained before the administration of the alkaline hypertonic salt solution were intensely acid and held albumin and casts in abundance. The urine obtained on the morning before the injection was begun showed casts as well as blood cells. Urine obtained a few hours after the treatment was alkaline and still contained albumin, but the casts had apparently disappeared. The mother stated that the child passed his urine on the morning following the injection after getting out of bed and securing the chamber himself. The patient was kept in bed for four weeks, and the recovery was rapid and uninterrupted. The urine remained alkaline to methyl red during this entire period. Albumin persisted in every sample examined, but no casts could be found. The quantity of albumin was always less than 1 gram as measured by the Esbach method. At the present writing (October 22) the boy is entirely well generally, but still has a small amount of albumin in his urine. Case XXVI. — (Drs. G. M. Allen and E. M. Baehe, Cincinnati, Ohio.) The patient, W. S., was a boy aged seven years, who had never before been ill. On December 10, 1911, his mother observed that he was ailing, tired, and listless. He complained of some distress in his neck below the left ear, where there could be felt an enlargement of the cervical glands. A low-grade fever was present at this time. His throat and ears were examined but nothing unusual was discovered. This state of affairs continued until December 27, when his temperature rose to 103° F., and albumin and casts appeared in abundance in the urine. The following is a brief synopsis of the development and course of the case: From December 27 to January 3, 1912: The temperature ran an irregular course varying in intensity from 99° to 104° F. The average quantity of urine passed in twenty-four hours was 416 cc. Albumin NEPHRITIS 593 and casts persisted. During this period he was given daily per rectum an average of 500 cc. of the following solution: Sodium carbonate (NaaCOa- IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. To January 10: The fever persisted, but was lower than during the previous week, fluctuating between 99° and 102° F. He passed an average of 640 cc. of urine daily. Albumin was constantly present, but was less in amount than before. On January 10 there was a slight transitory delirium; there was also observed a thin watery discharge from the left ear. The administration of the alkaline hypertonic salt solution was continued as before. To January 13: The temperature subsided gradually, reaching normal on January 12. The serous discharge from the ear continued. There was a leukocytosis of 34,000. On January 13, Dr. C. R. Holmes lanced the drums, but no pus was found. To January 17: The temperature remained normal on January 12, 13, and 14, and the urine free of albumin. The patient voided about 400 cc. daily. At this time a slight oedema of the face was seen, most pronounced under the eyes. His mental condition was quite good; on the morning of January 16 his mother read to him from his books, while he commented upon the pictures. That night without warning he developed a generaUzed convulsion which lasted two hours. The left side, apparently, was the worse involved. Dr. Holmes opened the left mastoid bone that night, the child having been given chloroform as an anesthetic. No pus was found. A second convulsion, lighter than the first, occurred later in the night. On the morning of January 17 the temperature was 103° F. The child was in a stupor from which he could not be aroused except with difficulty. There were continued spasmodic twitchings of the left arm and hand, and it was noticed that he did not move these parts as he turned in bed. The left leg also was not moved as freely as the right. Urine was voided involuntarily. Albumin was present in considerable quantities. At the request of Dr. Allen and Dr. Holt an intravenous injection was given by Dr. Baehr. One liter of the following solution was intro- duced into the superficial veins of the elbow. Sodium carbonate (Na2C03 • IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make 1000 cc. On account of the collapsed condition of the veins as well as the slight oedema of the tissues, the veins could be found only by dissec- tion. No anesthetic was required. During the ensuing night the patient was given 1600 cc. of water by mouth. In the course of the next twenty-four hours there were passed 1680 cc. of urine. Except in the earliest specimens there was no albumin 594 (EDEMA AND NEPHRITIS present even in traces. A blood examination made the morning after the injection showed a leukocytosis of 12,000. To January 21: The child rallied a bit in the course of the next twenty-four hours, the pulse rate and vascular tone remained quite satisfactory, and the mental condition cleared a little. It could be discerned definitely at this time that there was a complete left-sided hemiplegia, the arm and the hand being more severely involved than the leg. There was no fever, and the urine, for the most part, passed involunta- rily, was of satisfactory quantity, and always free of albumin. On January 19 and 20 he became fretful and irritable, the oedema of the face became more intense, and a suspicion was aroused that his vision had become impaired. On January 21 his general condition was bad; the pulse, which had been of excellent quality during the entire illness, now became very weak and rapid, and he slowly sank into a stupor which was practically a coma. His pupils responded feebly if at all to light stimulation, and Dr. Holmes believed he was able to discern a congestion of the retinal vessels, although no extravasation or oedema of the discs was found. After consultation a second intravenous injection of the alkaline solution was administered, this time in the right external jugular vein, as no other veins in the extremities could be located. Two liters were given. Chloroform was used as an anesthetic, chiefly to keep the child quiet during the operation. The coma seemed deep enough to allow of a much more severe manipulation. The entire time consumed was thirty minutes. To January 30: In the first twenty-four hours he passed copious quantities of urine involuntarily. Water was given him to drink in large quantities, and mUk alone was used for nourishment. Neither albumin nor casts were found at any time during this period. He was restless during the nights. His mental condition cleared rapidly. Three days after the injection a test of his vision was made. He was able to recognize his parents and the physicians about him. Convalescence was slow though uninterrupted: Albumin never reappeared in the urine except upon one occasion when there was found a httle circumscribed infection at the site of the first wound in the tissues of the elbow. As soon as proper drainage had been established the the albumin disappeared. The child had lost greatly in weight and strength during these weeks, and the entire period of convalescence consisted in obtaining an improve- ment of these conditions. The paretic disturbance of the left arm and hand persisted, but showed a slight improvement from week to week. During the following summer he was taken into the country about the Great Lakes, where he rapidly grew well. His father states that he left him there able to run about and paddle a canoe. The sole damage that remains is a tingling and stiffness in two fingers of the left hand. NEPHRITIS 595 §7 To this list in which the patients recovered I add a note on some acute nephritides in which the patients succumbed. A first fatal case is abstracted as Case XXVII. Case XXVII. — (Dr. Elizabeth Campbell, Cincinnati, Ohio). Miss E. T., aged forty years, a school teacher, consulted her physician a month before entering the hospital because she was constantly tired. She was thin, but without other physical findings of an abnormal nature. The urine was entirely negative. She was given an iron tonic, and urged to take a rest and to mcrease the amount of her food intake. On this regime her general health improved. February 24, 1912, she entered the hospital complaining of a sore throat which had developed two days previously. Both tonsUs were found enlarged and inflamed, and the lymphatic glands on both sides of her neck were swollen and tender. She suffered from pains in various parts of her body, and was nauseated. Her pulse was 112, and her temperature varied between 99.2° in the morning and 101.6° F. in the early evening. Her urine on the first half day after her entrance amounted to 300 cc, was dark amber, and acid, but free from albumin and casts. There was no oedema. She was kept at complete rest in bed on a milk and vegetable diet. By February 28 her throat had practically cleared, the neck glands had decreased in size, and her general symptoms had improved. Dur- ing this time she had taken frequent 0.3 gram doses of sodium bicarbonate with 0.25 gram doses of aspirin. Her temperature now varied between 99° in the morning and 100.6° F. in the evening. Her pulse ran between 96 and 112. On February 26 a trace of albumin was first noted in her highly acid urine as well as a few leukocytes. In the next two days the amount of albumin rose and granular, hyaUne, and epitheUal casts were noted. The urinary output per twenty-four hours was about 300 cc. At the same time diacetic acid and acetone in abnormal amount appeared in it. Sugar was absent as before. During these days the patient was frequently nauseated, and a marked generalized oedema developed. From February 28 to March 7 the patient's temperature steadily declined, so that on this day it was normal. Her pulse continued high, 104 to 120. Beginning on February 28, 480 cc. of the following solution were administered daily by slow drip into the rectum : Sodium carbonate (Na2C03- IOH2O) 10 grams Sodium chlorid 14 grams Distilled water, enough to make. .- 1000 cc. The urinary output in twenty-four-hour periods after this regime was instituted ran as follows: 330+ cc; 480+ cc; 540+ cc; 725 cc; 1080 cc; 1140 cc; 1050 cc; 1035+ cc. The urine remained acid all the time that these injections were given, and acetone and diacetic acid con- tinued to be present in abnormal quantities. The amount of albumin, which in the first three days of this period had been great, diminished 596 (EDEMA AND NEPHRITIS so that only a trace was noted in the later days. Medication during these days consisted of digitalin by hypodermic injection and occasional doses of strontium bromid at night. On March 7, when I first saw the patient, her general condition was so good that I merely approved of the scheme of treatment that was teing followed out. The rectal injections of alkaline hypertonic salt solution were continued as before. The day previously the patient had been drowsy, nauseated, and had had headache, but at the time of my visit the headache was less severe. That night she slept weU. The following day she was uncomfortable because of an accumulation of gas in the bowels. Her temperature was normal and the pulse 108. The evening of this day and in the night she vomited, though the rest of the night she slept fairly well. During these two twenty-foiir-hour periods 1140 and 1275 cc. of acid urine were voided, containing only traces of albumin, very few granular casts, a few red and white blood corpuscles, and diacetic acid and acetone in excess. The morning of March 9 was uneventful, but in the afternoon the patient began to complain of nausea. At 6 p.m. she vomited. The nausea and vomiting continued and became severe in the night. At 7 in the morning she complained of drowsiness; 1350 cc. of urine were voided in this twenty-four-hour period. At 7.30 A.M. (March 10) the patient was unable to swallow some proffered milk, and at 8 a.m. the nurse noted that the patient " had a far-away look in her eyes, and did not answer questions." Shortly afterward the breathing became labored. At 9.15 the nurse noted that the now unconscious patient looked constantly to the right, and at 9.30 a convulsion occurred. The convulsions were very severe and lasting, and chloroform was administered to control them. Crepitant and subcrepitant rales could be heard throughout the chest. Two 10-drop doses of tincture of veratrum viride were given into the muscles of the thigh. At 12, with no feeling that anything could be accomplished thereby, 1500 cc. of the alkahne hypertonic sodiimi chlorid solution were given intravenously. At 1.15, 300 cc. of urine were taken from the bladder by catheter, at 2.15 another 235 cc, and at 4.30 a final 75 cc. The first of these specimens, which included all the urine which had accumulated in the bladder since midnight, was heavily laden with albumin and casts. The second specimen contained relatively little albumin, the third again a large amount. All three specimens after the albumin had been removed reduced Fehling's solution heavily. The unconsciousness continued throughout the day, and the collection of fluid in her lungs increased. The patient could make no effort to cough it out, and the right to do a tracheotomy and practice insufflation was denied. Slight muscular twitchings were noted, but no more convulsive seizures. Oxygen played into the mouth helped but little. At 6 p.m., heavily cyanotic, the patient died. Autopsy performed immediately after death showed a well-marked oedema of the superficial tissues. The peritoneal cavity was dry. NEPHRITIS 597 The kidneys were somewhat swollen and of good color. The capsules stripped easily. The liver was smooth and somewhat swollen, the cut surface dry. The spleen was swollen so that the capsule was tense, but otherwise showed no changes. The pleural cavities contained a few ounces of free fluid. The lungs, crepitated throughout, were heavy and ran fluid from their cut surfaces. The apices showed some flat, thick scars in the pleura. The pericardial cavity was empty, the peri- cardium smooth. The heart muscle was flabby and slightly grayish. The endocardium was normal. Permission to open the head was not granted. In retrospect I feel responsible for the loss of this woman. In spite of the daily rectal injections of alkali, it is clear she did not get enough, as evidenced by the persistence of an acid reaction in her urine. I erred further on the day of my first visit in not' completely ignoring her good urinary output; and directing all attention to the well-marked brain symptoms evident the day before. As I have previously said, the state of one organ as evidenced by its function is not an index in these cases of the state of another. The alkali and salt should have been more indus- triously pushed on the day of my first visit, and subsequently. The patient should, moreover, have been given an adequate amount of dextrose (glucose) by rectum or intravenously. Her small daily intake of food with persistence of diacetic acid and acetone in the urine clearly indicated that a starvation " acidosis " was being added to the other conditions conspiring to produce her fatal brain cedema. §8. There are certain types of acute nephritis in which the toxic agent is of a kind to lead to irreversible colloid changes in the kidney from the start and in which therefore a hope of relief is small from the outset. Bichlorid of mercury when absorbed in sufficient amounts belongs in this class, as does phosphorus. While a number of my colleagues have reported relief with ultimate recovery of the partial or complete suppression of urine in bichlorid poisoning, I saw one patient in whom alkali, salt, dextrose, and water failed to elicit any response. She was seen for the first time in the third day of her anuria. When after twenty-four hours of alkali, salt and sugar intraven- ously no urine came, I urged a decapsulation, hoping to find a swelled kidney into which no blood was passing because of 598 CEDEMA AND NEPHRITIS compression of the blood vessels. Instead, the kidney was soft, gray and mottled in spots and streaks with yellowish-white areas of " fatty degeneration." The patient hved for nine days, in the course of which she developed no generalized oedema or any signs of a " uremia." Her blood pressure was normal at first, but fell on the eighth day. The volume of her pulse also fell slowly, disappearing at the wrist some eight hours before death. As the blood pressure fell her pulse rate increased and she became dyspneic. In this state, still clear mentally, she died. Two cases of phosphorus poisoning in children who had sucked the heads off some phosphorus matches died in almost identical fashion, though the urinary suppression in these had never been absolute. These three fatahties teach at the same time that not every- thing, including death, occurring in an individual showing casts and albumin, is at once to be regarded as consequent upon the kidney condition. The patients died of a " toxic shock " analogous to the " toxemic shock " that carries them away after the more protracted types of infection. Early in the eclampsial series that I have seen, the follow- ing fatality occurred. The patient, practically at term, threw herself out of bed in a convulsion early one morning. She was brought to the hospital late at night after the convulsions and coma had lasted through the day. No urine had been obtained since the night before, and catheterization was dry. She was injected intravenously with 1600 cc. of an alkaline hypertonic salt solution at midnight. Her convulsions stopped, and she cleared mentally so that at 4 a.m. she talked to her nurse. Between midnight and 6 a.m. she passed 412 cc. of urine filled with albumin and casts. She had no labor pains. At this time she was given ether and a vaginal Csesarean section was performed. There was marked hemorrhage, and at 9 she died. It remains a question whether this should really be counted a failure. The fatality occurred three years ago when I had less faith in the efficacy of a dehydration therapy and was less inclined than now to urge the quieter methods of delivery. The knowledge that an interruption of the pregnancy is synony- mous with a cessation of the intoxication is a constant argument in favor of speed. In looking at this side of the picture we forget all too easily that a third, and according to some statistics NEPHRITIS 599 a half, of all the convulsive seizures occuring in pregnant women do not take place until after delivery, in other words, not until the tremendous acid production of the muscular efforts of labor, the anesthetic, the bleeding, the pain, and the necessary- surgical procedures has been heaped upon that already incident to the pregnancy itself. The injurious consequences of all these must be subtracted from what we gain by speed before we obtain a correct estimate of the value of our therapeutic procedures. It is these facts that must also be kept in mind when the value of capsule stripping in nephritis comes up for debate. In at least some instances good has followed such a procedure. But this can be expected only if the deciding element between the recovery of the affected kidney and death is thought to be measurable in the increased circulation obtainable through the kidney by stripping the capsule. Even after the answer to this is given in the affirmative, then before operating, the effects of the anesthetic and the shock of the operation must be con- sidered, and not unless these are taken to be negligible should it be done, especially since experiment and clinical experience thus far is entirely one-sided in showing that all that can be gained through operation can be gotten by the simpler dehy- dration means of injecting alkali, salt and sugar. These injections are also of service in surgical operations in which by accident or design the blood supply to the kidney is temporarily occluded. The consequences of such a procedure are those of the experiments already detailed in which the blood vessels to the kidney were clamped. It has been shown by C. C. Guthrie ^ that perfusion with a physiological salt solu- tion or a Ringer solution of kidneys so treated affects them more deleteriously than if they are left alone. This is not because the sodium chlorid is poisonous to the kidney, as Lawrence Litchfield ^ has maintained, but because these salt solutions are not sufficiently concentrated and alkaline to prevent the swelling, etc., of the kidney cells. Most perfusion mixtures moreover lack the necessary colloids — the water in them is free, which is not the case in blood and lymph. 1 C. C. Guthrie: Arch. Int. Med., 5, 232 (1910). 2 Lawrence Litchfield: Jour. Am. Med. Assoc, 63, 307 (1914). 600 (EDEMA AND NEPHRITIS §9 A further recitation of cases could add little to what has been said. I should only like to emphasize once more the additional aid offered by a liberal use of the carbohydrates. When a desire for food is obtainable by an appeal to the apoetite so that the gastro-intestinal tract can be used in normal fashion, cereals of various kinds, with sugar, cream and salt work excel- lently. Candy in various forms is frequently desired and may well be given. Buttered and creamed toast, grapefruit juice with sugar, milk with milk-sugar added, etc., all not only prevent but help to relieve from a chemical point of view the " acidosis " so frequently observed in patients ill of any of a large number of causes. When the normal route is inadequate or unavailable then the carbohydrate must be given by rectum or intravenously. But when this is done it must be given in an immediately utilizable form, in other words, as dextrose (glucose) of a high grade of purity. As several hundred grams are necessary to cover the daily demands of the resting adult individual, too much dextrose can hardly be given. In mild cases a continuous administra- tion of dextrose with alkah is easily accomplished by rectum by using a solution containing 12 grams of sodium bicarbonate and 20 to 30 of dextrose to the liter of distilled water. This solution is somewhat " hypertonic," but does not irritate the rectum. When the patient begins to void neutral urine the absolute quantity of sodium bicarbonate injected daily may be reduced by lowering its concentration in the solution at the expense of the dextrose. The alkali and sugar must never be mixed until immediately before injection, as the alkali decomposes the sugar. For intravenous use I give the dextrose alone. When alkali and salt are needed they are injected at a separate time. In order to get as great a dehydrating effect as possible the sterilized dextrose solution is preferably given in highly concentrated form. Forty-five to ninety grams are dissolved in 100 to 200 cc. of freshly distilled water and injected very slowly (thirty minutes to an hour should be consumed) intravenously. To use haste is to lose the effect of the injection and to subject the patient to unnecessary risk. NEPHRITIS 601 §10 Space does not permit a detailed discussion of the problem but it must be apparent to the reader that this same alkali, salt, sugar and water therapy may be advantageously used and empirically has been used, in a number of other cUnical con- ditions in which a generalized or localized oedema is largely responsible for the signs and symptoms observed. Administration of alkahne hypertonic salt solution works well in the brain oedemas following injury,^ arsenic (salvarsan) injections, etc.; in the de- Urium, twitchings and convulsions seen in the acute infectious dis- eases; in scurvy; in angioneurotic oedema; in the marasmus of infants and children; in bronchial asthma; it frequently relieves the labored breathing of arteriosclerosis ^ and heart disease. C. C. FiHE has obtained excellent results by using alkah, salts and water in hay fever and mucous colitis, as has, in the latter condition, W. S. Kuder. When one deals with readily accessible oedemas, one can observe directly the good effects of alkali and salt when applied locally. A. E. Wright, J. L. Lohse^ and James J. Hogan have long employed alkahne hypertonic salt solutions (as 1 per cent sodium citrate and 2 per cent sodium chlorid, or 2 per cent sodium citrate and 1 per cent sodium chlorid) in all manner of superficial injuries, bums and infections as a wet dressing. A saturated solution of magnesium sulphate pro- duces a similarly agreeable dehydration of the tissues. This explains its long recognized virtues in reducing swollen rheumatic joints, etc. One can by wet bandaging with either of these 'W. B. Cannon (Am. Jour. Physiol., 1901) showed these to be due to changes in the brain itself. He held the increase in intracranial pressure to be due to an increased water absorption dependent upon an increase in the osmotic pressure of the cell contents. Brain swelling is more cor- rectly interpreted as a colloid-chemical phenomenon. Cannon's work has scarcely received its merited recognition. Had it, we should have been spared much modem clinical and surgical teaching which still considers blood pressure the source of increased intracranial pressure. This is para- mount to regarding the former as a source for energy greater than itseK, for the swelling brain is able to shut off its own arterial blood supply. A recent study of brain oedema in the terms of colloid- chemistry is that of S. J. KopETZKY (Trans. Am. Acad. Ophth. Oto-Laryn., 1913). ^ Dyspnea in arteriosclerotics is most commonly due to the high acid content of the blood secondary to cardiac insufficiency. It may, however, be secondary to an oedema of the lung dependent upon arterial disease of the bronchial arteries, or an oedema (acid intoxication) of the medulla itself dependent upon arterial change in the cerebral vessels. 'J. L. Lohsb: Lancet-Clinic: 107, 649 (1912). 602 (EDEMA AND NEPHRITIS solutions easily reduce the oedema of a vulva, penis or scrotum as observed in heart or kidney disease or in conjunction with certain infections. Similarly one can frequently render a patient who is hopelessly ill with heart disease more comfortable by dressing his oedematous legs, especially if we have made incisions into the skin, with these same solutions. John D. Long has injected sodium citrate and similar salts into the oedematous tissues in acute and chronic joint affections and attributes more than theo- retical value to the observed reduction in swelling thus obtained. And have we not long recognized the agreeable effects of applying mild alkalies to flea bites, mosquito bites and urticarial wheals and the value of using alkalies and calcium chlorid internally? Is the reduction of these annoying oedemas any different from the reduc- tion by like means of those produced experimentally on gelatin plates? At the same time such experiences yield ocular evidence of just what we are trying to do to internally situated organs when by any means we increase the alkali and salt content of the body tis- sues by administration of such by mouth, rectum, or intravenously.^ The large number of clinical conditions, involving so large a variety of organs, in which an alkali and salt therapy proves of service may seem at first sight somewhat strange. Surprise will disappear if we remember that the observed changes are merely a response to " injury." Whether such was induced by chemical, thermal, mechanical, or other means, it is always associated with an abnormal production and accumulation of acid in the injured part, as clearly evidenced by the electrical change called the " current of injury." Most Injuries to protoplasm are met by a reaction which in pathology is called " inflammation," ^ the most constant sign of which is, as we would expect, a " swelling " of the injured part, in other words, an oedema. 8. On (Edema as an Alleged Consequence of Sodixim Chlorid Retention It has been noted by different authors that in oedemas of various types, as in those associated with certain types of nephritis, with ' See in this connection the interesting experiments of Chiari and Januschkb (Wien. khn. Wochenschr., 23, 12, 1910), who observed the oedema of the conjunctiva following instillation of mustard oil to be markedly de- creased or entirely suppressed through sufficient calcium administration. ^ For a stimulating discussion of the colloid-chemical changes of infiam-- mation, gee Paul G. Woolley, Lancet-Clinic, 109, 360 (1913). NEPHRITIS 603 heart disease, etc., there is evidence of chlorid retention in the body. From this it has been quite generally concluded that in these conditions the kidneys are unable to eliminate chlorid (or, as ordinarily stated, are unable to eliminate sodium chlorid) and that its retention in the body is responsible for the oedema. Upon such reasoning has been based the widely approved therapy of sodium chlorid restriction, and since a lessening of oedema has at times been observed in patients following such restriction, the argument as a whole has been regarded as entirely sound. Against it have stood the failure of good observers to see any clinical improvement following careful efforts at apply- ing the principles of sodium chlorid restriction, and the experi- mental facts developed in this volume, which have all gone to show that the presence of salt, including sodium chlorid, in simple proteins, or in living cells, tissues, or organs always reduces the amount of water absorbed, either under normal circumstances or in states of abnormally great hydration (oedema) . These considerations have compelled the conclusion that sodium chlorid restriction as a scheme of therapy is not only wrong in principle but harmful in practice. It is the purpose of these paragraphs to indicate why sodium chlorid retention and oedema go hand in hand. Sodium chlorid retention is not due to an inability of the kidney to eliminate it, but to a change in the proteins {and other colloids) of the body as a whole. Sodium chlorid retention does not lead to oedema, but the changes which lead to oedema and to sodium chlorid retention are the same, consisting, in the main, of an abnormal production and accumulation of acid in the body. Proof of this may be brought from both a clinical and an experimental side. So far as the clinical aspects are concerned it is sufficient to emphasize that sodium chlorid retention is a constant accompaniment of all pathological conditions in which there is evidence of an abnormal production and accumulation of acid in the body, as betrayed through a persistently high hydrogen ion or titration acidity of the urine, an increased hydrogen ion acidity of the blood, an increased hydrogen ion acidity of the saliva or other secretions from the body, a high relative or absolute ammonia excretion in the urine, a low carbonic acid content of the blood or alveolar air, etc.^ Because ' See the succeeding pages 629, 638 and 648 (footnote) . 604 CEDEMA AND NEPHRITIS such acid intoxication is common to many different clinical states it is readily apparent why sodium chlorid retention has long been observed in pneumonia, in many of the infectious diseases, in the pernicious vomiting and intoxication of preg- nancy (eclampsia), in the cyclic vomiting of children, in diabetes, in carbohydrate starvation, in the severer types of the circulatory disturbances, in the severer anemias, in poisonings with arsenic, lead, phosphorus, chloroform, ether, and alcohol, and in the generalized parenchymatous nephritides when accompanied by a generalized oedema. The water retention likely to he observed in all these states is not secondary to the sodium chlorid retention, hut hoth are due to the existent acid intoxication. The presence of acid in abnormal amount in the body not only increases the hydra- tion capacity of the (protein) body colloids, but it increases at the same time their capacity for holding chlorid. Experimental proof of this may be easily brought. When protein (carefully washed fibrin or gelatin) is thrown into a salt solution, it not only absorbs water from the salt solution, but salt as well. Upon adding acid, the amount of water absorbed is increased, but the absorption of salt is also increased. The presence of acid, in other words, leads not only to greater swelling, but also to sodium chlorid retention. Experiments were made by placing weighed amounts of fibrin or gelatin in definite volumes of neutral or acidified sodium chlorid or calcium chlorid solu- tions of definite strength kept in carefully stoppered flasks. After varying periods of time the solutions about the swollen protein were filtered off and the chlorid in them determined by titration with silver nitrate according to Volhard's method. The following experiments will suffice to illustrate the values obtained : Experiment 93. — A dry gelatin plate weighing 0.813 gm. is placed in each of the following solutions : 1. 75 cc. m/6 NaCl+25 co. H2O. 2. 75 cc. m/6 NaCl+25 cc. n/10 HNO3. Titration (of an aliquot portion) of the surrounding fluids 19 hours later shows the former to contain 0.00303 gm. CI (=0.005 gm. NaCl) more than the latter. In other words, the gelatin in the acid solution has absorbed this amount of CI (or NaCl) more than the gelatin in the neutral medium. Twenty-three hours later the difference is still more striking. At this time the surrounding liquid in the first NEPHRITIS 605 mixture contains 0.00606 gm. CI ( =0.01 gm. NaCl) more than the second, or, conversely expressed, this much more CI (or NaCl) has been absorbed by the gelatin in the acid medium than by that in the neutral one. Experiment 94. — Three grams powdered fibrin are placed in each of four flasks containing, respectively, the following solutions: 1. 85 cc. m/8 NaCl+15 cc. H^O. 2. 85 cc. m/8 NaCl+ 5 cc. n/10 HNO3+IO cc. H2O. 3. 85 cc. m/8 NaCl+10 cc. n/lO HNO3+ 5 cc. H2O. 4. 85 cc. m/8 NaCl+15 cc. n/10 HNO3. Nineteen hours later the excesses of CI absorbed by the fibrin in the acid solutions over and above the CI found absorbed in the neutral mixture were, respectively, 0.00303, 0.004545, and 0.010605 gm., or, recalculated in terms of NaCl, 0.005, 0.0075, and 0.0175 gm. Experiment 95. — Three grams powdered fibrin are placed in each of five flasks containing, respectively, the following solutions: 1. 40 cc. m/6 NaCl+10 cc. H2O. 2. 40 cc. m/6 NaCl+ 2J cc. n/10 lactic acid+7i cc. H2O. 3. 40 cc. m/6 NaCl+ 5 cc. n/10 lactic acid+5 cc. H2O. 4. 40 cc. m/6 NaCl+ 7J cc. n/10 lactic acid+2^ cc. H2O. 5. 40 cc. m/6 NaCl+10 cc. n/10 lactic acid. Twenty hours later the excesses of CI absorbed by the fibrin in the acid solutions over and above the CI absorbed in the neutral mixture are, respectively, 0.0010605, 0.0037875, 0.006969, and 0.0113625 gm., or, ex- pressed in equivalents of NaCl, 0.00175, 0.00625, 0.0115, and 0.01875 gm. It has been customary in the modern writings on chlorid retention to assume that it is retained as sodium chlorid, and it is for this reason that in the experiments thus far detailed we have recalculated the value of the chlorid retention in terms of sodium chlorid. Chlorid is, however, better retained by acidified protein than by neutral protein even when it is offered in other form. Experiment 96 demonstrates the increased chlorid absorption under the influence of an acid in the case of calcium chlorid. Experiment 96. — Three grams powdered fibrin are placed in each of five flasks containing, respectively, the following solutions: 1. 40 cc. m/6 CaCl2+10 cc. H2O. 2. 40 cc. m/6 CaCl2+ 2^ cc. n/10 HN03H-7| cc. HjO. 3. 40 cc. m/6 CaCl2-|- 5 cc. n/10 HNOs+S cc. 11,0. 4. 40 cc. m/6 CaCl2+ 7^ cc. n/10 HN03-t-2| cc. H2O. 5. 40 cc. m/6 CaCU+lO cc. n/10 HNO3. 606 (EDEMA AND NEPHRITIS Twenty-four hours later the excesses of CI absorbed by the fibrin in the acid solutions over and above the CI absorbed in the neutral mixture are, respectively, 0.00303, 0.004545, 0.00606, and 0.010605 gm., or, expressed in equivalents of CaClj, 0.004735, 0.0071025, 0.00947, and 0.0165725 gm. It would take us too far afield to discuss why chlorid is better retained by protein in the presence of acid than in its absence. It is sufficient to point out that the observed behavior is probably nothing but an isolated illustration of the adsorption of a dissolved substance by a colloid,^ which, as is so frequently the case, is decidedly better when an acid is present than when it is absent. To the adsorption effect is, perhaps, added a chemical one, for there is some evidence that an acid protein combines chemically with neutral salts in a way that neutral protein does not. The increased amount of chlorid retained by proteins under the influence of acid is of a magnitude to cover easily any amount ever found retained by patients. Rarely are more than some 10 to 15 gm. of (sodium) chlorid held in the body. Choosing for illustration. Experiment 93 as a basis for calculation, we note that at the end of nineteen hours the acidified gelatin has absorbed an amount of sodium chlorid in excess of that absorbed by the neutral gelatin, amounting to 0.61 per cent of the original dry weight of the protein. At the end of another twenty-three hours the figure has risen to 1.22 per cent. If we choose the very liberal figure that but one-fourth the normal body weight is dry substance (of which more than 95 per cent is colloid) this means that a man weighing 75 kilos, developing the ability to retain even 0.61 per cent more salt, is already able to retain 114 gm. of sodium chlorid, or if the higher figure is chosen, twice this amount. These experiments have also a general biological interest in connection with the question of the " permeabihty " of cells to different substances. It is scarcely conceivable that anyone will maintain gelatin discs or fragments of powdered fibrin to be surrounded by " membranes " and yet observations of the type described in these paragraphs when made upon living cells or tissues are constantly cited as " proofs " of the existence of " membranes " about cells and of their alterable " perme- abihty." Thus, it has been argued that " hving " cells are surrounded by " osmotic " membranes " impermeable " to ' See pages 168 and 512. NEPHRITIS 607 sodium chlorid and to other salts, which become " permeable " upon the addition of acid or of substances which indirectly lead to a production and accumulation of acid in the cells (chloro- form, ether, potassium cyanid, etc.). Would anyone by similar reasoning maintain that gelatin plates and fibrin flakes lying in a " physiological " sodium chlorid solution are " osmotic " systems surrounded by " impermeable " membranes which be- come " permeable " to sodium chlorid when an acid is added? Perhaps these experiments will illustrate anew the fruitful consequences of the application of colloid-chemical principles to medical and biological problems. As they have proved adequate in the explanation of the many phenomena characteristic of water absorption they will also explain without contradiction the absorption and secretion of dissolved substances, 9. On the Treatment of (Edema. Comment on the Sodium Chlorid Restriction Therapy A generalized cedema constitutes so prominent a feature of certain cases of nephritis that it of itself becomes at times an object of treatment. From what has been said it is clear that this generalized oedema is not to be considered a consequence of the kidney state as is so widely done, but the " nephritis " is rather to be regarded as in good measure an csdema of the kidney, and so as part of the general process which gives all the rest of the body an increased water content. Since both theoretically and practically it is found that the swelling of the kidney may be reduced through salts, the rec- ommendation that the nephritic try to keep the salt concen- tration in his kidneys high follows as a matter of course. The thought, of course, at once suggests itself that this same scheme of treatment may be extended to the general oedema. While such a course has for decades been approved of empirically, as evidenced by the use of saline purgatives, saline diuretics, etc., in the treatment of oedema, a marked reaction against the giv- ing of salts in nephritis and in the oedemas accompanying it and other conditions has more recently set in. Of the scores of salts that might have been attacked in this way, sodium chlorid has been especially marked out, and to-day it is a widely accepted belief that the presence of this particular salt in the body is 608 CEDEMA AND NEPHRITIS responsible for the retention of water and so the oedema of neph- ritis, of certain cases of heart disease, etc. Evidence in support of this view has been entirely clinical. From our knowledge of their general physico-chemical activ- ities it cannot be understood why sodium chlorid should, of all the common salts that are found in the living organism, act in this specific way. That, as a matter of fact it does not, seems to me proved conclusively by everyday experience and the experiments detailed in this volume. Does the butcher not complain because his meats shrink when he salts them? And neither in the normal nor in the cedematous animal does an increase in its salt con- tent, sodium chlorid included, lead to an aggravation of the oedema. When rabbits are injected with progressively stronger solutions of sodium chlorid they lose progressively more water (shrink),^ while frogs developing a generalized oedema in con- sequence of poisoning with uranium (uranium nephritis (!) with casts, albumin and diminished water secretion), absorb decidedly less water if treated with sodium chlorid and other salts than when not so handled.^ These remarks hold for all the tissues of the body. Evarts Graham ^ showed me recently a pair of guinea pigs which had been subjected to the same degree of chloroform poisoning. One was subsequently treated with alkali and hypertonic sodium chlorid, the other not. On autopsy the kidneys and liver of the untreated one were swollen, dry and strongly mottled with grayish white patches of necrosis; those of the treated were of normal consistency, bled normally and showed less evident patches of destroyed tissue. The salts decrease cedema wherever found including that of certain types of nephritis, and sodium chlorid is no exception to this rule. How best to deal with the accumulations of fluid which so often occur in the peritoneal, pleural and pericardial cavities in the oedemas associated with heart lesions, kidney lesions, etc., is a matter to which colloid chemistry can also give answer. We know, of course, that a considerable ascites, hydrothorax, or hydropericardium may develop and disappear without ever assuming enough importance to demand clinical consideration. At other times, however, they become so great that they of 1 See page 287. 2 See page 218. " Evarts Graham: Personal Communication. NEPHRITIS 609 themselves give rise to trouble or at least add additional burden, as through their pressure effects, to the circulation, respiration, etc. These so-called transudates are identical with lymph and blood plasma, and it is for this reason that they may persist for days, weeks, or months in the body cavities being without absorbed. They are colloid solutions in which the solvent is bound to the colloid, and not until the solvent is rendered "free" can it be absorbed. When nature does not spontaneously remove them they can be gotten rid of only by tapping. That this is true is borne out not only by our previously described experiments ^ but by the well-known fact that blood and lymph extravasations into the peritoneal, pleural or peri- cardial cavities, whether encountered in man or produced in entirely healthy animals, remain here unchanged and undi- minished in amount for periods of time in which other aqueous solutions not containing such colloid material (which, in other words, contain " free " water) are readily absorbed. The fol- lowing experiments prove this : ExPBBiMENT 97. — A black and white rabbit is taken from its hutch, catheterized, and then weighed. Its weight is found to be 1493 grams. A slight opening is made in the abdominal wall and traction made on this so as to make the entrance of fluid into the peritoneal cavity easy. A second rabbit has the carotid laid bare for as great a distance as possible in the neck. It is Ugated high up, an artery forceps is attached to the coat of the vessel, a small forceps is placed below this, and the carotid is severed. This second animal is now placed in such a position that the blood will flow directly from its carotid into the abdominal cavity of the first animal, when the forceps is removed. The blood passes in a stream directly from the cut artery of the second animal into the peritoneal cavity of the first. This procedure is carried out at 2.40 p.m. The abdominal wound is closed immediately and the animal is weighed a second time to see how much blood has flowed in. The second weighing registers 1504 grams, which means that 11 grams of blood have flowed in. At the end of an hour the animal is killed by a blow on the head and immediately autopsied. The blood is found uncoagulated in the folds of the intestine. It is carefully aspirated into a tared flask and weighed. 11 grams of blood are recovered. Experiment 98. — In an entirely similar way a guinea pig, weigh- ing 520 grams, has a small opening made in its abdomen, and the blood from the carotid of a rabbit is made to flow directly into it. An increase 1 See page 261 and the first edition of my " CEdema." See also Jambs J. HoQAN and Martin H. Fischer: KoUoidchem. Beihefte, 3, 385 (1912). 610 (EDEMA AND NEPHRITIS in the weight of the guinea pig of 2.3 grams is thereby brought about. At the end of li hours the pig is iilled by a blow on the head and the unabsorbed blood is aspirated into a tared flask. 2.1 grams are recovered. ExPEBiMBNT 99. — A black and white rabbit, weighing 1630.5 grams, receives intraperioneally in the already described way enough blood from the carotid 'of a second rabbit to raise the weight of the for- mer 26 grams. At the end of an hour the rabbit is killed, and the unabsorbed blood is carefuUy recovered by aspiration into a tared flask. 26 grams of blood are recovered. ExPBBiMENT IOO.-7A white rabbit, weighing 767 grams, receives intraperitoneaUy 45 grams of blood from the carotid of a Belgian hare. At the end of seventy minutes the animal is killed by a blow on the head and the blood found in the peritoneal cavity is aspirated into a tared flask. 42.2 grams are recovered. There is nothing strange in the fact that the removal at times of a comparatively small amount, say of an ascitic accumula- tion, may be followed by a rapid absorption of the rest. As the amount of fluid in a serous cavity increases, the circulation through the surrounding tissues becomes more and more em- barrassed, and so the possibilities for absorption progressively poorer. To relieve this pressure even somewhat improves the circulation, not alone as to quantity, but as to quality of blood passing through the part (a blood more nearly arterial in character replacing a highly venous one). By thus- favor- ing the removal of carbonic and other acids always found in such serous accumulations ^ the power of the colloids here for holding water is decreased, and so further opportunity for the abstrac- tion of water from the transudates found in these cavities is brought about. What holds for the " transudates " and their absorption holds also, of course, for the absorption of inflam- matory " exudates." Various authors have claimed that, the administration of sodium chlorid (and other salts) is bad practice because it increases the accumulations of fluid in the serous cavities in the oedemas encountered in " parenchymatous nephritis," in heart disease, etc. How the salts may be effective in this regard is explained by the following. When any salt is given an cedematous individual his tissues give up water as do the frogs that have ^G. Stbassburg: Pfliiger's Arch., 6, 65 (1872); A. Ewald: Arch. f. (Anat. u.) Physiol., 663 (1873); Felix Hoppe-Seyler: Physioloeische Chemie, 1, 601, Berlin (1877.) NEPHRITIS 611, been described. But where does the water go? The body- weight as a whole can diminish only if this water is lost from the body through the urine (skin, gastro-intestinal tract, or lungs). But in a generalized nephritis and in heart lesions the kidney does not so readily rid the body of water as in health, and so this freed water must go somewhere, else. If it does not come out through some other emunctory (as in watery stools or sweat) this water can only escape into the cavities. What happens is identical with what is observed in experimental animals when they are made to give up their water very rapidly (especially after first rendering them oedematous by any means we choose) by injection of concentrated salt solution. I saw a good clinical illustration of the process in a patient of W. S. KuDER. A woman who for several weeks had been in bedj suffering from an extensive generalized oedema, with col- lections of fluid in the pleural cavities and abdomen, secondary to a heart muscle insufficiency of several years' duration, had the abdominal effusion removed by paracentesis. In order to keep up the drainage some strands of silk were left in the opening made by the trocar. Seepage stopped at the end of twenty- four hours, but the silk was left in place. On the third day a Hter of water containing 14 grams of sodium chlorid and 10 grams of crystallized sodium carbonate (Na2C03 • IOH2O) was given intravenously to combat the tissue oedema. This went down enormously, and as it disappeared the abdominal wound began to seep once more so that pad after pad had to be applied to absorb the liquid. James J. Hogan observed the same in a ten-year old child with a practically complete suppression of urine following an infection of the kidneys, in which a drain was left in the abdomen following paracentesis for an ascites. It is clear, therefore, that while the oedema of the tissues is reduced when salts (or alkali) are given an oedematous individual, the collection of fluid in the cavities may be increased. The thirst consequent upon the dehydration may lead the patient to drink water. In this way his total body weight (which in turn is taken as a measure of his oedema) may at times actually increase. But such a secretion of fluid into the peritoneal or other cavity is not by itself a particularly serious thing, for water and various salt solutions are readily absorbed from the peritoneal 612 (EDEMA AND NEPHRITIS (and other serous) cavities. If the effects are lasting it can only be because the added fluid has given rise to more permanent changes through added pressure, etc., or because it has had colloid material added to it (albumin) which after secretion renders the water unabsorbable. With the origin of the albumin we are not immediately concerned, though it is of interest to recall, after what was said regarding the origin of albumin in the urine, that the ascitic fluid may be looked upon as an albumin-con- taining secretion from the peritoneal tissues which, in its general composition and mode of origin, finds an analogue in the highly albuminous urine secreted by the kidney in acute nephritis. Syneresis and a " solution " of previously solid colloids occur in both.i How now, if my views are correct are we to explain the good results reported by Widal and his school when patients with oedema are salt restricted. Paradoxical as seems the answer at first sight, I believe that in this scheme of treatment a salt con- centration is also brought about but by indirect means. When the salt is taken out of the diet of a patient his appetite for water is tremendously decreased. If the patient is put upon the Karell dietary regime, care is taken from the start to assure not only a low salt intake, but a low water intake also. It is this water restriction in both cases that to my mind produces the reduction in oedema. We ordinarily overlook the fact that even under normal circumstances more than half the total water excretion is lost through the lungs and skin. Against the normal of, say, 1500 cc. of water lost as urine, man loses 500 cc. from his lungs, and 1000 to 2000 cc. through his skin. If the patient is in a warm room, or if he is warmly covered, or if, as is so frequently done, he is sweated, then these figures rise still higher. Even a com- pletely anuric individual therefore suffers a substantial loss in weight if for any reason his water intake is restricted. But no colloid mass, be this a drying gelatin plate or a water-starved human being, loses its salt to the atmosphere as rapidly as it loses its water. The salts therefore become more concentrated in the cells and tissues of the body. To realize what must be the effect of all this upon the total body weight is not difficult. Successive daily weighings will 1 See pages 240 and 434. NEPHRITIS 613 show such a patient to be losing a kilo or more a day and this is exactly what these clinical observers report. What a good effect this or any other effective dehydration scheme must have upon a nephritic kidney, a swelled brain or any similar condi- tion, can easily be appreciated when the vicious circles that may and do become established in any of these organs or in an extremity are kept in mind. I have repeatedly emphasized how an oedematous organ, not given free play to swell, tends to make itself progressively worse. As the organ swells, a ham- pering capsule, bony walls, or a small foramen make the sweUing tissues compress the blood vessels entering the affected organ, and so a lack of oxygen, in itself capable of producing an oedema, is added to the already existing factors responsible for such. Thus, sweating, drink restriction, or salt restriction, which indirectly amounts to drink restriction, all lead to a drying out of the tissues. If now a kidney in its process of swelling, say from an intoxication of some sort, has succeeded in rendering itself anuric by squeezing upon its blood supply, a more normal state may once more be attained if we rob the body tissues and fluids of their water content, and so also draw upon the water content of the swollen kidney. The net result must be a better blood supply through the kidney, and if the initial insult to the kidney was not too severe or of too lasting a character, the restitution of a better circulation may well be equivalent to a relief of the kidney condition — just as when alkali and salt are gotten into this organ. Then why not the water restriction-salt restriction scheme of therapy? I would answer that salt restriction constitutes an unnecessarily roundabout and by no means pleasant way of accomplishing a water restriction which can better be obtained by direct means. Second, as previously emphasized, intoxica- tion depends upon concentration. We can dull the effects of an intoxicant only by diluting it, and since we can guard our- selves against the bad effects of water by giving with it properly chosen salts in the right concentration, I have not been able to see the superiority of waiting for nature to dehydrate an organ which we can dehydrate as well and more rapidly, especially when in the former case we rob our patient of the advantage of having water available with which to float off his poisonous products. 614 CEDEMA AND NEPHRITIS I should like to re-emphasize a last point in this entire matter of alkali, salt, and water therapy. It must be remembered that no solution is either absorbed or secreted as such, but that in every case the water and dissolved substances move inde- pendently of each other, at times in the same direction, at others in opposite directions, and usually at entirely different rates. A primary purpose in giving the hypertonic solutions that I have advised is to secure an increase in the salt concentration in the body. Now the body cells and fluids in any state associated with an oedema have an increased capacity for holding water, and so take up water more easily than do the normal. The oedematous individual will therefore absorb water from any given salt solution more easily than will the normal individual. A solution, " hypertonic " for a normal individual, may be " iso- tonic " for one whose colloids have an increased hydration capacity. So it need not surprise us to find that when even a strongly hypertonic salt solution is given certain nephritics they may show an initial increase in their oedema. This only means that the colloids of their tissues were not previously saturated with water. Neither does it prove that the salt injected was responsible for this oedema, but only that the concentration of the salt which we succeeded in attaining in the body was not sufficiently high to decrease the existent high hydration capacity of the colloids of those tissues which showed the increased swelling. To remedy the situation we must give still more salt. (And at this time, as at all other times when we are trying to increase the absolute salt concentration in the body, we must give with the salt as little water as possible.) 10. Judging the Nephritic. Prognosis How we are to regard the patient with albumin, casts, etc., in his urine, and what must be the prognosis for him must be apparent from our previous considerations, and yet at the risk of repetition I again emphasize some rather obvious points because I am constantly questioned regarding them. The finding of casts and albumin of kidney origin in the urine merely tells us that destruction is occurring in the kidney and nothing more. It tells us nothing of the nature of the destroy- ing forces nor of their importance from the standpoint of the NEPHRITIS 615 patient. This information has to be worked out by other methods than urinalysis. Speaking generally, large numbers of casts and much albumin mean greater destruction than a smaller number with little albumin. Yet so far as the patient is con- cerned either finding may be of trivial or of great importance. The occasional cast with a trace of " albumin assumes great importance if it is the expression of blood vessel disease which involves the kidney, but not because of the kidney findings them- selves, but because a diagnosis of blood vessel disease with its important possibilities is thereby made. On the other hand, large quantities of casts with albumin in an active athlete or in a man who works to the point of becoming dyspneic (as in running for cars, making close connections, hard walking, etc.) may mean little or nothing, for any healthy man may get into such a state and over it in a few hours. Similarly, a man whom we first see in a convulsion and in whose urine we then discover casts and albumin must not at once be called a " chronic interstitial nephritic " the subject of a " uremic " attack. He may be such, but a convulsion from any cause (which really only means hard muscular work with respiratory interference) as in an epileptic fit, the spasms of strychnin poison- ing, or in the course of an infection, will put casts and albumin into a urine that never contained them before. Only if no such temporarily acting factors are at work are an increase in casts and albumin to be regarded as significant. The nephritic who with bed rest and proper treatment shows increasingly larger quantities of casts and albumin, is certainly not getting better, but good judgment must be used before it is stated too flatly why the patient is getting worse. Especially is restraint in order before any change in the patient's general condition is said to be due to the more evident signs of kidney involvement. More casts and more albumin in the urine mean, of course, more kidney involvement, but only clinical judgment can say whether the added factors he within the kidney itself or outside of it. A progressing blood vessel disease may involve larger and larger areas in the kidney up to the whole organ; or an infection originally Hmited to a spot or a few spots may spread to involve the whole kidney. But all too often, especially in the so-common chronic interstitial nephritides associated with blood vessel disease, the increase in the casts and albumin 616 CEDEMA AND NEPHRITIS coincides with the time the patient gets out of the ambulatory class, and the causes for the increase lie entirely outside of the kidneys. A failing circulation is discovered — a failing heart muscle with dilatation, leaking aortic and other valves, dilata- tion of the aorta itself, etc. — and the prognosis for the patient becomes that of his circulatory disturbances with little emphasis upon the kidney findings which in toto are now so largely dom- inated by and secondary to the heart failure. It is a safe rule whenever confronted by a patient showing simultaneously the signs of heart disease and of kidney involve- ment to look upon the heart as the greater offender and not the other way around, as is so often done. The matter is of much importance from the standpoint of clinical diagnosis, prognosis and treatment. The problem is often brilliantly illustrated by some of the so-called " orthostatic " cases of albuminuria. Many of these are really undiagnosed cases of cardiac insufficiency. Such patients may not show a single abnormal urinary feature when at bed rest, but the increased work incident to mere maintenance of the upright position for an hour or two make albumin and casts appear and continue as long as the upright position is persisted in. In the ambulatory nephritic presenting cardiac signs or symptoms or, to turn it about, in the cardiac patient showing albumin and casts in the urine I make use of the simple expedient of two or three days' rest in bed by way of estimating and eliminating the cardiac element in the production of his urinary signs. Many a patient with valvular disease, whose heart is efficient for certain low degrees of physical endeavor and in whom the cardiac process is not of a progressive type may be taught how to live and be given a more cheerful outlook upon life by not having a diag- nosis of " Beight's disease " superadded. But the impression must not be gotten from these remark^ that a cardiac element is the only one which accounts for the orthostatic types of albuminuria, nor yet that all which is eliminated by bed rest is of cardiac origin. The blood of an anemic individual may supply his kidneys when at rest in bed with enough oxygen to keep the urine free from albumin, but prove inadequate when he assumes the erect posture with its increased need for oxygen; or kidneys which in the prone poEi- tion are freely supplied with blood may yield albumin and casts NEPHRITIS 617 because their blood vessels are dragged upon when the patient rises. As previously noted, the CEdema observed in a sufferer from nephritis is never to be regarded as secondary to the kidney disease. In the earlier stages of the chronic interstitial types of nephritis associated with vascular disease we are not in the habit of expecting an oedema. When these patients do show such it means exactly what a generalized oedema always means — a generalized intoxication of some type. Such may be any of the many kinds which overwhelm the previously normal individual, but in actual practice their origin in the chronic interstitial types with blood vessel disease is almost always trace- able to something which can interfere with the normal (oxida- tion) chemistry of the whole body. The commonest of these are disturbances affecting the general circulation or respiration, wherefore careful consideration of the heart and its efficiency is not only demanded, but usually reveals the real source of the trouble. As a consequence of the cardiac disturbance with its resulting generalized oedema is to be expected an oedema of the kidneys, whence the increased casts and albumin and the decreased urinary output. The patients with chronic inter- stitial nephritis associated with vascular disease rarely die of their kidneys. Exclusive of the fatal accidents which may overtake anyone and which are especially prone to attack the vascular case, as hemorrhage, thrombosis, coronary disease, etc., they die almost half and half of faihng hearts which wUl not supply enough blood to keep the remains of their kidneys work- ing, or of oedemas of the brain secondary to the vascular disease (and wrongly called " uremias "). The prognosis of chronic interstitial nephritis associated ivith vascular disease is almost entirely the prognosis of vascular disease and of heart efficiency Exactly as our judgment of the chronic interstitial nephritic cannot rest too exclusively upon certain kidney findings, so also will their too intense contemplation in the parenchymatous types lead us into false paths. While in rare instances the kid- neys may be picked out alone to exhibit the picture of a gen- eralized parenchymatous nephritis, an intoxication is commonly at the root of the process, which affects simultaneously all the organs of the body. Hence the so common association of the oedema of the kidney with an oedema of all the tissues of the body. 618 (EDEMA AND NEPHRITIS Yet the one is not the consequence of the other, and so we may see complete suppressions of urine without a degree of surface oedema that can be recognized clinically, or a generaUzed (toxic) oedema in which the eyes are swollen shut, the skin stretched and shining, the fingers and toes swollen until they stand apart, and yet no urinary findings; or such developing after the gen- eralized cedema has persisted for days. The oedemas of the different organs are each to be sought for in turn and their intensity and importance judged alone from the point of view of the organ in which they occur. An oedema of the medulla and brain is more important than one of the kidney, and this than one of the liver. An oedema of the optic nerve is not dangerous to life, but, if persistent, destructive to vision. Many times the same degree of swelling in the skin has no lasting consequences.^ The variations from the normal blood pressure bring much clinical light, but no quick statements regarding their meaning can safely be made. The kidneys may be almost or com- pletely destroyed without the slightest increase in pressure. It may safely be said that generalized parenchymatous neph- ritis due to a poison or to an infection of the kidney never alone increases the pressure. Early in some intoxications and infec- tions there is a slight rise in blood pressure because of their effects upon the circulatory system, but when they persist they almost always lead to a gradual decrease in blood pressure (ending in the general picture of toxic or toxemic shock). A heightened or even very high blood pressure may be observed in patients having parenchymatous nephritis, but not because of this. The presence of vascular disease and certain heart lesions should first be considered. If they can be eliminated as factors, the pressure is nearly always the expresssion of brain cedema. An increasing or high blood pressure in the intoxication of pregnancy, for example, points not to kidney disease, but to a developing brain cedema due to the same poisoning which is producing such kidney signs as may be present. Except for such causes of high blood pressure as these in ^Exactly as the albumin content of the urine rises in oedema of the kidney (nephritis) the albumin content of the cerebro-spinal fluid rises in cedema of the brain or cord. Edmund M. Baehr first called my atten- tion to the' high protein content of the spinal puncture fluid in three "uremia" cases. The care necessary to avoid the trap of a wrong diag- nosis of brain syphilis on the basis of such findings is self-apparent. NEPHRITIS 619 patients showing abnormal urinary findings, a high blood pressure is most commonly due to blood vessel disease or certain types of heart lesions. Of the two, blood vessel disease is, of course, the commoner offender. A high blood pressure with cardiac enlargement, a few casts and albumin, when an old primary heart lesion can be ruled out, calls for a primary diagnosis of vascular disease. The maintained high blood pressure in such patients is not in itself to be regarded as a bad sign, it may mean a well- marked vascular involvement, but it also means good heart muscle. After the accidental variations due to work, excitement, etc., have been taken into consideration, a rising blood pressure usually means progress of the vascular disease or development of a brain oedema. A falling blood pressure is good when we can trace it to improvement in the vascular disease, but when such is not there, it becomes too often the sign of a failing heart muscle. It should also be remembered that a high systolic pressure alone, while it means a good heart muscle, does not yet mean an effective circulation of the blood. The diastolic blood pressure must be correspondingly high. A low diastolic blood pressure means that the effect of the systole in moving the blood forward is being largely lost, and the finding of an aortic leak or a dilating heart too commonly explains why it exists.^ These considerations render it clear why, in the so-called chronic interstitial types of nephritis, therapy should be directed primarily toward the relief of the blood vessel condition so far as such may be possible, and toward the maintenance of an effective heart action. To relieve the symptoms of the patient and to reduce the blood pressure, persistent, daily administration of sufficient alkali to keep the urine constantly neutral is un- doubtedly the most effective single thing we can do. Arthur D. Dunn came to this same conclusion entirely independently. It is remarkable how the high blood presssure will fall and remain low. As attempts to meet the vascular disease itself may be fisted (because of my belief in its infectious origin) the approved hygienic means generally considered effective in com- bating such, as fresh air, rest, and good food, or when syphilis is suspected, iodids, mercury and arsenic preparations in addi- tion. To bring aid to a distressed heart the use of digitalis and 'See in this connection the interesting studies of Willard J. STOffE: Jour. Am. Med. Assoc, 61, 1256 (1913); Lancet-Clinic, 111, 247 (1914). 620 (EDEMA AND NEPHRITIS other cardiac stimulants, even though they raise the blood pres- sure, have long been known to be of service. Such therapeutic measures yield better results than the drastic dietary restrictions or the empirical incantations so often invoked over this long- suffering group of patients. It is apparent also why measures which merely reduce blood pressure (except in cases of hemorrhage) are not only disappointing in their results, but may actually do harm. I have several times seen alarming falls in the urinary output and once a complete suppression of urine with death of the patient eight days later after the administration of nitrites to reduce blood pressure in cases of chronic interstitial nephritis associated with vascular disease. Suppression of urine is bound to follow the lack of blood supply thus induced to kidneys which are already barely getting enough with a high blood pressure. There is no justification for giving nitrites in chronic interstitial neph- ritis unless we can show that while reducing general blood pres- sure we are not at the same time reducing the blood supply to the kidney down to a dangerous point. And yet these considerations must not be interpreted as an interdiction of the use of nitrites in all clinical conditions. When we deal with symptoms referable to vascular disease or " vascular spasm " affecting particularly but one organ, as in angina pectoris, the nitrites may well be tried in the hope of obtaining a vasodilatation in the affected organ without losing enough in other organs to do harm. But whether we will or will not get such an overplus of good effects can be determined only by careful observation of the individual patient when such vasodilators are first tried. It should, moreover, be borne in mind that, as A. W. Hewlett I think first showed, the nitrites show an initial increase in blood pressure before the fall is obtained which is so generally looked for. This initial rise in blood pressure, assuring a better blood supply to a stricken organ, may possibly be the real source of the relief experienced. The possible danger of this preliminary effect in cases of hemorrhage needs no special emphasis. We should not fail in looking for the factors capable of pro- ducing the acid and similarly acting substances which we hold to be directly responsible for the development of the signs and symptoms of a nephritis to consider foci of infection in the NEPHRITIS 621 kidneys themselves or elsewhere in the body. As , previously emphasized, the effect of a poison upon a kidney is the same whether it is produced locally in the kidney or elsewhere in the body and brought into the kidney by the blood stream. The importance of searching for foci of infection in the tonsils, teeth, ears, and mastoids, in the sinuses of the head and face, in the genito-urinary tract of men and women, as well as elsewhere in the body is often brilliantly emphasized by the clearing up of long-standing cases of albuminuria. How comparatively trivial and constantly neglected foci of this type lead to destructive changes in many of the organs of the body (including the kidney) has been beautifully demonstrated by the chnical, bacteriological and experimental studies of the Chicago school (Frank Billings, Edward C. Rosenow, Rollin T. Woodyatt, D. J. Davis, and their collaborators).^ With these facts in mind, which show how the whole patient and not only his kidneys must be seen if we would judge rightly the meaning of his urinary findings, we shall consider next the meaning and value of some methods which have been resorted to by way of determining kidney efficiency. 11. Underlying Principles and Clinical Value of Kidney Efficiency Tests Clinicians have for many years and in many ingenious ways tried to determine the physiological efficiency of different organs. How valuable would be at times a correct knowledge of the working power of the heart, of the chemical activity of the liver, of the secretory activity of a kidney is, of course, obvious, and yet the bitterness of controversy which surrounds the various 'To the long list of "metabolic'' diseases already eliminated by these workers we must, I think, add gout. In five patients — 4 men and 1 woman — falling strictly within the text-book type of the disease (nocturnal big toe attacks, urate tophi in ears and skin, unsymmetrical urate deposits in finger joints, susceptibility to nucleo-protein feeding, etc.) badly infected teeth were observed in all. Two have recovered absolutely since losing all their teeth. A third who would develop "gouty" attacks whenever his teeth were treated by a dentist is better. The remaining two still have their teeth — and their gout. May we not better regard these patients as suffering from emboUc infectious arthritis with the urate deposits secondary thereto and analogous to the formation of "stones" in infected gall bladders, kid- neys or urinary bladders? The same infectious embolism in other organs will then explain the alleged consequences of gout, as the occasional fever, 622 (EDEMA AND NEPHRITIS attempts that have been made to estimate such scientifically suffices to indicate how far we are still from the desired goal. Before my own position in the matter, so far as the kidney is concerned, is expressed, I would emphasize the necessity of bearing in mind some physiological truths, the ignoring of which has led to the expectation of obtaining by efficiency tests on different organs impossible ends, or to a condemnation of such as have real, even though rather limited, value. With the exception of certain portions of our central ner- vous system it is characteristic of our organs that they each contain several times as much material as is necessary to meet the work requirements of every day. Seven-eighths of our pan- creas may become functionless and still enough internal secre- tion be obtained to keep us from becoming diabetic; the work demands on the heart may be trebled or quadrupled and yet it goes on; half or three-quarters of the liver may be destroyed with no signs of a hepatic insufficiency. Similar claims may be made for each of the organs of the body. We possess, in other words, a potential power many times that actually used. The important corollary of this fact is that an organ continues to show a normal function as long as more than such a physiologically necessary minimum remains preserved, even though large pieces of it may actually be temporarily functionless or totally destroyed. The more definitely more than such a phy- siological minimum is still preserved the less can any ordinary- test give us an inkling of the actual amount of damage. The test has to be heightened to the point of straining all of the normal or imagined remnants of an organ before a defect here will manifest itself. These considerations hold for the functional testing of all organs. How they apply to the specific problem of the kidneys is evident from what follows. The mystic element of human nature is well expressed in its so commonly voiced faith that it is the function of the kid- the muscular involvement, the nephritis, the occasional leucocytosis, etc. We are only slowly beginning to learn the multiplicity of pathological effects exerted by one and the same organism when grown under different environments. See in this connection the fundamental work of William B. Wherry: Jour. Inf. Dis., 2, 436 (1905), cholera red reaction; Arch. f. Protistenkunde, 30, 77 (1913), flagellation of amebae; Centralbl. f. Bakt., Parasitenk. u. Infektionskr., 70, 115 (1913), "spore" formation in tubercle bacilli; Jour. Inf. Dis., 13, 114 (1913), acid proofness in tubercle baciUi. NEPHRITIS 623 neys to excrete the poisons produced in the body. Such a view ignores the fact that they are quite as likely to be excreting substances for a lack of which the animal is perishing, as when in salt stjarvation they continue to secrete this even though its lack in the body is kilhng the animal. But however we may choose to look upon such obvious facts, it is evident that the kidney's ability to give off dissolved substances may serve as one type of kidney efficiency test. To most this has appeared the only test, and yet we need but recall that a dry kidney can get rid of nothing at all to show how its abiUty to give off water is really quite as important, in fact, as previously emphasized,^ it is the primary function of the kidney, and the secretion of all dissolved substances is secondary to this. The efficiency of a kidney may therefore be gauged by its ability to secrete water. Before discussing the utilization of these considerations in clinical practice, some animal experiments in which we have conditions under better control than in our clinical practice had best be borne in mind. The best way to do away with any amount of organ function is to remove parts of the organ. In the case of the kidneys it is an easy matter to take away three-fourths and even more of the total substance and yet have the animal live. Such removal means, of course, a loss of three-quarters of the normal kidney function. Animals so operated upon excrete water and all dis- solved substances, such as the various dyes, exactly as do normal animals. Even excessive amounts of water given such animals by way of placing " strain " on the remaining portion are easily cared for. The important conclusion to be drawn from this is that mere loss of three-quarters of the normal kidney function (and even more) does not yet betray itself to any of our ordinary functional tests. This fact is of great clinical importance because it means that ordinary efficiency tests tell us nothing until we are working on the last quarter or less of our total kidney efficiency. Efficiency tests, as we shall see, still retain a useful purpose in spite of this, though it might as well be emphasized now that when we are thus working on the last elements of a still function- ing kidney very obvious clinical signs will in most cases tell us quite as much. It is well now to re-emphasize that the secretion of dis- solved substances is secondary to the secretion of water, not 'See page 324. 624 (EDEMA AND NEPHEITIS in the sense that if a liter of urine brings out a gram of dissolved substance two liters will bring out two, but in the physico-chemical sense ^ that if no water at all is secreted into the uriniferous tubules no dissolved substance at all can be washed out of the if kidney parenchyma and therefore none be secreted. If some water comes through then some dye comes out also, while with more water more dye will be washed along. But the time that the water remains in contact with the parenchyma plays an important part in permitting diffusion, etc., wherefore the con- centration of the dye may be greater with less urine than with more, but other conditions the same, the absolute amount never. The increased output of any dissolved substance when, other conditions remaining the same, the amount of water passing through the kidneys is increased, has for half a century been one of the well established facts of physiology. Experimenting upon himself C. Genth,^ for example, found that during a con- trol period he voided a daily average of 1252 cc. of urine with 40.2 grams of urea. In a subsequent period when everything was kept constant except that an added two liters of water were consumed he secreted 3250 cc. of urine with 46.6 grams of urea. When the water intake was increased to four liters the urine rose to 5500 cc. and the urea to 54.2 grams. Similar results are obtained if the observation periods are measured by hours instead of days.^ While such increases have been quite generally attributed to an increased " protein metabolism," they probably represent pure washing-out processes, for upon return to the original water intake the urea excretion falls not only to the nor- mal, but for a period to a figure even below this. What has been said here of urea holds also for the total urinary nitrogen or some of its other fractions, as ammonia, creatin,* creatinin, and alantoin.^ It holds also for such salts as sodium chlorid.® The whole matter is easily demonstrated on animals with nor- mal or reduced kidney substance. With ordinary feeding rabbits 1 See pages 164, 324 and 512. ^ C. Genth: Ueber den Einfluss des Wassertrinkens auf den Stoffwechsel, Wiesbaden (1856). 'Oppenheim: Pfluger's Arohiv, 23, 465 (1880); see also Neumann: Arch. f. Hyg., 36, 248 (1899). *C. C. Fowler and P B. Hawk: Jour. Exp. Med., 12, 388 (1910). 'L. T. Fairhall and P. B. Hawk: Jour. Am. Chem. Soc, 34, 546 (1912). « S. A. RuLON and P. B. Hawk: Arch. Int. Med., 7, 536 (1911). NEPHRITIS 625 eliminate 55 to 70 per cent of a dye (1.5 milligram phenolsul- phonephthalein) in the two hours following its introduction intramuscularly, but if the water secretion is heightened by intravenous injection of salt solution 65 to over 80 per cent will be obtained. While the concentration in the individual samples of urine in the latter case is lower than in the former the absolute amount given off is not. With plenty of water coming through the kidney the dye also appears earlier than when such is not the case. Exactly the same occurs in patients. With little water coming through the kidney the output of potassium iodid, of milk sugar, of sodium chlorid, of a dye, etc., is lower than when the same test is repeated on the same patient, but a heightened wa'ter secretion is assured by a greater water intake. I have been told that J. Albekkan first taught that a " poly- uria " following the administration of water was the best evidence of a patient's kidney efficiency. Alberran's original communication is not accessible to me, but it expresses a view with which I heartily concur. I have never seen a patient showing a normal water secretion who did not also show a normal output of any of the dissolved substances customarily used in kidney tests. To have a water, or any other type of kidney test mean any- thing, it is evident that all the factors outside the kidney which are capable of influencing it must be eliminated as far as possible. If, for example, we do not get a normal water response this does not at once mean that the kidneys are affected. The patient's body colloids may not have been previously saturated with water, or exercise may have temporarily increased their water- holding power, or a circulation may be so defective that the blood never becomes properly arteriahzed and so " free " water for urinary secretion appears only slowly in it. To eliminate as many of these factors as possible it is nec- essary in making efficiency tests on a patient to put him at bed rest for at least twenty-four hours and let him have an unre- stricted diet. The best time to make a water test is two and one-half hours after breakfast or two hours after the last water was drunk. It takes a normal person about this time to elim- inate whatever free water exists anywhere in his body. When this state has been attained, but not one beyond this in which the tissues are beginning to dry out, the patient empties his bladder (or if absolutely necessary a catheter may be inserted) 626 CEDEMA AND NEPHRITIS and 500 cc. of drinking water are consumed. A normal person (or one having what will be called a normal kidney efficiency) excretes 400 to 500 cc. of the consumed water in the next two hours. The excretion shows an optimum point at the end of about an hour. The results of a few actual tests will show how the curves run. TABLE LXXIX Urinary Output in cc. Infection of kidney (un- known type). Casts One kidneyremoved Time in minutes and hours. Vascular disease with and polymorphonuclear for infection. Re- Normal. high blood pressure, cardiac hypertrophy, leucocytes from the kidneys constantly in maining kidney has shown large num- no heart murmurs, urine. Occasional slight bers of casts, leuco- casts and albumin rises in temperature. cytes and albumin constantly in urine. No abnormal findings for several years. in rest of body. .15 Test 1 Test 2 25 40 19 20 37 .30 50 48 31 31 43 .45 76 119 65 59 43 1.00 72 148 , 119 96 53 .15 62 51 128 92 59 .30 65 44 78 58 54 .45 57 34 47 25 49 2.00 28 26 29 23 42 Total. . . 435 510 518 405 380 A secretion of water above the total of the 500 cc. consumed is usually attributable to the fact that free water over and above that administered existed in the body at the time the test was started. If a patient consumes water against instructions, or if in the digestion of his food considerable amounts of water are freed, these quantities are, of course, added to the 500 cc. admin- istered. Similarly, a low secretion does not at once mean an involvement of the last quarter of the total kidney substance. It may simply mean that the patient was not saturated with water and calls for an immediate repetition of the test. As the first administration of water usually suffices to saturate the body colloids the second test is usually more reliable than the first. We need not specially emphasize that the outcome neither of the water test nor of any other functional test depends solely upon the functional abiUty of the kidney. A heart lesion, extensive interference with the respiration, etc., may all give rise NEPHRITIS 627 to a deficient water output, and yet the kidney itself be but little affected. Neither do any of these tests tell us what is the pathological background for the disturbance in the kidney. They only tell us if the functional capacity has fallen below a certain minimum, regardless of the fact whether this disturbance is of a type which in a few hours will be removed, or is due to permanent destruction. 1 make it a rule to use with the water test one of those in which the elimination of a dissolved substance is followed, and the well worked out phenolsulphonephthalein test of J. T. Geraghty and L. G. Rowntree ^ is perhaps the simplest and best of these. Geraghty and Rowntbee provide for an adequate water secretion in their test in advance by having the patient consume several hundred cubic centimeters of it. While I lay greatest stress on the water secretion capacity of a kidney and have never found one which if it would eliminate water would not also eliminate a dissolved substance, the use of a dissolved substance capable of rapid elimination, as phenolsulphonephtha- lein, along with the water test furnishes a valuable check. If a kidney is functioning well, but the water elimination is low due to some accidental factor such as insufficient water consump- tion, the disproportionately higher elimination of the dye at once betrays the fact, for the dye is washed out of the kidney by very little water. The real significance of the low water excretion due to such accidental factors is therefore immediately disclosed. A set of observations on the excretion of phenolsulphone- phthalein and other dyes in experimentally induced nephritides, interesting because of their bearing on the interpretation of clinical findings, has recently been made by 0. Schwarz.^ He finds that when the acid content of the body is increased (resulting, as we have previously seen, in a retention of water and a diminished urinary output with casts and albumin in the urine) phenolsul- phonephthalein under these circumstances also fails to be excreted. This is what would be anticipated on the colloid-chemical basis. These experiments teach at the same time how extrarenal factors play an important part in determining dye output. Phenol- sulphonephthalein is subject to the distribution law as is every ^J. T. Gebaghty and L. G. Rowntbee: Joufn. Am. Med. Assoc, 57, 811 (1911); 60, 191 (1913); 61, 939 (1913); Arch. Int. Med., 9, 284 (1912). 2 O. ScHWABz: Pflliger's Archiv, 153, 87 (1913). 628 (EDEMA AND NEPHRITIS other substance, and so under the influence of acid, salts, etc., is taken up or rejected by protoplasm elsewhere in the body as it is in the kidney. Extrarenal factors, therefore, help in determining what shall be the output of dye by the kidney, a fact which must teach us not at once to attribute to changes in the kidney alone any observed deviations from the normal dye output. Classical parallels of such experimentally demon- strable facts are common. I need but state that in the acuter types of heart failure the dye excretion drops down at once, but rises again as soon as the heart recovers. Here the whole body, and incidentally the kidney, is suffering from a lack of oxygen with its accompanying abnormal production and accumu- lation of acid. How much more important .it is in this illus- tration to recognize the heart condition rather than the incidental kidney involvement needs no emphasis. Schwarz has, more- over, brought experimental proof of what was predictable theoret- ically.^ After acid injection (with its accompanying nephritis) some dyes are excreted even better than normally. If these considerations are borne in mind it will serve to explain why I hold that the best evidence of kidney efficiency is its ability to excrete water. A kidney capable of secreting water is secreting in proper fashion everything else, and as long as suf- ficient water is supplied the patient and comes through, his state is not to be attributed to defective kidney elimination. This is true even of the alleged " uremias " encountered in chronic interstitial nephritis. The use of dyes in testing kidney efficiency finds its greatest service, perhaps, in deahng with one sided kidney lesions. While ureteral catheterization will give us the water output from each of the two kidneys it is not devoid of danger. Cystoscopic observation of a properly chosen dye coming freely from one ureteral opening while little or none of it is coming from the other often suffices to betray the almost functionless character of one of the kidneys. The use which may be made of functional tests is self-evident. They are neither to be condemned as universally as they have been by some writers, nor yet to be regarded as alone capable of giving a correct index of the state of the kidney, as assumed by ^ See page 510. NEPHRITIS 629 others. Except in one-sided kidney lesions where it is important to determine, as before surgical interference, the functional capacity of the kidney to be left, very evident clinical symptoms usually suffice to let us know when we are working on the last quarter (or eighth) of our total normal kidney capacity. When with adequate water supply by mouth a kidney secretion remains persistently low, mere measurement and examination of the twenty-four hour output suffice to tell us of the state of the kid- neys, and functional tests will not add much. The claims of some authors who have felt it possible to predict the onset of " uremic " attacks and death in patients with chronic interstitial types of nephritis from their low output of some dissolved sub- stance through the kidney need to be restudied. As previously emphasized, animals and patients deprived of their kidneys do not die as do those which clinically we diagnose " uremic." These " uremic " deaths are due to cedemas of the brain, and such are secondary, not to loss of kidney function, but to vascular disease, to thromboses of the cerebral vessels, and to hemorrhage. Patients with chronic interstitial nephritis associated with blood- vessel disease die once in three times of a failing heart. And since individuals with uncompensated heart lesions fail to excrete water and dissolved substances exactly as do such as have their kidneys primarily involved, it may well happen that death is correctly prognosticated in such from a low dye output, but, to repeat, the death is again not due primarily to kidney disease, but to cardiac involvement. 12. On Acidity Measurements of the Urine Such stress has been laid upon the importance of an abnormal production and accumulation of acid in various organs of the body in giving rise to an oedema in them, or in the case of the kidney, in leading to the development of the signs of a nephritis, that the question naturally arises whether we cannot in some way measure this acid content by way of getting an index of the severity of the changes occurring there. We have not at present any methods which can be used clinically for measuring changes in the acid content of the different organs in the body. We can, however, measure the acidity of the blood which bathes them 630 CEDEMA AND NEPHRITIS and of the different secretions from the body, as the urine, the saliva and the sweat. Other things remaining the same, the laws of chemical equihbrium then permit us — from the acidity changes in these secretions — to conclude that there must have been similar changes in the tissues from which they came. Thus, an increase in the acidity of the urine means, generally speaking, an increase in the acid content of the kidney from which it came, and vice versa. We have now to say exactly what we mean by urinary " acidity." In the days before physical chemistry, acidity and degree of acidity was usually measured by titrating with an alkaU of known strength. This gives the so-called titration acidity of a hquid, and this type of analysis has been appHed to the urine, as to many other fluids derived from the body. The titration acidity of the urine in nephritis has by scores of inves- tigators been found to run much above that of normal urine and we may still use it to advantage in studying our clinical cases to-day. It must be clearly understood, however, that the titration acidity of the urine is not an absolute guide to the degree of acid intoxication occurring in a kidney. The reasons for this are, of course, obvious. When, for example, we compare the poisonous effects upon the tissues of equivalent concentrations of phosphoric acid, ammonium dihy- drogen phosphate, diammonium hydrogen phosphate, and tri- ammonium phosphate, the first is found to be highly poisonous, the second more mildly so, and the third and fourth still less poisonous, in the order named. Yet the titration acidity of all four as ordinarily determined by titrating with standard sodium hydroxid solution gives the same reading. These facts must be kept in mind when judging the clinical significance of the titration acidity of the urine. For, clearly, were a urine filled with pure, highly poisonous phosphoric acid instead of the com- paratively innocuous dibasic or tribasic salt, the titration acidity would not betray this fact. Determination of the titration acidity has nevertheless a distinct value if what has been said be kept in mind. It is capable of giving us definite evidence of the existence of an abnormally high acid content in the urine (and therefore in the kidney) and of the changes in this from hour to hour or day to day. Titration acidity does not, how- ever, vary directly as the degree of intoxication, and only igno- NEPHRITIS 631 ranee of the elementary facts of chemistry would ever lead anyone to expect such complete parallelism. The physical chemists have more recently distinguished between the latent and the active acidities of a fluid, meaning by the first the total replaceable hydrogen, by the second the hydrogen ions yielded upon solution in water. When the physical chemists speak of acidity they usually refer to the active or hydrogen ion acidity, and a large portion of them believe that all so-called acid effects are dependent exclusively upon the presence and the number of these hydrogen ions. When the acid content of a system is increased there follows usually an increase in the hydrogen ion acidity. Increasing the amount of acid in a beaker of water is followed by an increase in the number of hydrogen ions, and when we deal with very dilute solutions the increase in hydrogen ion acidity is very nearly proportional to the increase in the amount of acid. It is for this reason that the hydrogen ion acidity of urine coming from a kid- ney containing a more than usual amount of acid, as in nephritis, is usually increased. The increased acid content of the kidney, to satisfy the laws of chemical equilibrium, demands an increased acid content in the urine coming from it, and as an expression of this we find an increased hydrogen ion acidity. The hydrogen ion acidity of the urine can be measured in various ways, and since some of these can be used clinically it constitutes one figure of value in judging a kidney case. It must,, however, he clearly understood from the outset that hydrogen ion acidity determinations of the urine can alone he no ahsolute index to the severity of the intoxication occurring in the kid- ney, nor yet that every increase or decrease in hydrogen ion acidity is or must he followed hy a corresponding increase or decrease in the severity of the intoxication in the kidney. The reasons for this are, of course, perfectly obvious. The attempt was made some fifteen years ago to show that the toxicity of various acids, as determined by their effects upon growing plants, the sense of taste, the absorption of water by muscle, the aggregation of infusoria, etc., followed their degree of ionic dissociation. It was early learned, however, that no such parallelism exists. Thus, it was found that acetic and other organic acids with their low dissociation produced physiologically as great effects as the highly dissociated hydrochloric, nitric, and other acids. On the 632 (EDEMA AND NEPHRITIS other hand, the rather highly dissociated sulphuric acid produced physiological effects far below the weakly dissociated organic acids. In other words, physiological effect is not determined solely or even in the main by the degree of dissociation. I believe I was the first to show that an entirely similar disproportion between degree of ionic dissociation and effect produced holds for the swelling of various protein colloids, and in so doing emphasized that the observed physiological reactions depend, in the main, upon the protein constituents of the tissues under consideration. It would, therefore, have been manifestly absurd for me ever to have claimed that any physiological effect could be measured by merely determining quantitatively the hydrogen ion acidity, and this whether we deal with purely physiological reactions or with the question of the development of the signs of a nephritis in a kidney. An increase in the hydrogen ion acidity of the urine above a normal standard may serve as evidence of an abnormal acid content in the kidney itself, but it can never be a complete measure of the degree of the intoxication. To make the matter more concrete wo need but illustrate this by saying that on poisoning a kidney with hydrochloric acid there occurs a great increase in the hydrogen ion acidity of the urine, yet if we produce a similarly great intoxication by the use of lactic acid only a slight rise is observed; on the other hand, intox- ication with sulphuric acid again gives us a great rise in hydro- gen ion acidity, and yet comparatively little effect on the kidney. To these considerations needs to be added the further fact, which I have so often emphasized, that an increase in the acid content in such an organ as the kidney does not alone determine the degree of effect produced. The presence and kind of salts found in a protein influence markedly its swelling and solution, and mere measurement of the hydrog"i ions in the urine tells us nothing of these factors. As a matter of fact, the addition of salt (even of neutral salt) to an acid protein mixture brings about an actual rise in hydrogen ion acidity of the liquid about the protein as this shrinks. The same fact can at times be observed clinically when a temporary rise in the hydrogen ion acidity of the urine follows' the active administration of salt alone. Unless such simple principles of physical and colloid chemistry be borne in mind we shall never come to a correct NEPHRITIS 633 understanding of the value and limitations of such hydrogen ion determinations.! ' By ignoring what I have written here and in my previous books and papers L. J. Henderson and his co-workers, W. W. Palmer and L. H. New- burgh, have wasted energy in attempts to disprove what I have never said. My constantly reiterated claim that certain changes in tissues are due to an " increased acid content " cannot at wiU be made to read by these authors an " increased (hydrogen ion) acidity." The latter may under otherwise constant conditions become evidence of the former, but the reverse need not foUow. Thus, the generally higher hydrogen ion acidity of a tissue or of a fluid coming from that tissue is evidence of an increased acid content in the tissue, but considerable quantities of either acid (or alkaH) may be introduced into any of our tissues, thereby raising their " acid content " even to the point of kilUng them, without any appreciable increase in the hydrogen ion acidity. When we add acid to a protein in the presence of an indicator the " acid content " of the protein rises from the moment we begin adding acid, but a long time may elapse and much may be added before the hydrogen ion acidity changes, as evidenced by a change in the color of the indicator. Would Henderson and his school hold that because the indicator had not turned no acid had yet been added? Beyond this the findings of Henderson and his co-workers do nothing but corroborate my own teachings. While in his first communication Henderson thought his results to dispose of my views entirely, his more recent conclusions are less sweeping. Thus, he finds " The mean acidity in cardio-renal cases is undoubtedly high . . . which other observations lead us to consider a form of acidosis" (Jour. Biol. Chem., 13, 404 (1913)). " This lack of response to alkali occurred most frequently in patients with kidney disease" (Arch. Int. Med., 12, 163 (1913)). "The hydrogen ion concentration from individuals with severe cardiac decompensation is higher than normal " and " the hydrogen ion acidity follows the general clinical course, becoming normal when compensation is restored" (Arch. Int. Med., 12, 146 (1913)). These conclusions sound strangely like my own. In the former of these articles Henderson advises against unchecked use of alkali in treatment because in too high concentrations it leads to albuminuria. This is a fact I emphasized long ago, but why is Henderson so willing to concede the albuminuria to be due to excessive alkali when he strains so over the analogous behavior of acid? Of course, Newburgh, Palmer and Henderson found that " it could not be shown that there was any definite relation between hydrogen ion concentration of the urine and cedema in the oases studied" (Arch. Int. Med., 12, 146 (1913)). But then, as pointed out above, we learned in the late nineties that physiological effect is nowhere proportional to the hydrogen ion concentration. So far as the general biological views of Henderson are concerned he errs in these, as do most of the physical chemists working in biology. While I doubt not that the maintenance of neutrality in the organism is in good measure dependent on a play between monobasic and dibasic salts, some caution is necessary before all the conclusions of the physical chemists obtained on dilute aqueous solutions are bodily heaped upon protoplasm. Protoplasm is not a sack of water with a few salts dissolved in it. The water of pro- toplasm is hydration water and the dissolved substances are by no means all in simple solution. Not until it is shown that reactions occurring in gels are identical with those occurring in pure water — which we already 634 (EDEMA AND NEPHRITIS Of the methods that have been or may be used to measure the hydrogen ion acidity of the urine (or any other body fluid) nearly all are too complicated for routine clinical use. Those of S. P. L. SoRENSON and of L. J. Henderson are among the simplest, but even these require more time and skill than is always available. A series of phosphate or acetate mixtures having a known hydro- gen ion acidity are first prepared to which indicators (dyes) are then added. The urines properly diluted and containing the same indicators are then matched against the colors in the stand- ard mixtures. The method has yielded valuable figures, but it is too compUcated to find universal employment, and the degree of accuracy attained is, after all, not necessary for good judgment in medical practice. To get a method which would yield for clinical purposes suf- ficiently accurate data and still be simple enough to be employed by anyone, I have made use of graded indicators such as the know is not the case, for, as Findlay and Creighton (Biochem. Jour., 5, 294 (1911)) have shown, so simple a phenomenon as the solubility of oxygen in serum is only one-fifth as great as that in water — can we continue to apply without modification to protoplasm the physico-chemical laws governing reactions in dilute watery solutions. The latest criticisms of Henderson, Palmer and Nbwburgh (Jour. Pharm. and Exp. Therap., 5, 449 (1914) ), upon which the Journal of the American Medical Association lays great stress editorially (Jour. Am. Med. Assoc, 62, 2033 (1914), are equally inconclusive. Must it be reiterated that all salts, and particularly the phosphates and acetates used by Henderson, Palmer and Newburgh as they were used by Max Koppel (Deut. Arch. f. klin. Med., 112, 594 (1913) ), even earlier — decrease and may suppress com- pletely the swelling of proteins no matter what hydrogen ion concentration surrounds them? Since gelatin and fibrin show increased sweUing even in the lowest concentrations of carbonic acid it is not true that " no influence to increase colloid swelling has ever been observed through the action of hydrogen ions varying within the ranges of acidity known to occur in the body or in the urine." Further, it has never been maintained that labora- tory gelatin or fibrin was at once to be made identical with the proteins in our body. The body proteins are far more sensitive to acids, alkalies and salts than the long-suffering and mutilated materials we have salted and boiled out of them. The variations in the " acidity " of the blood noted by these authors are incorrectly maintained by them to be of no significance, for the fact remains that such as accompany mere change from arterial to venous blood already make the corpuscles hold 15 per cent more water, and if there is a passive congestion, 30. The clinician bases a diagnosis of oedema on a much smaller increase in general body weight than this. The arguments on the significance of " osmotic pressure " in water absorption by protoplasm may be referred back to the biologists who gave up the kind restated by Henderson, Palmer and Newburgh ten and twenty years ago. NEPHKITIS 635 physical chemists use.^ By using a number of dyes which show color changes at definite hydrogen ion concentrations and then using the same indicators on the urine it is possible to determine its hydrogen ion acidity. The indicators are so chosen that their turning points vary from each other approximately by the power of ten. Of the many indicators which might be used those are best which do not give colloid precipitates when added to urine. The following series, the end points of which are sharp and can be readily recognized even in highly colored urine, have given excellent results in my hands. Name of indicator and method of preparing same. Methyl orange (0.5 gram in 100 cc. distilled water) Paranitroph^nol (2 grams in 100 cc. alcohol) Sive's red ^^ (2 grams in 100 cc. water) Methyl red (0.2 gram in 100 cc. alcohol) Rosolic acid (0.5 gram, in 50 cc. alcohol +50 cc. water) Phenolphthalein (1 gram in 100 cc. alcohol) Thymolphthalein (0.5 gram in 100 cc. alcohol) Concentration of hydrogen ions when indicator changes color. 10-* lO-o 10-6 to • io-« 10-' Color of indicator. In acid solution. Salmon pink Colorless Red Magenta red Orange-yellow Colorless Colorless In alkaline solution. Orange-yellow Greenish-yellow Canary yellow Canary yellow Magenta Bluish-red Blue ^ This is the hydrochlorid of paramonomethylaminoazobenzencorthocarbonic acid. It does not turn until a hydrogen ion acidity more than that necesary to turn methyl red is attained, and yet shows an acid reaction before such is discoverable with paranitro- phenol. Under the direction of Lauder W. Jones, B. Sive worked this out to meet the need for an indicator lying between these points. In practice 10 cc. of urine are placed in a clean vessel (pref- erably a porcelain dish, which if distilled water is not available is first rinsed in the urine to be tested) and two drops of one of the indicators is then added to it. By trying successive indicators one is finally found toward which the urine is neutral. The urine has then the hydrogen ion concentration represented by the turning point of that indicator. As the acidity of the ^ See Arthur A. Noyes: Jour. Am. Chem. Soc, 32, 815 (1910), where is given an excellent discussion of the whole question of measurement of hydrogen ion acidity. See also Eduard Salm: Zeitschr. f. physik. Chem., 57, 471 (1907). Fritz Glaser: Indikatoren, Wiesbaden (1901); S. P. L. Sorensen: Biochem. Zeitschr., 21, 131 (1909); L. J. Henderson: Biochem. Zeitschr., 24, 40 (1910). 636 (EDEMA AND NEPHRITIS urine runs up, it will, of course, show an acid reaction to the upper members of the list, and as it runs down, to the lower. The turning point of the commonly used litmus is about that of rosolic acid. Urine not acid to phenolphthalein is alkaline to Htmus, while thymolphthalein still remains colorless in urines which are distinctly alkaline to litmus. As is to be expected, the hydrogen ion acidity of the urine shows great variations even in health. A man doing muscular work, or on a predominantly meat diet shows a higher acidity than one ui bed or on a predominantly vegetable diet. The urine after meals is less acid than that before meals, and the night and early morning urines are more highly acid than those obtained after breakfast. The measurement of the hydrogen ion acidity of the urine is one of the few tests in which averages and twenty-four samples give us least information, and one less valuable than isolated tests at frequent intervals. The reasons for this are obvious. An athlete starting with a urine alkaline to methyl red secretes one highly acid to this shortly after going to work. But the urine returns to the originally alkaline state after a short rest. In the period of observation the urine originally free of albumin and casts becomes rich in these and loses them again. Had we measured only the average acidity as obtained by mixing the three samples of urine we should never have discovered the acid wave and perhaps maintained that the hydrogen ion acidity never went above the normal, as do some of my critics. The same is true of the alleged " physiological " and orthostatic albuminurias. At bed rest the urine shows a degree of hydrogen ion acidity which increases as the patient assumes the erect position (while albumin, casts, etc., appear at the same time) to fall again on resumption of the horizontal. Only many tests at frequent intervals will betray these constant changes. What we are interested in particularly, therefore, are the highest acidities registered and the length of time these remain active. Other things being equal, it is these two factors which determine how much effect is going to be produced on the colloids of the kidney. In practice, now, when will we say that our patient is not exceeding a safe hydrogen ion acidity of the urine? To get at this value I chose the highest hydrogen ion acidity registered NEPHRITIS 637 by healthy men on a full diet at bed rest. Such individuals do not show a hydrogen ion acidity sufficient to turn methyl red to the acid side except, perhaps, in the night urines voided between two and seven in the morning. The urine of healthy individuals who are up and about and on a full diet is also alkaline to methyl red for most of each twenty-four hours, though for obvious reasons, muscular exercise, high meat and fat diets, etc., may increase these hydrogen ion acidities. In actual 'practice, therefore, methyl red should he used as the routine indicator for all urines. Those which have an acidity for the major portion of each twenty-four hours, or always, above this point I consider abnormally acid. Figures below this point and down to the turning-point of litmus or phenol- phthalein I consider normal. Phenolphthalein rarely shows an alkaline reaction (if ammoniacal decomposition of the urine is not present) unless alkali is being fed to the patient. When the urine becomes alkaline to thymolphthalein too large quan- tities of alkali are being given, and the possibility of getting an albuminuria due to alkali is at hand. When methyl red is used in routine fashion on all patients it will be observed that a large number run constantly acid to this indicator. This serves to bring home how common are low- grade types of acid intoxication. The more acute and the protracted infections, starvation cases, diabetics, patients with cardiac and respiratory disease, and patients with generalized parenchymatous nephritis all show an abnormally high hydrogen ion acidity. In ambulatory patients with chronic interstitial nephritis secondary to vascular disease such an abnormal acidity may for obvious reasons be lacking, even though casts and albumin be present in the urine. In the later stages of the disease, especially when the circulation is beginning to fail, a high hydrogen ion acidity is the rule. When the acidity of the urine lies constantly below the turning-point of methyl red, or when by the administration of alkali it can be made to do so and be kept there, it augurs well for the patient. On the other hand, I cannot recall a single -patient in whom it was difficult or impossible to hold the urinary acidity below that of the turning point of methyl red who did not die. The correlation between increase in the hydrogen ion acidity of the urine and the appearance of albumin and casts in it can 638 OEDEMA AND NEPHRITIS be easily observed in athletes who voluntarily produce much acid, as well as in patients with orthostatic albuminuria, or in heart cases showing the first evidences of insufficiency. After exercise or on assumption of the erect position the acidity mounts from somewhere balow the turning-point of methyl red to a place above, and if this is maintained for a little time casts and albumin are likely to appear. The more definitely neutral the urine before such added efforts, the longer does it take for the casts and albumin to appear. If the facts here outlined are borne in mind the simple methods of measuring the hydrogen ion acidity described here prove of much clinical use. They apprise us of the existence of low degrees of acid intoxication in patients in whom we do not ordinarily consider them. By recognizing and meeting them by dietary regulations and alkali, we increase the reserve of these patients against the effects of such further intoxication as may be due to infection, anesthesia, or the trauma of opera- tion. Or, in the established case, a fall in the hydrogen ion acidity of the urine tells us that our therapy so far as alkalinizing the patient is concerned is of a successful type. Since the indicator method is exceedingly simple we can follow the patient's condition from hour to hour, an important fact when we deal with the acuter manifestations of nephritis and allied conditions. In cases of complete suppression, in other words, when there is no urine to tell us when we have succeeded in getting an adequate amount of alkali into our patient the reaction of the saliva serves as a useful guide. Ordinarily this is neutral to litmus paper, but it turns acid in various intoxications. Alkali should be given until it again turns neutral or even slightly alkaline to this indicator.^ 13. Ammonia Determinations In addition to measuring the titration or hydrogen ion acidity of the urine there is available another scheme of analysis which gives light regarding the existence and degree of an acid intoxication within the body. As is well known, the living animal ^ The ordinary litmus paper is well-nigh worthless. It should always be tested for its sensitiveness before dependence is placed upon it. Only Kahl- baum's make of neutral litmus paper has proved of service in my hands. NEPHRITIS 639 has to labor constantly to keep as low as possible and to render as innocuous as possible the acids steadily produced in its normal or abnormal metabolism, or introduced into it from without. Thus, whenever possible it converts the products of its metabolism into carbonic acid, which can be lost from the body through respiration without simultaneous loss of an equivalent amount of alkali. The non-volatile acids such as sulphuric or phosphoric in escaping through the urine take alkali with them. In many metabolic disturbances there is an interference with the normal production of carbonic acid, and then other, largely non-volatile acids, are produced in its place. These can be eliminated only in the form of salts produced through union with the bases of the body. When such abnormal production or accumulation of acid is not too intense, the reserve of bases in the body is usually sufficient to meet the need and the whole derangement comes and goes without disastrous effects upon the organism. But let the process be continued and this is not the case. In the herbivora the animal survives acid intoxica- tion until its fixed alkali (the metallic bases) has been exhausted down to the physiological minimum. But in the carnivora (with which man is to be reckoned) a second source of base is available which, if it exists in the herbivora, is negligible in amount. When the fixed bases of the carnivora are heavily drawn upon these animals begin to produce ammonia and to use it to neutralize the acid. The carnivora can therefore withstand a continuous acid intoxication longer than can the herbivora. In these simple considerations are embodied certain principles of analysis which have not yet found their deserved place in the practical handling of the nephritic. While they have long been used to advantage in diabetes and other metabolic disturbances they have been largely discarded in nephritis. The error of so doing will become apparent as we proceed. Clearly, when an individual for any of a number of reasons becomes the subject of an intoxication with acid this may betray itself in the fluids coming from his body, as in the urine (feces, sweat, or saliva), not only by an increased titration or hydrogen ion acidity, but in either or both of the following two ways : (a) By the appearance in it of acids which do not normally occur there, or 640 (EDEMA AND NEPHRITIS (6) By a relative or absolute increase in the amount of ammonia given off. Many of the now available tests for the presence of diacetic and betaoxybutyric acids in the urine as well as for the related acetone can be quickly performed and are so simple as to be - within the reach of the busiest practitioner.^ The recogni- tion of lactic acid is also easy. As none of these or similar substances are found in appreciable amounts in normal urine their mere qualitative recognition becomes evidence for an abnormal body chemistry. But it means no more than this. One is not immediately to conclude that an " acidosis " ^ exists, if by this term is meant an intoxication with acid. An acetone body, lactic acid or any other normal or abnormal acid may be produced in great quantities in the organism and appear in the urine, yet if sufficient base is available they are neutralized and so are practically without effect. Even good men seem to for- get this constantly. On the other hand, no acids of abnormal kind need be present in the urine (or elsewhere) and yet the organ- ism be suffering from an intense degree of acid intoxication (as with phosphoric or sulphuric acids) if the available supply of base has been exhausted. Since the normal amount of ammonia in the urine of an animal (like man) does not exceed a certain value, and since ammonia is not used to neutralize acid until the available fixed bases have first been heavily drawn upon, an absolute increase above the normal of the amount of ammonia in the urine becomes evidence of an acid intoxication. Furthermore, since the ammonia is formed at the expense of some of the other products of protein metabolism which escape in the urine, a relative increase in the amount of ammonia as compared with the total urea (or total nitrogen) of the urine is observed ' We cannot detail here the various methods that may be used and their relative merits. An excellent small guide thereto is S. W. Cole's Practical Physiological Chemistry, Cambridge (1913). ^ The word " acidosis " has had its meaning twisted so greatly to suit the whims of different authors that it might well disappear from our medi- cal and physiological writings. If it is used at all it must be used in its original sense as synonymous with " intoxication with acid." The mere finding of various abnormal acids or " acidosis compounds " in the urine or elsewhere does not yet mean an acid intoxication. The acetone bodies of a diabetic do not betray an acid intoxication, but an altered chemistry from which an acid intoxication may result. NEPHRITIS ■ 641 whenever an acid intoxication is under way, and becomes evidence for such. As comparatively simple methods are now available for the quantitative estimation of ammonia and of urea (or total nitrogen) in the urine they are sure to' find renewed application to many patients in whom an acid intoxication followed by nephritis is feared or existent. The considerations which led the older observers to discard such analyses as valueless depended upon the wrong interpretations which they made of their findings. As is well known, routine examination of the urine of all patients reveals the fact that acetone, diacetic acid, betaoxy- butyric acid, etc., occur in a large number of different clinical entities. Not only are they found in diabetes, but they occur in all types of general starvation, carbohydrate starvation, in many of the fevers (frequently in consequence of the bad feeding incident thereto), in cyclic vomiting, in certain of the so-called autointoxications, etc. Lactic acid, on the other hand, is found in uncompensated heart lesions, in severe anemias, in severe respiratory disease, in various intoxications, and in other conditions in which a lack of oxygen in some or all of the tissues of the body is in evidence. The finding of any or all of these abnormal constituents in the urine is evidence of an abnormal chemistry which may be followed by the signs and symptoms of a nephritis. Recognition of them in the urine is therefore a warning that the signs of a nephritis may supervene if care is not taken to feed properly (carbohydrate feeding) or to see that an adequate amount of alkali gets into the body to neutralize the acids. In the established nephritis the disappearance of these substances from the urine indicates that our therapy has been of an adequate and successful kind, a conclusion often brilliantly confirmed by the rapid disappearance of albumin and casts from the urine and of oedema from the body generally. I re-emphasize the necessity of keeping in mind that even large quantities of these abnormal substances may be found in the urine without any signs of nephritis, and conversely, that a severe nephritis may show but little of these substances or none at all. Thus, in diabetes large quantities of the acetone bodies may be found without the signs and symptoms of a nephritis, as long as an adequate amount of alkali is present to neutralize them. When the alkali becomes deficient the diabetic begins 642 (EDEMA AND NEPHRITIS to show casts and albumin in the urine, to retain water (develop an oedema) and to have his brain affected in this general process (brain oedema and coma). Some of my critics seem constantly to forget these facts, as also that 40 per cent of all diabetics, especially those in the latter stages of the severer types of the disease, show casts and albumin. What has been said of the diabetic (who really represents in essence but a severe grade of carbohydrate starvation) is true, of course, of all who suffer from carbohydrate starvation from any cause whatsoever, as in cyclic vomiting, in the protracted fevers, in the starvations associated with gastric ulcer, malignant tumors, tuberculosis, etc. To judge properly of the actual degree of acid intoxication from which such patients are suf- fering we cannot rely solely upon the discovery of abnormal urinary constituents. A better index, as we have already emphasized, is found in the measurement of the titration and hydrogen ion acidities of the urine. To this we may add the quantitative estimation of ammonia in the urine and the deter- mination of the ammonia coefficient. For oft-emphasized reasons the ambulatory case of chronic interstitial nephritis associated with vascular disease need show no abnormal total ammonia excretion or abnormal ammonia coefficient (total ammonia divided by total urea or total nitro- gen)^ The dying spots of (parenchymatous) nephritis in the kidney do not produce enough acid to affect the mixed urine coming from the whole kidney. Such a patient has large amounts of normal kidney substance left, and as long as his heart and circulatory apparatus is not too heavily taxed he does not develop acid in sufficient amounts to call into play this ultimate neutralization mechanism of ammonia production. In fact, such a patient inay die in coma (so-called " uremia ") and yet show no increased ammonia output. This coma is, to my mind, an oedema of the brain, but since it is usually secondary to vascular disease the acid production is largely local and there is no reason why the body as a whole should show evi- dence of a general acid intoxication, to betray itself through an abnormal ammonia output in the urine. If the coma persists and the patient is not fed, or if convulsions supervene, the 1 The figure obtained is practically the same, as normally, urea con- stitutes over ninety per cent of the total nitrogen. NEPHRITIS 643 ammonia output is regularly found to rise. Because of this rise in ammonia output in the course of the coma it has been reasoned that the original " uremia " caused it and that the " uremia " itself cannot have been an acid intoxication of the brain (oedema of the brain). If what has just been said be borne in mind, it becomes readily intelligible how in the begin- ning there was an oedema of the brain (due to local acid intoxica- tion) and how through this with its attendant bad feeding, etc., the patient's condition was made steadily worse. An increased total and relative ammonia output is, as we should expect, particularly common in the so-called generalized parenchymatous types of nephritis, especially when protracted as in those incident to pregnancy. As a rule, these cases show with increase in albumin, casts, etc., or with increase in the gen-- eralized oedema, or with increase in the severity of the symptoms premonitory of coma and convulsions, an increase in either the total or the relative ammonia output. Where normally the ammonia coefficient does not exceed 5 or 6 per cent of the total urea or nitrogen (or in absolute amounts, 0.3 to 0.6 gram a day) such cases will show 10 per cent of ammonia or even more (with an absolute output of 2, 3 or even 4 grams a day). These considerations suffice to indicate how important it is to follow the absolute and relative ammonia excretion in any case in which a nephritis is feared or existent. The value of such determinations in diabetes by way of foreseeing and forestalling coma has long been recognized. The coma incident to neph- ritis, like the coma of diabetes, is an oedema of the brain, and its appearance, particularly in the so-called generahzed' parenchy- matous types of nephritis, can be foretold with the same degree of assurance. A long series of observations on the absolute and relative amounts of ammonia found in the urine in cases of nephritis are available. What has been lacking is their proper interpreta- tion. Rudolph von Jaksch, finding the ammonia output high in a number of nephritics, declared the nephritides to represent a type of acid intoxication. It is a view with which I heartily concur in spite of the many criticisms of it that have appeared since von Jaksch's original- communication. 644 (EDEMA AND NEPHRITIS 14. Prophylactic Measures Against Nephritis It becomes apparent from what we have said above that the mere diagnosis of nephritis has little meaning. It is as com- plete as a diagnosis of "dropsy," "fever" or "headache." We found the nephritides to divide themselves into three groups — into a toxic type in which a poison from some source affects the kidney, an infectious type again essentially toxic in nature but harboring organisms in the kidney itself, and a vascular type in which destructive lesions are consequent upon involvement of the blood vessels of the kidney by vascular disease. The first of these necessitates search for the source of the toxic agent; ' the second for the source and type of the infection in the kidney; the third for foci of infection responsible for the primary changes in the blood vessels. The infections of the kidney and vascular disease of the kidney come to us, at present, largely ready made, and so in these patients we must content ourselves with stopping, if possible, the progress of their pathological states and counter- acting as well as we may their effects upon the kidney. We can do most from a prophylactic standpoint for the toxic nephri- tides. Very evidently, if alkali, salt, carbohydrate and water relieve the signs and symptoms of the established toxic nephritis (and of the toxic portion that may be present in the other nephri- tides) the same measures will largely suppress or prevent their development if used prophylactically. It is more spectacular, by the use of alkali and salt to save a woman in convulsions, say from a pregnancy intoxication, but it is better medicine to instruct her in the methods of preventing such a condition alto- gether. And what is true of a pregnancy is true of any other established or anticipated type of intoxication in which dis- turbances from the side of the kidney are feared. Our surgical and medical wards are filled with patients about to be operated upon or ill of various maladies in whom we can do much to lessen or shut out entirely such possibilities. An interesting and easily manageable group is furnished by the surgical patients. Much of their ante-operative preparation used to be and, in some places still is, of a type to guarantee from the outset a maximum of discomfort and even danger to the patient. Routine examination of the urine of patients about to be operated upon, even when not suffering from infections NEPHRITIS 645 or intoxications likely to be complicated by nephritis, shows that in the days before their operations they are usually pushed into a state in which post-operative complications are very likely to appear.^ Commonly, the urine shows even before operation a high hydrogen ion acidity (acid to methyl red or even para- nitrophenol) . Such patients make poor surgical risks, and should be given the benefit of preliminary treatment. The mental anxiety so common in the surgical case expresses itself physiologically by increased muscular tone, and this declares itself chemically by a great acid production which mirrors itself in the high acid findings in the urine. How much the intelligent reassurance of the surgeon must mean to such a patient is obvious. Patients are also thoughtlessly ordered upon a " light diet." In this way the acid products of a starvation diet are added to those already present from other sources. Unless there are specific reasons against it, a surgical patient should be fed to within a few hours {about six) of his operation, and since carbo- hydrate starvation is the commonest and earliest form, special attention should be paid to getting an adequate sugar-starch ration into him in any form he may desire (potatoes, mush, toast, bread, sugar, candy). In addition the patient should be fed alkali in some agreeable form. Any scheme for accomplishing this is good, but the use in large quantities of artificial or natural alkaline waters is perhaps simplest and best. If conveniently possible the alkali should be used for several days before the operation and up to the point where the patient has a persistently neutral or some- what alkaline urine. A patient that goes upon the operating table with highly acid urine, or with a high ammonia coefficient, or shows qualitative disturbances in his metabolism as evidenced by acetone, diacetic and betaoxybutric acid in the urine, is in this proportion a bad surgical risk. An apparently poor risk, free from such findings, is almost certain to withstand anesthetic, the necessary trauma of operative interference, etc., without difficulty. James J. Hogan, Hayward G. Thomas, and Gordon F. McKiM have long prepared their surgical patients in this manner, and find they do better than on the old expectant scheme. The 1 For a report of fatalities due to " acidosis " in surgical patients see W. B. Russ: Jour. Am. Med. Assoc, 61, 1618 (1913). 646 CEDEMA AND NEPHRITIS patients recover more rapidly from the effects of their anesthesia, they are without headache (absence of brain cedema) they vomit little or not at all (absence of oedema of the medulla), they urinate an hour or two after the operation (absence of kidney oedema and early presence of " free " water), and the urine is practically free from the acetone, diacetic acid, albumin and casts so common post-anesthetically. Moreover, the trauma- tized tissues at the seat of the operation swell less and are less painful (less oedema). McKiM, for example, found that in a series of eighty sur- gical operations on the prostate, treated on the old expectant plan, he encountered much post-operative nausea, vomiting and gaseous distention of the bowels and some evidences of shock. Two patients became maniacal after the operations. Of a second series of fifty cases, in which nothing in operative technic or general hospital care was changed except that the patients were fed to within a few hours of operation and were given sodium bicarbonate and magnesium oxid, with much water until the urine was persistently neutral to litmus, he writes: " There has not been a single case of post-operative vomiting, practically no gas distention, and not one case of shock. The excellent condition of patients so handled is not to be compared with that of such not previously alkalinized. They are in better spirits mentally, in less pain and heal more quickly. This sounds strange, but my last cases have healed more quickly by a week." The toxic effects of an anesthesia so far as the kidneys are concerned deserve consideration from several viewpoints. In the first place, no anesthetic, be this chloroform, ether or nitrous oxid, can produce its desired effects without interfering with the oxidation chemistry of the body. In fact, anesthesia depends largely, if not entirely, upon such an effect. We need not, however, add to this load by giving the patient too little oxgyen. One of the superior merits of nitrous oxid anesthesia resides not so much in the anesthetic itself as in the fact that oxygen accompanies it. Nor may we say offhand that the bad effects of an anesthesia are simply a function of its length and the quan- tity of anesthetic used. Alvin Powell ^ found in a carefully studied series of operation cases that patients in whom but little anesthesia was used (and in whom perfect muscular relax- ' Alvin Powell: Personal Communication (1913). NEPHRITIS 647 ation was not obtained) showed more casts, albumin, etc., in the urine than those anesthetized longer and more deeply. On the other hand, very deep anesthesia was again followed by more albumin and casts. These facts are to be interpreted as follows: The bad toxic effects avoided by use of but little anesthetic are more than counterbalanced by those incident to the great acid production consequent upon imperfect muscular relaxation. The imperfectly anesthetized subject responds with muscular con- traction to the irritation due to surgical trauma. A medium degree of anesthesia increases the toxic factor of the anesthetic but eliminates the acid factor due to muscular rigidity. In deep sleep a maximum of interference with normal oxidation chemistry is again assured by the anesthetic itself. Similar considerations hold in determining the value of morphin, atropin, scopalamin, and sunilarly acting drugs admin- istered before or after an operation. With a good anesthetist their administration before an operation is only a handicap. With a poor anesthetist, their bad effects may be offset by the assurance of better muscular relaxation, with elimination of acid production from this source. The use of narcotics after operations becomes a matter of balancing their bad effects in interfering with the oxidation chemistry of the body against the good effects incident to elimination of the great muscular reaction, etc., consequent upon pain. The value of local anes- thetic measures (cocain, novocain, etc.), as in nerve blocking and in tissue infiltrations, is similarly explained. Their use pre- vents pain impulses from reflexly expressing themselves in increased muscular tone, but it should be clearly borne in mind that their careless or immoderate use is not without bad effects.^ In the after treatment of these surgical patients, an early reestablishment of a carbohydrate-rich diet with fruit juices ' I cannot help but endorse here the excellent surgical methods urged by George W. Cbile in the protection of his patients against shock, and this in spite of the fact that I am not of a mind with him regarding its nature and cause. I cannot escape the conviction that the central nervous system changes which he describes (swelling of cells with changes in their staining properties) are not the causes but the consequences of shock and to be explained in the same way as the csdemas which characterize the swollen glandular and body tissues elsewhere in the organism. I shall return to this question in detail at another time. 648 OEDEMA AND NEPHRITIS and alkaline drinks does much to hasten convalescence. The reasons for this are self-evident. In much the same way that we have discussed the protection of a patient against the effects of an anesthesia intoxication we may also guard him against the consequences of other intoxica- tions. The effects of arsenic (salvarsan) offer a case in point. It may be only good fortune, but I think not, that I have never had a serious salvarsan complication and that none has occurred in the practice of my colleagues who before injection take the precaution of thoroughly alkalinizing their patients. The blindness (optic nerve oedema), headache (brain oedema), vomit- ing (medullary oedema), generalized cedema, decrease in urinary output with blood, casts and albumin (cedema of kidnej's), pain in various sensory nerves, etc., which have been so fre- quently described, and which, when sufficiently severe, may end in death are more easily interpreted as oedemas due to arsenic intoxication than as syphilitic manifestations fanned to fire by the arsenic injections. In consequence of syphilitic disease, tissues on the verge of cedema through connective tissue over- growth, defective blood supply, pressure, etc., are pushed over the line by the arsenic intoxication. Thorough alkalinization beforehand (and treatment with potassium iodid has much the same effect) tends to prevent the disastrous consequences of such superimposed oedema. And let me add that I have used the salvarsan in the very cases in which it is ordinarily forbidden, namely in tabes, optic nerve and brain lesions, arteriosclerosis, aneurysm and kidney disease (so-called chronic interstitial nephritis with high blood pressure) when I felt syphilis was a factor in the case. What has been said of surgical patients holds with equal force for many of the medical ones. Their routine examination reveals a great majority suffering from medium and at times high grade states of acid intoxication. The urine is our best guide in the matter, but the retention of water by the patient (as evidenced by weighing him), the slight general cedemas so frequently observed, the headache, the vomiting, the rapid pulse, the quickened respiration ^ and the appearance of albumin and 1 The patient suffering from an acid intoxication from any source whatso- ever cannot hold his breath as long as can a normal person. Yandbll Henderson (Jour. Am. Med. Assoc, 63, 318 (1914) ) with his customary NEPHRITIS 649 casts in the urine all further betray the fact. The sources for the acid production .are, of course, many. The various intox- ications incident to the infections lead to great acid production. To this are often added the acid effects of inadequate feeding. The food may be badly chosen or be insufficient in amount, or in his illness the patient may not take enough. If he becomes mentally excited or develops a convulsion, say in the course of an infectious disease, then the products of such excessive mus- cular work are added, and his precarious state is further aug- mented. It must be self-evident how much we can do both to prevent and to relieve these conditions. It is not difficult in beginning cases to feed enough alkali by mouth to keep the urine persistently neutral. If such proves inadequate, enemas of sodium bicarbon- ate (12 to 18 grams to the liter) or of sodium carbonate and salt may be used. The necessity of sufficient carbohydrate feeding has long been emphasized by different authors. When the oral route is not adequate, rectal injections of dextrose (glu- cose) do much good. But it should be remembered that only dextrose is easily absorbed, and that the common practice of giving starch, milk, milk-sugar or cane sugar enemas and like concoctions is valueless, for these higher carbohydrates are scarcely absorbed. Moreover, several hundred grams of carbo- hydrate are required per day — a fact which will suffice to em- phasize the complete inadequacy of the teaspoonful methods of feeding so commonly encountered in practice. When both mouth and rectum are inadequate, good, and at times, startlingly good results are obtained by giving chemically pure dextrose intravenously. For reasons previously empha- sized, this is best given very slowly in concentrated form (45 grams dextrose per 100 cc. of water). To the mind which never asks what is the nature of the processes that characterize disease and what is the purpose and the ultimately accomplishable in therapy, the attempts at analysis, and the suggestions for treatment outlined in these pages, can mean but little. What I have said has been variously commented upon. To Theodore C. Janeway my reports on keenness has indicated how this fact may be used as a simple and accurate guide to the degree of acid intoxication clinically. 650 (EDEMA AND NEPHRITIS the relief of patients, in whom objectively judging colleagues had felt only a fatal issue possible, " read like the cures of the nostrum venders." ^ To others the patients would have recov- ered anyway. Some believe the whole procedure valueless. Henderson, Palmer and Newburgh find it " harmful and productive of human suffering." ^ How proper use of alkali, salt and sugar can produce such strange results is incompre- hensible. Some hold what I have written as essentially true.^ In the dilemma I advise anew that the objective thinker in medicine reject my views and first treat his patient with nephri- tis and its alleged consequences by more approved methods. If he should feel that his patient is going to die, alkali, salt and sugar might be tried. If the patient dies, the expected will merely have happened. If he lives, it proves nothing, but it may en- courage repetition of the experiment. And this, I feel, is all that is necessary. 1 Theodore C. Janbwat: Unsigned review of first edition of my " Nephritis." Arch. Int. Med. 9, 637 (1912). '^ Henderson, Palmer and Newburgh: Jour. Pharm. and Exp. Therap., 6, 466 (1914). The opinion is shared by Lawrence Litchfield: Jour. Am. Med. Assoc, 63, 307 (1914). The value of Litchfield's evidence has been analyzed by Pattl G. Woolley : Jour. Am. Med. Assoc, 63, 596 (1914). Similar criticism by A. R. Moore: Univ. Calif. Pub. Physiol., 4, 111 (1912); Jour. Am. Med. Assoc, 59, 423 (1912) and 60, 345 (1913), I have answered myself. Fischer: Jour. Am. Med. Assoc, 59, 1429 (1912); 60, 348 (1913). * Arthur D. Dunn: Lancet-Clinic, 108, 8 (1912); Albert J. Bell: Am. Jour. Med. Sci., 144, 669 (1912); personal communication (1914); Magnus A. Tate: Bull. Acad. Med. Cincinnati, 1, No. 22 (1912); Lancet- Clinic (1912); Edgar G. Ballenger and Omar F. Elder: Jour. Am. Med. Assoc, 62, 197 (1914); Rufus Southworth: Lancet-Clinic, Sept. 6 (1914); Gordon F. McKiM, Personal Communication (1914); H. Lowenburg: Jour. Am. Med. Assoc. 63, 1906 (1914). PART SEVEN GLAUCOMA PART SEVEN GLAUCOMA ON THE NATURE AND CAUSE OF GLAUCOMA From a pathological standpoint, glaucoma represents simply one of the local cedemas. From a clinical point of view, all its signs and symptoms have since von Grabfb's teachings (1860) been correctly referred to the increased intraocular pressure induced through the abnormally large amount of water held by the eye. How does it come to do this? A glance at any of the standard works on ophthalmology ^ shows no dearth of attempts to answer the question, but experi- ments planned to support the views advanced by the various authors have been singularly unsuccessful. For the most part, when not simply referred to the occult properties of " living " matter, these explanations are identical with those given for CBdema anywhere else in the bod^^ They are familiarly mechan- ical in character in that an increased lymphatic or blood pressure is supposed to force an abnormally large amount of liquid into the tissues of the eye. Such increased pressures are generally held to be induced through interference with the outflow of lymph or of blood from the eye occasioned through obliteration of the " filtration angle," etc. The experiments detailed in a previous section of this volume,^ and instituted in order to ground experimentally the colloid- •See, for example, Ernst Fuchs, Augenheilkunde, Zwolfte Auflage, 523; Leipzig u. Wien, 1910; Priestley Smith, Glaucoma, London, 1891. 2 See page 127. 653 654 (EDEMA AND NEPHRITIS chemical conception of oedema, showed clearly that the most in- tense grades of glaucoma can be induced experimentally in an eye in the entire absence of any circulation. This fact coupled with the well-known observation that any experimental increase in the pres- sure of the liquids circulating through an eye is not followed by glaucoma arraigns all the explanations thereof which look to an increased pressure as in itself of essential importance in its causa- tion. Such considerations compel the conclusion that the cause of glaucoma resides in the tissues of the eye itself, and that it becomes glaucomatous not because fluid is pushed into it, but because through changes in it, it absorbs an increased amount. That the amount of such absorption is sufficient to explain the severest grades of glaucoma is clearly evidenced by the fact that through the mere presence of a little acid, a beef eye can be made to absorb enough water to rupture its enormously thick sclera. This is a grade of glaucoma that exceeds anything ever seen clinically. Our experiments further show that this increased absorption of water is dependent upon the colloids in the eye, for not only is it built up of a series of different colloids {sclera, cornea, lens, vitreous humor), but the same conditions which govern the absorption of water by protein colloids also govern the absorption of water by the eye. On the ground of these experiments we can, therefore, no longer insist that an eye becomes glaucomatous because water is forced into it. It does this because chemical changes occur within it which increase the capacity of the ocular colloids for hold- ing water so that these are enabled to absorb water from any avail- able source. In our experiments with enucleated eyes this source is the solution into which the eye has been dropped; in the body it is the liquids flowing about or through the eye.. The chemical changes iii the eye which clinically lead to glau- coma are the same as those which may give rise to an oedema anywhere else in the body. In a large number of cases, circulatory disturbances in the eye are unquestionably present which per- mit of an accumulation of carbonic acid and of such other acids as are a constant accompaniment of states of oxygen want. In the glaucomas due to infections or, in general, to the toxic agents capable of producing " degenerative " or inflammatory changes (in the strict pathological sense of the term) in the eye we have to look to the chemical alterations thus induced for the cause of the altered hydration capacity of the ocular colloids. Many GLAUCOMA 655 of these intoxications lead to an altered oxidation chemistry in the affected tissues and an accumulation of acids; but the direct action of chemical substances (like urea or certain amins) , which in their ability to increase the hydration capacity of protein colloids act like acids, must also be kept in mind. Under the influence of proteolytic ferments, proteins having a low hydra- tion capacity can also be converted into such as have a higher one. Ordinary gelatin can thus be converted into Beta-gelatin. Wolfgang Ostwald's studies show Beta-gelatin to be capable of greater swelling than the unchanged. It is therefore con- ceivable that in inflammation (whether in the eye or elsewhere) an increased hydration capacity of the involved tissues may result from " autolytic " changes occurring in them even when no abnormal storage or production of acids in the part occurs. Perhaps the best evidence of the correctness of this colloid- chemical conception of glaucoma is furnished by the following clinical observations. II ON THE RELIEF OF GLAUCOMA 1. Local Measures The experiments on the swelling of enucleated eyes famil- iarized us not only with ways and means by which an intense glaucoma can be induced in an eye, but showed how the develop- ment of such can be prevented, or, once established, be made to go down again. While under ordinary circumstances little is gained by simply reducing an oedema, there exist a number of clinical forms of it which are in themselves dangerous. Glau- coma is one of these, which through its existence for even a short time may permanently blind an eye. To be able to combat the oedema in such a case is, therefore, not a useless procedure. In the experiments on the swelUng of eyes we learned that the presence of any salt markedly decreases the amount of water that an eye wifl absorb in an acid solution. The question therefore arose whether the instillation of salt solutions into the eye might not be followed by relief in chnical cases of glau- coma. 656 (EDEMA AND NEPHRITIS Haywaed G. Thomas and I decided to test the matter.* The instillation of salt solutions was not, however, to be entered into hastily, for experiment had shown that while all salts reduce the amount that an eye will swell in an acid solution, a large number also increase its tendency to develop corneal opacities. There would be little gained, except so far as relief from certain subjective symptoms might be concerned, by guarding an eye from blindness through glaucoma while blind- ing it through the agency employed for its relief. There exist, however, a number of salts which inhibit markedly the swelling of eyes in acid solution and at the same time not only do not increase, but even decrease the tendency to the development of these corneal opacities. In other words, the use of these salts tends to prevent the development of even that well-known turbindess of the cornea which is so constant a sign in clinical cases of glaucoma, and which one never fails to get in the experiments on eyes that I ,have described.^ These salts are the citrate-, tartrate, sulphate and phosphate of sodium and potassium. After a number of preliminary tests sodium citrate was chosen as the salt best suited for clinical use. Only the chem- ically pure salt (made by Kahlbaum) should be used, in con- centrations varying from m/8 to m/6 solution. Expressed in percentage, the former is equivalent to a 4.05 per cent solu- tion, the latter to a 5.41 per cent solution of the ordinary crys- tallized sodium citrate. The m/8 solution has an " osmotic " pressure below that of the human tissue fluids, the m/6 one shghtly above. The injections are made with a fine-needled hypodermic under the conjunctiva in the usual manner adopted by ophthalmologists, and are preferably preceded by the use of cocain and adrenalin solutions. Enough of the sodium citrate is injected to distend gently the connective-tissue spaces (5 to 15 drops). Immediately following the injection the patient suffers some pain. While this is usually insignificant, it is fairly severe in certain cases. Alternate hot and cold com- presses laid over the eye may help to ease it. In any event it disappears in a few minutes. In the severer cases of glaucoma ^Hayward G. Thomas and Martin H. Fischer: Annals of Ophthal- mology, 19, 40 (1910); Hayward G. Thomas: Journal of Ophthahnology and Oto-Laryngology, July (1911). * See the succeeding page 664. GLAUCOMA 657 we use the stronger sodium citrate solution, in the milder ones or for subsequent treatment the m/8 is sufficient. This will, in fact, rapidly reduce the tension in even the severe cases of glaucoma. Later in the treatment a mixture of one part of the m/8 sodium citrate solution, with two to four parts of a " physiological " (0.9 per cent) sodium chlorid solution, is sufficient. It need hardly be emphasized that such citrate solu- tions must be sterile and that since bacteria readily decompose them, they need constantly to be freshly prepared. Our results maj^ be summed up as follows : Subconjunctival injections of m/8 to m/6 (4.05 to 5.41 per cent) solutions of the crystallized, chemically pure sodium citrate in clinical cases of glaucoma are harmless and always followed by a prom,pt fall in ocular tension. The fall may be appreciable within ten minutes after the injection and ultimately so great as to make the eye have a subnormal tension. The effect of such injections lasts from three to six days {or even more) and is accompanied by a relief of all the subjective symptoms of glaucoma (except, of course, any blindness due to permanent structural changes). 2. Systemic Measures As the ophthalmologists have long recognized, many factors lying outside of the eye play a r61e in the development of the glaucomatous attack. The problem is, as a matter of fact, analogous to the acute oedemas that may develop in any of the other organs of the body, as in the brain (uremia), the kidney (nephritis), the optic nerve (papillo-oedema) , or the liver (cloudy swelUng, liver necrosis). And as in such oedemas we err when we observe only the specifically involved tissues, so also in glaucoma. It is self-apparent that a hydrating agency such as an accumulation of acid in an eye leads to a swelling of the ocular colloids, no matter whether its origin is purely local (say the consequence of an arterio-sclerosis of the blood vessels of the eyeball), or whether to this local acid production is added the effect of acid produced elsewhere in the body and carried into the eye through the circulation. The conditions that lead to an abnormal production or accumulation of acid in the body as a whole are many, and 658 (EDEMA AND NEPHRITIS constitute a list that is familiar to every ophthalmologist under the heading of etiological factors concerned in the production of glaucoma. Starvation, an excessive protein diet, hard muscular and mental work, excessive consumption of sour wines, various intoxications (anesthetics, alcohol, arsenic), the infections, the severe anemias, generalized arterio-sclerosis, uncompensated heart lesions, exposure to cold, are all associated with an abnormal production or accumulation of acid in the body. Any of these may be the deciding factor that pushes an eye on the verge of glaucoma from a local condition, over the line, and so precipi- tates the glaucomatous attack. Local treatment alone, he this a subconjunctival injection of sodium citrate or one of the more popular iridectomies, sclerotomies, or trephinings, does not affect the contributions which are being made by the extraocular factors. To meet the situation we must treat the whole man. To begin with, it is clearly indicated, therefore, that we remove as many as possible of these extraocular conditions. But we are likely to find that some of them cannot be removed, or at least not with sufficient speed to make it count in our patient who finds himself in the midst of a glaucomatous attack. Under these circumstances we have only one other door open to us, and that is to combat their consequences. In practical terms this again means a neutralization of the abnormal acid content by giving alkali; the administration of salts to reduce the swelling of all the colloids in the body including those of the eye; an administration of carbohydrate if indicated; and after the glaucomatous attack has subsided, water, day and night, to help wash out through the kidneys the acid and other toxic bodies from the various tissues. How in actual practice this is accomplished may be illus- trated by the following history of a case of glaucoma which Hayward G. Thomas invited me to see. Case XXVIII.— Mr. F. C, aged seventy-two, goes to his oflace daily- He has for fifteen years had some albumin and casts in his urine- Unless his carbohydrates are consumed in moderation, he also has sugar- All his superficial arteries are easily palpable and tortuous, and his heart is hypertrophied. The second heart sound is accentuated. His blood pressure is constantly 190, and rises to 210 mm. of mercury. He has never had a generalized oedema. GLAUCOMA 659 On July 16, after a day of mental and physical fatigue, he developed pain in his left eye and left temple, noticed that his eye was " bloodshot," and that he could not see the outline of objects clearly. The condition continued through the night, the pain being so severe as to keep him awake. In the morning of July 17 his state had not improved, and his eyesight had fallen off still more. He tolerated his condition through- out this day, and through the succeeding night and day, by which time he declared himself completely bUnd in the affected eye. In the middle of the afternoon of July 18 he summoned Dr. Thomas, who found the eye hard (tension +3), pupil dilated to size 5, Morton scale, conjunctiva very much chemosed, cornea slightly steamy — a typical attack of the so-called " acute inflammatory glaucoma." Instilla- tions of eserin were at once begun, and the patient moved to the hospital. The instillations were entirely without effect. At 9 P.M. a slow injection of the following solution into the rectum was started : Sodium carbonate (monohydrated, NajjCOs-HjO).. 4.3 grams Sodium chlorid 14 . grams Distilled water, enough to make 1000 cc. The patient retained the solution well, and by midnight had taken up the whole Uter. The tension in the eye had fallen appreciably an hour after the injection was started, and at midnight was normal to the touch. At the same time the subjective symptoms of the patient improved, and he went to sleep. At 4 a.m., 500 ce. more of the above formula were injected and retained. At daybreak the patient was able to recognize gross objects, and through the day his vision became steadily clearer. He remained under observation in the hospital for two days longer. No new symptoms developed, and he was discharged with completely restored vision. In interpretation of the clinical history just detailed, which is characteristic of the larger number of glaucomas encoun- tered in patients beyond forty, I would say that vascular disease was primarily responsible for a diminished oxygen supply to the eye. For years such a change had led to no appreciable symptoms so far as the eye was concerned, but one day in con- sequence of unusual muscular and mental fatigue, aided possibly by an " acidosis " incident to his sugar intolerance, the acid accumulation from these sources added to that initially incident to the bad blood supply to the eye, sufficed to increase so materi- ally the hydration capacity of his ocular colloids that they swelled to the point of giving him easily recognized signs and symptoms — a frank oedema of the eyeball, a glaucomatous attack. But the reduction of this attack did not change his blood vessel disease, 660 CEDEMA AND NEPHRITIS and so it could safely be predicted that in consequence of another period of hard work or dietary indiscretions he would again get eye symptoms. As a matter of fact, Dr. Thomas reports that after two months of freedom from symptoms the patient tired of his restricted activities, and his alkalinized diet and had two more attacks of increased tension, though not of a severe type. The first of these was controlled by the same eserin solution which in the initial severe attack had been unable to reduce the tension. In the second of these milder attacks the eserin again proved unavail- ing, even though a contraction of the pupil resulted. An active administration by mouth of alkali and table salt with calomel and magnesium citrate was turned to, and while using these the tension returned to normal. When we have succeeded in relieving the frank glaucomatous attack in which we are likely first to see our patient it becomes our purpose to prevent further attacks. To do this we make use of the principles already enunciated, though it is not necessary, of course, to work so aggressively. We need again to recognize and avoid as far as possible those conditions which directly or indirectly threaten to increase the hydration capacity of the ocular colloids and to increase the margin of safety against such. This means a sane restriction of the physical and mental activities and a quieter insistence on a diet rich in alkalies, salts and water. Ill SOME COMMENTS Prompt as may be the relief of tension with its associated symptoms in glaucoma after subconjunctival sodium citrate injections or the use of alkahne hypertonic salt solutions by rectum, it must be clearly understood that neither of these con- stitute a " cure " for it. As a cure of glaucoma we could only consider a removal of the condition or conditions which are responsible for the development of the substances which increase the hydration capacity of the ocular colloids. If these are acids, the product of a circulatory disturbance or of an infection, then clearly, the real cure resides in a correction of the circula- tion to the eye, or in the removal of the infection. But even GLAUCOMA 661 toward such ends do these dehydrating methods help. In the progressive development of a glaucoma the swelUng of the col- loids tends to compress the blood vessels passing into and out of the eyeball. The natural tendency of a glaucoma is, therefore, to make itself worse. Writers on ophthalmology are in the habit of laying great stress on the obliteration of the filtra- tion angle. Obliteration of the filtration angle is frequently said to be the cause of glaucoma. It is a consequence, as evidenced by the fact that enucleated eyes rendered artificially glaucom- atous by being placed in acid solutions show the same progressive decrease in the depth of the anterior chamber that is noted in cUnical cases. The matter is easily explained through the unequal swelling of the different colloids of the eye, those posterior to the lens (sclera, choroid, vitreous) being capable of greater swelling than those anterior to it (cornea, aqueous). Through this unevenness in swelling the cilary body is crowded against the sclera — a process in which the blood vessels of the ciUary body become pinched. The resulting embarrassment in the circulation (with its lack of oxygen, accumulation of carbonic and other acids, etc.) is then added to whatever con- ditions are already active in producing the glaucoma. To reduce the swelling of the ocular colloids, even though but tem- porarily, is, therefore, to improve the circulation through the eye and in this way to contribute not inconsiderably toward the restitution of normal conditions within it. If the glaucoma is the consequence of some acute accident then its prompt relief may not only save the eye from blindness through pressure, but by helping toward the re-establishment of a normal circulation through the eye furnish the necessary conditions required for the repair of any pathological process. The principles underlying both the local therapy of glau- coma (subconjunctival injections of sodium citrate) and the general therapy, as touched on here, are of course the same, and so the fact will not prove strange that both produce a lower- ing of tension. I now urge and rely chiefly on th euse of the rectal injections, not alone because glaucoma is so often but a local ex- pression of a general state, but because, while various observers ^ have reported uniformly favorable results from the use of sub- ' See for example VAN DER Hobvb: Klin. Monatsbl. f. AugenheUk. N. F., 13, 602 (1912). 662 CEDEMA AND NEPHRITIS conjunctival injections of sodium citrate, others have objected to them, and some have even maintained that their use increased the tension.^ While I have myself never observed such a result, the proper use of alkali and salt by rectum produces so effective a dehydration of the body including the eye that I have in the past two years made it a rule to try first this simpler method for two or three hours before turning to subconjunctival injections. But whatever scheme is used let me here again insist that the various mixtures of alkali and salt and sugar and water which I have suggested and to which my name is often linked have not in themselves any special virtue. Virtue resides in working out a really effective scheme for dehydrating swollen colloids, and the attending physician or surgeon is at liberty to accom- plish this by any means which he thinks best. Some of my critics have utilized their failure to obtain a fall in tension after subconjuctival sodium citrate injections to com- bat the colloid-chemical theory of glaucoma. Without charg- ing them with improper preparation or improper use of their solutions — I find men constantly modifying the concentration or the amount of the injection to suit themselves — it might be well to inquire why they failed. When a solution of sodium citrate is injected subconjunctivally, we desire to have the sodium citrate diffuse into the eye and so decrease the hydration capacity of the ocular colloids (shrink them) . We use water along with the salt only of necessity. But if the glaucoma is of a severe type it means that the hydration capacity of the ocular colloids is exceedingly high. It may hap- pen in consequence that when we inject an aqueous solution subconjunctivally the water is absorbed before the salt gets in, in which case the swelling would actually be further augmented. The bad result is not due to the sodium citrate, but to the inabihty to get the salt into the eye in sufficient concentration. Or, if extraocular factors are playing a good part in the produc- tion of the glaucoma as is the case when there is a generalized acid accumulation resulting from a weakened heart, overwork, starvation, etc., this cannot, of course, be neutralized by the instillation of a few drops of some salt under the conjunctiva. Others of my critics have objected to the lack of tonometric readings in my reports. A desire always to substitute actual 'See Gilbebt: von Graefe's Arch. f. Ophth., 32, 438 (1912). GLAUCOMA 663 figures for the results of human judgment together with the easy availability of the Schiotz tonometer certainly tempts one to fill out this gap, and yet I have hesitated in this direction, and for the following reasons. First, the reduction of tension in an eye must be of so marked a character if it is to serve as the basis for a suggested therapy that it must be readily discernible even to the, perhaps, but slightly practiced touch of any physi- cian. I would think little of a suggested therapy for glaucoma which reduced tension so slightly that only a tonometer could recognize the change. Second, tonometric measurements cannot be made without instillations of cocain, holocain or similar substances into the eye, and manipulations of the eyeball which are not without effect. Such instillations and manipulations themselves lead to an increased cedema, and so tend to main- tain or augment whatever increased tension already exists in the eye. There is evidence of a marked tendency in the recent literature on the treatment of glaucoma to urge more strongly than for- merly the use of myotics and constitutional remedies for the relief of glaucoma. This has largely grown out of the fact that iridectomy all too often fails to give more than temporary relief. The problem is really the same as that encountered in brain cedenia, in nephritis, etc., which represent in the involved organs processes which if they affect the eye are called glaucoma. A decompression operation, a stripping of the capsule, etc., may bring clinical relief (by permitting a better circulation through the swollen parts), as do sclerotomy, trephining, and similar sur- gical procedures when applied to the eye. But statistics on the after effects of operations in glaucoma are no better than those following surgical interference in brain oedema, nephritis, etc. The reasons for these failures, as well as the explanation of the occasional brilliant result, are, of course, not far to seek. Behind an oedema of the eye lie the same possibilities which produce an oedema anywhere else in the body. Rarely are such confined to the eye alone. And the effects and relative merits of a surgical operation, or a myotic, or sodium citrate, or alkali and salt by rectum can be foretold here as well as when a capsule stripping, a vasodilator, a diuretic salt or alkali are used in a brain oedema or a nephritis. When the swelling of the eye is due to a temporarily acting poison all may yield brilliant 664 OEDEMA AND NEPHRITIS and permanent results, for when the tension has once been reduced the eye is saved, for the causes leading to the oedema have then also gone. But when blood vessel disease — by far the commonest cause in our older patients ^ — is responsible for the increased tension, it may again be reduced, but since this does not abolish the blood vessel disease, the tension is again liable to increase even if an iridectomy or a trephining has been done or alkali and salt has been properly used. Many ophthalmologists know now and all will shortly learn that a diagnosis of glaucoma is as complete as a diagnosis of " dropsy," and as modern medicine is not content with the latter it will not long remain so with the former. Glaucoma, except when it follows trauma or direct infection, is not a local disease but nearly always a local expression of systemic derangement. It will be blotted off the pages of ophthalmic literature not by more surgery but by a prophy- laxis and therapy which recognizes and treats vascular disease, infection and intoxication involving the eye. IV ON THE NATURE OF CORNEAL OPACITIES In clinical cases of glaucoma there is noted as one of its most constant signs more or less opacity of the cornea. In an entirely similar manner the cornea loses its transparency in the experi- mentally induced glaucomas already described. Since the essential change in the eye in glaucoma consists of an abnormal increase in the amount of water held by it, the view generally advanced by ophthalmologists that the observed opacities are due to the absorption of water by the cornea does not surprise us. Such an origin for the opacities observed here has been extended to include those found in the other transparent media of the eye. Especially has the lens been believed to owe its loss of transparency in many conditions to an imbibition of water. ' In 22 patients with glaucoma 19 showed high blood pressure and other frank signs of vascular disease. The remaining 3 were younger individuals who had suffered from "rheumatism" with metastatic, infectious involvement of the eyes themselves. The eyes of one of these had been operated upon 13 times without benefit. GLAUCOMA 665 Serious objections seem never to have been raised against such a view, and this in spite of the fact that clinical cases of absolute opacity of the cornea or the lens may exist without any evidence of an increased absorption of water, while, on the other hand, even severe cases of glaucoma may come and go without more than a mere haziness of the cornea. These paragraphs confine themselves to the question of the origin of corneal opacities, simply because these have been studied with greatest care. It seems, however, that what is here ■ said regarding the cornea holds also for the lens ^ and the other transparent media of the eye. The opacities referred to, it need hardly be said, include only such as are the consequence of chemical disturbances in the eye, and have nothing to do with such as are the result of leucocytic deposits, connective tissue scars, etc. Neither the presence of an increased or a decreased amount of fluid in the cornea is responsible for the appearance of an opacity. Such is produced whenever some of the colloid constituents of the cornea are precipitated, and depending upon whether the precipita- tion is only slight or very great, these opacities vary from being barely visible (steaminess of the cornea) to such as are intensely white (leukoma). The effect of different solutions on the transparency of the cornea was judged in two ways, first in regard to the rate at which they permitted the development of an opacity, and second, in regard to the intensity of the opacity. The outer limits of the former vary from a few minutes to several days, for the latter from a turbidness scarcely visible to the naked eye to a white- ness like that of boiled albumin. The italicized conclusion is based upon the following facts.^ (a) If an eye is simply allowed to dry, no opacity of the cornea develops. Mere loss of water, therefore, does not lead to its appearance. (6) If an eye is laid in distilled water it gains in weight. In this process of water absorption the cornea takes a prominent ^ For experimental details which may all be explained in the terms of colloid-chemistry on opacities of the lens and water absorption by it, see Phil. Botazzi and N. Scaunci: Arch. ital. Biol., 51, 96 (1908); Rend, della Accad. del Lincei, 27, 305, 445, and 566 (1908); ibid., 28, 225, 326 and 379 (1909). 2 See Martin H. Fischer: Pfluger's Arohiv. 127, 46 (1909). 666 (EDEMA AND NEPHEITIS part, yet no turbidness of this structure develops until quite late. Simple absorption of water, therefore, does not lead to an opacity. (c) The presence of any acid favors the development of an opacity, but the different acids are unequally powerful in this regard. Nitric acid induces a corneal opacity more quickly than an equinormal oxalic acid, and this more quickly than an equinormal hydrochloric acid. Still less powerful are sulphuric and acetic acids in the order named. Clearly, therefore, the order in which acids induce corneal opacities is entirely different from the order in which they make eyes swell. (d) We note a further discrepancy between the amount of water absorbed by an eye and the rate of development, or better, the intensity of a corneal opacity as soon as the effect of adding equimolar salt solutions of different kinds to any acid solution is compared. While every salt reduces the amount of water absorbed by an eye in an acid solution, some salts favor the development of an opacity while others distinctly inhibit it. The citrate, acetate, and sulphate, for example, inhibit the development of a corneal opacity, while the sulphocyanate, nitrate, bromid, and chlorid favor it. (e) The effect of any salt seems to be made up of the algebraic sum of its constituent radicals. When a series of salts having a common base are compared, the order of the acid radicals is always the same, and when a series of salts having a common acid are compared, the order of the basic radicals is always the same. These orders are indicated in the two following lists, in each of which the radical most effective in producing an opacity is given first, that most effective in inhibiting it last. Sulphocyanate, nitrate, bromid, chlorid, sulphate, acetate, citrate. Iron (ferric), copper (cupric), calcium, strontium, barium, magnesium, ammonium, sodium, Uthium (?). The order in which different salts or, as we had best say, their constituent radicals, affect the production of corneal opacities, is, therefore, an entirely different one from the order in which they influence the amount of water absorbed by the eye as a whole. The disproportion is illustrated in Fig. 160. In a is shown the thickness of the cornea of an eye which has lain in distilled water for thirty-six hours and is still perfectly GLAUCOMA 667 clear; in b that of an eye which has remained for the same length of time in n/110 hydrochloric acid. This eye burst six hours after being placed in the solution. The cornea is very thick, but only slightly opaque (ground-glass appearance), c was left for thirty-six hours in a similarly concentrated hydro- chloric acid solution, containing magnesium nitrate in addition (20 cc. n/10 HCl-l-200 cc. m/3 Mg(N03)2). Even though the steamy (yellowish) white wjiite nucleus Figure 160. cornea is not swelled — it is even thinner than normal — it has the intensely white color of boiled albumin. About the same condition of affairs is shown in d, which indicates the appearance of an eye thirty-six hours after being placed in n/110 hydro- chloric acid solution plus ferric chlorid (20 cc. n/10 HCl-|-200 cc. m/3 ferric chlorid). In spite of the great loss of water, the thin cornea is intensely white (and stained slightly yellow from the iron chlorid). (J) In the experiments on the swelling of eyes it was found that non-electrolytes in low concentration do not markedly 668 CEDEMA AND NEPHRITIS affect the swelling of eyes in an acid solution. Nevertheless, most non-electrolytes appreciably inhibit the de-velopment of corneal opacities. (g) All the above facts show clearly that no parallelism exists between the total amount of water absorbed by the cornea and the intensity or rapidity of the development of an opacity in it. The facts outhned are easily harmonized as follows: While the eye as a whole swells, due to an increased hydration capacity induced in some of its colloids, a second colloid {of the type of casein) is being precipitated {dehydrated). As wollen eyeball with opacities in its clear media {glaucoma) is the analogue of what in other organs is known as " cloudy swelling," ^ and the conditions which bring it about are exactly the same in both. V CLOSING REMARKS This volume must not be ended without the request that should its contents tempt any clinician to try the therapeutic methods here advocated, it tempt him also to study the considera- tions upon which they are based. It will prevent misunderstand- ing in criticism, and the disappointment incident to application of the suggested remedial measures to improperly chosen clinical cases. Our studies^ were made originally with an eye to analyzing in the terms of colloid chemistry a series of physiological and pathological phenomena which are associated with the problem of water absorption, and with no immediate ideas of applying clinically any of our deductions. We think that we have suc- ceeded in showing that the water of the body cells and fluids is carried as hydration water in combination with the hydro- philic, more especially the protein colloids found in them. We have extended this view to include oedema, which we have defined as a state in which the hydration capacity of the body colloids is abnormally increased. As causes of oedema we have cata- logued any substance or condition which is capable, under the '■ See page 455. ^ For references to them, see the bibUography at the end of this volume. GLAUCOMA 669 circumstances existing in the body, of increasing the hydration capacity of any of its hydrophilic colloids. We have mentioned that an abnormal production or accumulation of acids constitutes one of these conditions, but in spite of its dominant role we have never maintained this to be the only one. As we discovered that all salts, including the neutral salts, decrease the hydration capacity of certain proteins swelling in the presence of an acid, it is but natural that we should have insisted that use of such a fact could and should be made in combating the increased hydration which in the body we call oedema. And since such an oedema, as it involves special cells, special organs, or the body as a whole, goes by many names, it is only natural and logical that we should have proposed the same principles for the treatment of all of them. Upon such considerations, learned in the laboratory and upon animals where alone we can obtain strictly reproducible results, is based all that we have tried to formulate into some principles that should guide us in the treatment of those clinical conditions in which an oedema is a prominent feature, and inde- pendently of whether it involves the whole body or individual organs like the skin, mucous membranes, kidney, brain, liver, eye, or optic nerve. I do not believe that these fundamental propositions have been or can be validly attacked. Progress will be furthered by those who make the more positive contributions which tell us how chemically or physically conditions are brought about in any organism which alter the hydration capacity of its con- stituent colloids. In this problem, as always, science will be moved less by the cry of what is not, than by the whisper of what is. BIBLIOGEAPHY This is a list of the publications in which were originally expressed the views summed up in running form in the foregoing pages. 1. Maetin H. Fischer: The Physiology of Alimentation, New York (1907). 2. Maktin H. Fischer and Gertrude Moore: On the Swelling of Fibrin. American Journal of Physiology, 20, 330 (1907). 3. Martin H. Fischer: On the Analogy between the Absorption of Water by Fibrin and by Muscle. Pfliiger's Archiv, 124, 69 (1908). 4. Martin H. Fischer: Further Experiments on the Swelling of Fibrin. Pfliiger's Archiv, 125, 99 (1908). 5. Martin H. Fischer: The Nature and the Cause of (Edema. Jour- nal American Medical Association, 51, 830 (1908). 6. Martin H. Fischer: On the Swelling of Eyes and the Nature of Glaucoma (Preliminary Communication). Pfliiger's Archiv, 125, 396 (1908). 7. Martin H. Fischer: On the Swelling of Eyes and the Nature of Glaucoma (Second Communication). Pfliiger's Archiv, 127, 1 (1909). 8. Martin H. Fischer: On Corneal Opacities. Pfliiger's Archiv, 127, 46 (1909). 9. Martin H. Fischer: Remarks on a Colloid-Chemical Theory of Hemolysis. Kolloid-Zeitschrift, 5, 146 (1909). 10. Martin H. Fischer and Gertrude Moore: On the Antagonistic Action of Neutral Salts on the Swelling of Fibrin in Acids and Alkalies. Kolloid-Zeitschrift, 6, 197 (1909). 11. Martin H. Fischer and Gertrude Moore: On the Passive Con- gestion CEdemas of the Kidneys and the Liver. Kolloid-Zeitschrift, 5, 286 (1909). 12. Martin H. Fischer: CEdema as a Colloid-Chemical Problem, (with Remarks on the General Nature of Water Absorption in the Living Organism). Kolloidchemische Beihefte, 1, 93 (1910). 671 672 BIBLIOGBAPHY 13. Martin H. Fischeb: ffidema; a Study of the Physiology and the Pathology of Water Absorption by the Living Organism. New York (1910). (Also Available in German and Russian.) 14. Haywaed G. Thomas and Martin H. Fischer: On the Treatment of Glaucoma with Subconjunctival Injections of Sodium Citrate. Annals of Ophthalmology, 19, 40 (1910). 15. Martin H. Fischer: On the Nature of Cloudy Swelling. KoUoid-Zeitschrift, 8, 159 (1911). 16. Martin H. Fischer: Further Remarks on the Colloid-Chemical Analysis of Nephritis. KoUoid-Zeitschrift, 8, 201 (1911). 17. Martin H. Fischer: Contributions to a Colloid-Chemical Analysis of Absorption and Secretion. (Absorption from the Peritoneal Cavity). Kolloidchemische Beihefte, 2, 304 (1911). Reprinted in English in Cincinnati Lancet-Clinic, 107, 684 and 702 (1912). 18. Martin H. Fischer: Some Practical Points in the Treatment of Nephritis. Ohio State Medical Journal, 7, 400 (1911). 19. Martin H. Fischer: On the Nature, Cause, and Relief of Glau- coma. Trans. Amer. Acad. Ophth. Oto.-laryn., 193 (1911). 20. Martin H. Fischer: Nephritis; An Experimental and Critical Study of Its Nature, Cause and the Principles of Its Relief. New York (1912). (Also available in German and Russian.) 21. William H. Strietmann and Martin H. Fischer: On the Con- traction of Catgut and the Theory of Muscular Contraction. KoUoid-Zeitschrift, 10, 65, (1912). Reprinted in English in Cin- cinnati Lancet-Clinic, 108, 205 (1912). 22. Marian 0. Hooker and Martin H. Fischer: On the Absorption of Water by Nerve Tissue. Kolloid-Zeitschrift, 10, 283 (1912). 23. James J. Hogan and Martin H. Fischer: On the Theory and Practice of Perfusion. Kolloidchemische Beihefte, 3, 385 (1912). 24. Martin H. Fischer: A Response to Some Criticisms of the Colloid-Chemical Theory of Water Absorption by Protoplasm. Biochemical Bulletin, 1, 444 (1912). 25. Martin H. Fischer: A Further Response to Some Criticisms of the Colloid-Chemical Theory of Water Absorption by Protoplasm. Journal American Medical Association, 59, 1429 (1912). 26. Martin H. Fischer: Further Remarks on the Treatment of Neph- ritis. Trans. Association American Physicians, 27, 595 (1912). 27. Martin H. Fischer: Physical Chemistry in Pharmacology and Therapeutics. Section in F. Forchheimer's Therapeusis of Inter- nal Diseases, 1, 1, New York (1913). 28. Martin H. Fischer: A Third Response to Some Criticisms of the Colloid-Chemical Theory of Water Absorption by Protoplasm. Journal American Medical Association, 60, 348 (1913). BIBLIOGRAPHY 673 29. Martin H. Fischer: Further Remarks on the Treatment of Neph- ritis and Allied Conditions. KoUoidohemische Beihefte, 4, 343 (1913). 30. Martin H. Fischer: The Treatment of Nephritis and Allied Con- ditions. Journal American Medical Association, 60, 1682 (1913). 31. Martin H. Fischer and Anne Sykes: On the Colloid-Chemical Action of the Diuretic Salts. (Preliminary Communication), Science, 37, 845 (1913). 32. Martin H. Fischer and Anne Sykes: On the Colloid-Chemical Action of the Diuretic Salts. Kolloid-Zeitschrift, 31, 112 (1913). 33. Martin H. Fischer and Anne Sykes Non-Electrolytes and the Colloid-Chemical Theory of Water Absorption. Science, 37, 486 (1913). 34. Martin H. Fischer: On the Nature, Cause and Relief of Nephritis, Journal Medical Society, New Jersey, 11, 116 (1914). 35. Martin H. Fischer and Anne Sykes: On the Effect of Some Non- Electrolytes on the Swelline of Protein, Kolloid-Zeitschrift, 14, 215 (1914). 36. Martin H. Fischer and Anne Sykes : On the Colloid-Chemistry of Sugar Diuresis. Kolloid-Zeitschrift, 14, 223 (1914). 37. Martin H. Fischer: On the Relation between Chlorid Retention, (Edema and "Acidosis." Journal American Medical Association, 1914. {In press). 38. Martin H. Fischer and Anne Sykes: On the Non-Acid Hydration of Gelatin. KoUoid-Zeitschrift, (1915). (In press). AUTHOE INDEX Abderhalden, Emil, 540 Adams, L. P., 575 Alberran, J., 625 Aldbn, B. F., 346, 347, 348, 349 Allen, G. M., 592 Araki, Trasaburo, 193, 194, 197, 203, 230, 413, 418, 422, 423, 426, 426 Arnold, Julius, 470 Arnold, Rudolf, 137, 396 B Badt, 413 Baehr, Edmund M., 494, 561, 591, 692, 593, 618 Baetjer, W. a., 501, 510 Ballenger, E. G., 542, 650 Barcroft, 314, 413, 501, 508 Bauer, 267 Bauer, J., 137, 149 Batliss, 444 Bbchhold, H., 137, 506 Bell, Albert J., 650 Berghausen, Oscar, 371, 424, 529 Beutnbr, R., 394 Billings, Frank, 500, 621 Blatherwick, N. R., 540 Bock, C, 565 Bolduan, 170 BoRowiKow, G. A., 376 Botazzi, Phil., 665 Bowman, W., 316, 410, 505 Bright, Richard, 176 Brodie, T. G., 314, 507, 508 Brucke, E., 109 BUCHNER, 425 Bugarsky, S., 403 Bunge, G. von, 427, 640 Burton-Opitz, R., 490 Calvin, J. W., 102, 105, 106 Campbell, Elizabeth, 569, 596 Cannon, W. B., 601 Chiari, O., 602 Clark, W. A., 577, 582 Coffey, W. B., 349, 360 CoHNHEiM, Julius, 176, 177, 226, 234, 236, 243, 244, 245, 456, 470 CoHNHEiM, Otto, 255, 267, 279, 403 Cole, S. W., 640 CoNZEN, F., 688 Cotton, 413 Creighton, 634 Crile, George W., 647 CuRDTS, C. E., 585 Cushny, a. R., 267, 268 CzAPEK, F., 375 D Davenport, C. B., 372, 373 Davis, D. J., 621 Dernoschek, a., 332 Dieulafoy, G., 688 Dreseh, H., 406, 407, 408, 409, 410, 412, 516 Driesch, 372 Duclaux, 425 Dunham, H. K., 571 Dunn, A. D., 619, 650 Durig, a., 151, 156 675 676 AUTHOR INDEX E Edlbfsen, G., 419 Ehrlich, Paul, 170 ElCHBERG, J. H., 573 EijKMAN, C, 111, 176, 192 Elder, Frank R., 223 Elder, Omar F., 542, 650 Engelmann, T. W., 378, 386, 390 EtrsTis, Allan, 224 Evans, 413 EwALD, A., 193, 422, 610 Fairhall, L. T., 624 Farkas, G., 402 FiCK, Adolph, 279 FiHE, C. C, 581, 601 Findlay, 634 Fischer, Martin H., 43, 52, 55, 81, 112, 127, 137, 149, 224, 227, 255, 287, 295, 305, 313, 319, 330, 333, 364, 377, 391, 394,. 395, 424, 455, 629, 565, 609, 650, 656, 665, 671, 672, 673 Fletcher, W. M., 388, 418 forster, j., 427 Fowler, C. C., 624 Fraenkel, p., 402, 588 Fraser, 348 Frerichs, F. T., 448, 588 Freundlich, H., 42 Fhey, Ernst, 321 Friedlander, 236 Frolich, Theodor, 195 Ftjchs, Ernst, 653 FtJRTH, VON, 396 G Garniee, 289, 336 Gedroiz, K., 137, 376 Geier, Otto P., 572 Genth, 624 Geraghty, J. T., 627 Gettler, a. O., 540 GiES, W. J., 223, 242 Gilbert, 662 Glaesgen, 588 Glasbr, Fritz, 635 Goodridqe, F. G., 242 goppelsroeder, f., 326 Gottlieb, R., 322, 506 Graefe, von, 653 Graham, Evarts, 608 Graham, Thomas, 40, 240, 431, 433 Grober, J., 396 Grutzner, p., 409, 411 Gryns, 155, 192 GtJRBER, August, 111, 177, 271 Guthrie, C. C., 599 H Haake, B., 319 Halliburton, 466 Hamburger, H. J., Ill, 152, 155, 177, 192, 232, 233, 255, 258, 267, 271, 272, 273, 280, 357, 358, 359, 457, 458, 464 Hamilton, N. A., 578, 584 Hammarsten, 0., 466 Handovsky, Hans, 106, 107, 252, 332, 391, 469, 489 Hardy, W. B., 105, 106, 433, 489 Hart, E. B., 466 Hauch, 451 Hawk, P. B., 624 Hedin, 155 Hbidbnhain, R., 255, 267, 274, 279, 315, 316, 406, 409, 410, 411, 412, 431, 505, 507, 516 Heilner, E., 314 Henderson, L. J., 402, 633, 634, 635, 650 Henderson, Yandell, 351, 413, 648 Henle, J., 283 Hermann, L., 388, 391, 392 Herrmann, Max, 425, 529 Herter, E., 236 Hewlett, A. W., 620 HiROKAWA, Waichi, 232 HoBER, Rudolf, 151, 155, 161, 255, 267, 268, 367, 377, 402, 403, 404 HOEFFT, F. VON, 414 HOESSLIN, R. VON, 588 HOEVE, VAN DBR, 661 HOFF, VAN 't, 155 Hoffmann, F. A., 565 HoFMEisTER, F., 57, 59, 106, 110, 151, 267, 274, 280, 389, 390 HoGAN, James J., 287, 313, 333, 346, 347, 350, 500, 566, 570, 575, 601, 609, 611, 645, 672 Holmes, C. R., 593, 594 holst, axtel, 195 AUTHOR INDEX 677 Holt, O. P., 593 Hooker, Marian O., 137, 149, 672 Hopkins, F. G., 388, 418 Hoppe-Seylbr, F., 193, 194, 230, 418, 423, 425, 610 Huxley, T. H., 372 Irasawa, T., 423 Jaksch, R. von, 194, 197, 405, 423, 643 Janewat, Theodore C, 649, 650 Januschke, 602 JOLLES, A., 197 K Kahlbnberg, Louis, 375 Karell, 612 Kennedy, R. D., 424 KiELY, W. E., 572 KiLiANi, H., 425 Kiss, Julius, 366, 367 Kite, G. L., 244 Klemensiewicz, Rudolf, 242 Klose, H., 137 Koeppe, 155, 156, 366 kopetzky, s. j., 601 KoppEL, Max, 634 KoRNER, M., 242 KovBsi, G., 267 Krau^, F., 405 KuDER, W. S., 601, 611 Landsteiner, Karl, 457 Laqueur, E., 466 Laub, 413 Lenk, 396 Leubb, W., 419, 588 Leubuscher, 272 Lewis, 413 LiCHTHEIM, LUDWIG, 176, 243 LlEBERMANN, L., 403 Liesbgang, Raphael Ed., 137 Limbeck, C. von, 111, 177, 192, 271, 319 Litchfield, Lawrence, 599, 650 Loeb, Jacques, 110, 111, 116, 155, 156, 177, 178, 359, 374, 394 LoHSE, J. L., 601 Long, J. D., 601 LowiT, M., 234 Luciani, Luiqi, 195 LuDWiG, Carl, 274, 315, 316, 410, 505 LuNiN, N., 427, 540 M Magnus, R., 245, 246, 292, 319, 322, 505, 506 Majors, E. A., 580 Mandel, J. A., 466 Marchand, Felix, 242, 591 McDougall, William, 388, 389, 392, 393, 394 McKiM, Gordon F., 645, 646, 650 Meigs, Edward B., 387, 388, 389, 393, 394 Meisbnhbimer, 425 Mbloy, J. C, 197 Mbltzer, S. J., 233 Mendel, E., 193, 423 Mendel, L. B., 270 Mering, von, 270 Meyer, Hans, 166, 364, 366, 413, 511 Miller, Edgar G., Jr., 223 Miller, Joseph L., 590 Moore, A. R., 650 Moore, Gertrude, 43, 203, 227, 370, 671 Munk, J., 258 Munzbr, E., 413, 423 N Nasse, O., 109, 110 Nathansohn, a., 161 Nbf, J. U., 425 Nbilson, C. H., 339 Neumann, W., 624 Newburgh, L. H., 590, 591, -633, 634, 650 Noorden, C. von, 419 Noyes, a. a., 42, 43, 635 NussBAUM, M., 406, 409, 516 O Oliver, Wade W., 159 Oppbnheim, H., 197, 624 Orlow, 258 Osborne, W. A., 466 678 AUTHOR INDEX OSLER, 588 OsTWALD, Wolfgang, 42, 43, 57, 58, 59, 60, 63, 152, 201, 220, 240, 332, 476, 655 Overton, E., Ill, 151, 155, 156, 158, 159, 160, 161, 166, 177, 255, 359, 361, 362, 363, 364, 366, 394, 511 Palma, p., 413, 423 Palmer, W. W., 633, 634, 650 Pauli, Wolfgang, 57, 106, 107, 152, 252, 332, 359, 366, 391, 396, 466, 469, 489 Peiper, E., 405 Pblbt-Jolivet, L., 170 Pemsel, W., 403 Perrin, J., 42, 43 Pfbpfbr, W., 151, 154, 155, 158, 160, 359, 363 PiECK, C. G., 591 POKROWSKY, 236 PoNFiCK, E., 292, 505 Post, W. E., 543 POTZL, O., 137 Powell, Alvin, 646 Priestley, J. T., 492 Przibram, E., 137 R Ravine, William, 423 Rbcklingshausen, von, 470 Rbg^czy, E. von, 433 Rbid, E. Waymouth, 255, 267, 272, 278, 279 RhOREB, LtTDWIG VON, 403 Richards, D. N., 348 RiNDFLj;iscH, E., 456 Robertson, T. B., 368, 403, 466 Roger, H., 289, 336 Romberg, E., 588 RosENOw, Edward C, 500, 621 RosENSTEiN, 258, 588 ROWNTKBB, L. G., 627 RuLON, S. A., 624 Rumpf, W. H., 405 Russ, W. B., 645 Ryffbl, 413 Rysselberghe, VAN, 376, 377 S Sachs, 374 Sackur, O., 466 Sahli, H., 234 Saiki, 270 Salant, W., 543 Salm, Eduard, 635 Salus, G., 446 ScALiNCi, N., 665 Schade, 425 SCHBLTBMA, M. W., 588 ScHLOSs, Ernst, 540 Schmidt, C, 258 schmidter, w. c., 571 Schoep, Alfred, 506 Schorr, Karl, 106, 332, 391, 469 schrobdbr, p. von, 106, 489 schuller, a., 137 schutzenbekgeb, 425 SCHWARZ, O., 627 Sellards, a. W., 414 Senator, 588 Sherman, H. C, 540 Sjoquist, J., 403 Slyke, L. L. van, 466 Smith, Dudley, 576 Smith, Priestley, 653 Sollmann, Torald, 319 SoRENSON, S. p. L., 534, 635 southworth, rufus, 650 Spiro, K., 57, 61, 233, 319, 403 Starling, E. H., 255, 257, 322, 444 Stone, W. J., 619 Steassbtjrg, G., 193, 422, 610 Stkibtmann, William H., 377, 672 Sykbs, Anne, 52, 55, 81, 224, 295, 305, 313, 673 Tait, Dudley, 486 Tate, Magnus A., 650 Terray, p. von, 422 Thomas, Hayward G., 645, 656, 658, 660, 672 Thompson, G., 423 Tolman, Richard C, 332 Tract, Grover, 223 Traubb, Isador, 137, 200, 305, 507, 508 Traubb, Moritz, 154 True, Rodney, 375 AUTHOR INDEX 679 TsucHiTA, 420, 421 Tubby, 257 txtechteb, j. l., 572 U Underhill, F. p., 543 Upson, Fred W., 102, 105, 106 V ViRCHOw, R., 426, 455, 456 VoiGT, H., 137 VoiT, 267 Vhies, Hugo de, 154, 155, 160, 359 W Walker, C. A., 349, 350 Wallace, G. B., 267, 268 Webster, R. W., Ill, 155, 156 Weigert, 448 Weiske, H., 195, 543 Welch, William H., 233, 234 Wells, H. G., 543 Wherry, William B., 621 WiDAL, 612 WiNTERNITZ, M. C., 197 WiTTICH, W. VON, 280 Wolf, 413 Wood, T. B., 105, 106, 433 WooDYATT, R. T., 426, 621 WooLLEY, Paul G., 371, 469, 602, 650 Wright, A. E., 195, 601 Zangger, Heinrich, 170 ZiBGLBR, C, 242 Zillessen, Hermann, 193, 194, 197, 230, 422, 425 ZuNTZ, 427 SUBJECT INDEX Absorbing Membbane, physical chemistry of, 253. Absorption, of dissolved substances by protoplasm, 164; of blood from peritoneal cavity, 261, 609; of salt solutions from peritoneal cavity, 262; of non-electrolytes from peritoneal cavity, 264; of acid and alkali from peritoneal cavity, 265; in dead animals, 266; from gastro-intes- tinal tract, 267; theory of, 271; of dissolved substances from peri- toneum and alimentary tract, 271, 273; filtration theory of, 272; osmotic theory of, 273; physiological and vitalistic theories of, 273, 277; selec- tive, 275; of hypertonic, hypotonic and isotonic solutions, 275, 276; of blood from gastro-intestinal tract, 279. See also, Absorption op Water. Absorption of Water, by fibrin, 41, 53; by gelatin, 57; by gluten, 102; by muscle, 109, 359; by eye, 127; general problem of, 249; in complex organism, 249; from peritoneal cavity, 255; from gastro-intestinal tract, 268; by spermatozoa, 356; by epithelial cells, 356; by white blood corpuscles, 356. Acetone, and gelatin, 92; and nephritis, 425, 540. Acetone Bodies in nephritis, 540. Acid, and fibrin, 43, 435; salt antagonism to, 46; and gelatin, 58, 440; and muscle, 112; and eye, 128; and nervous tissue, 138; accumulation of, in cedema, 190, 191; and pulmonary oedema, 236; absorption of, from peritoneal cavity, 265; action of, on spermatozoa, epithelial cells and white blood corpuscles, 358; formation of, in plants, 377; and catgut, 379; r61e of, in muscle contraction, 392; accumulation of, in nephritis, 401; injection of, and nephritis following, 415; and cloudy swelling, 459; and secretion of dissolved substances, 512; and staining of coUoids, 514; in diet of nephritic, 539. Acid-forming elements in food, 540. Acid Fuchsin, and staining of kidney, 407. Acidity of normal and abnormal urine, 404; measurements of, in urine. 629. Acidosis, definition of, 640; in nephritis, 640. Adbouact, of colloid-chemical theory of absorption, 162. Adsorbent, 168. Adsorption, and distribution, 168; r61e of, in hemolysis, 368. Agar-agar, 269. Albumin Test, 108. 681 682 SUBJECT INDEX Albuminuria, test for, 108; of newborn, 426; general remarks on, 430; theories of, 431; solution theory of, 432, 443. See also. Nephritis. Albuminuric Retinitis, 496. Alcohols, and fibrin, 52; and gelatin, 82, 87, 88, 91; and muscle, 122, 362; and eye, 135; and nervous tissue, 146; absorption of, from peri- toneal cavity, 264; and urinary secretion, 317, 318; and nephritis, 425; in nephritic diet, 540. Alkali, and fibrin, 45; salt antagonism to, 46; and gelatin, 61; and muscle, 114; and eye, 130; and oedema, 224; absorption of, from peritoneal cavity, 265; action of, on spermatozoa, epithelial cells, and white blood corpuscles, 358; and production of nephritis, 428; and cloudy swelling, 462; in albuminuria of hard work, 535; in treatment of nephri- tis, 538, 541, 588; and rule for giving, 563; in heart disease, 590; in other conditions than nephritis, 601, 602; in high blood pressure, 619; in preparation of surgical patients, 645. Alkalinity, of blood in oedema, 194. Amins, and fibrin, 55; and oedema, 224. Ammonia, excretion of, in nephritis, 413; determination of, in nephritis, 638. Ampoules, 558. Amyl Nitrite, and nephritis, 425. Analogy, between protein and protoplasmic swelling, 109; between pro- tein and muscle swelling, 109; between soap and muscle, 110; between protein and eye swelling, 127; between protein and nervous tissue swelling, 136; between catgut and muscle contraction, 387. Anasarca, 195. Anemia, pernicious, and oedema, 194; and nephritis, 423. Anesthetics, and secretion, 317; and nephritis, 425, 566; protection against after-effects of, 646; local, 647. Angioneurotic CEdema, 601. Antagonism, between neutral salts and acid and alkali, 46, 63, 102, 107; between urea and sugars, 56, 98, 100; between pyridin and sugars, 57, 99, 101; history of discovery of, 357. Apparatus, for injection of sodium carbonate-sodium chlorid solution, 558, 561, 562. Arsenic, and oedema, 203; and nephritis, 425, 601: avoidance of after- effects from, 648. Arteriosclerosis, see Vascular Disease. Ascites, 608, 611. Ascitic Fluid, transfusion of, 346. Asphyxial nephritis, 424, 517; treatment of, 519, 529. Asthma, bronchial, 601. Atheroma, see Vascular Disease. Athletes, and nephritis, 419; relief of albuminuria in, 535. Atropin, and urinary secretion, 317; and lymph formation, 328. B Baehr Apparatus, 561. Base-forming elements in food, 540. Basham's Mixture, 548. Beer, 547. Betaimidazolylethylamin, 225. SUBJECT INDEX 683 Bibliography of colloid-chemical theory of water absorption, 671. Biology, importance of colloid hydration in, 151; and colloid-chemical theory of water absorption, 355. Blindness, 557. Blood, alkalinity of, in cedema, 194; physical chemistry of, 254, 283; compared with lymph, 258; composition of, 258; absorption of from peritoneal cavity, 261, 272, 609; changes of corpuscles in arterial and venous, 271; gastro-intestinal absorption of, 279; effect of, on urinary secretion, 292; why it remains in the blood vessels, 333; intravenous injection of, 335, 346; neutrahty of, 402; hydroxyl and hydrogen ions in, 403; decreased alkalinity of, in nephritis, 405. Blood Pressure, and theory of oedema, 228; decrease of, as cause of oedema, 228; and urinary secretion, 315; treatment of abnormally low, 345 increased, not due to kidney loss, 483; benefits of increased, 490 increase of, in general intoxications, 618; and bronchial oedema, 618 treatment of increased, with alkali, 619. Blood Serum, intravenous injection of, 337, 346. Brain (Edema, in sulphuric acid poisoning, 591; treatment of, 591, 601 different types of, 601; in nephritis, 618. Briqht's Disease, see Nephritis. Bronchial Asthma, 601. Bronchial Arteries, 235, 601. C Cachexia, oedema of, 244. Caffein, and urinary secretion, 318. Calomel, in nephritis, 542. Calcium Salts, in oedema, 540. Cane Sugar, see Saccharose. Cannula, 560. Carbon, 168. Carbon Monoxid, nephritis in poisoning by, 423. Carmin and fibrin, 368. Casein, intravenous injection of, 341, 343; physico-chemical behavior of, 466. Cases, of shock, 347; of nephritis, 566; of chronic interstitial nephritis, 573; of glaucoma, 658. Casts, origin of, 475; types of, 476; significance of, 615. Catgut, contraction of, 378; effect of acid on, 379; effect of time on, 380; residual contraction in, 381; effect of salts on, 382; effect of Ringer solution on, 384; analogy with muscle contraction, 387; as oroblem in colloid-chemistry, 387. Cause of oedema, 178, 190; of shock, 345; of nephritis, 400. Cellulose, 269. Chemical Differences, and distribution, 170. Chbyne-Stokes Respiration, 555. Chloroform, and nephritis, 425. Chlorosis, and ctdema, 194. Chromium, and nephritis, 425. Chronic Interstitial Nephritis, 450, 495; interpretation of signs asso- ciated with, 495; terminal manifestations of, 498; clinical histories of, 573; treatment of, 587, 589; prognosis in, 617. 684 SUBJECT INDEX Circulation, maintenance of, 333; interference with, and nephritis, 424, 517, 529. Circulatory Disturbances, and oedema, 191; and pulmonary cedema, 234. Citrus Fruits, 537. Clamping of renal vessels, 517, 529; effect of sodium chlorid on, 529. Classification of the nephritides, 448. Clinical Histories, of shock, 347; of nephritis, 566; of chronic interstitial nephritis, 573; of glaucoma, 658. Cloudy Swelling, in glands, 330; general problem of, 455; theories of, 455; of kidney, 459; effect of acid on, 459; effect of salts on, 461, 463; effect of alkali on, 462; microscopic description of, 463, 468; behavior of granules in, 463 ;_ as colloid phenomenon, 464, 465, 466; and casein, 466; interpretation of clouding and swelling, 469; glaucoma as example of, 668. CocAiN, and oedema, 203; and nephritis, 425. Coefpicibnt of distribution, 166. CoHEsivBNBSs of gluten, 105. Cold, and nephritis, 423. Colloid-chemical Theory, of water absorption, 162; of oedema, 190, 220; of muscle contraction, 387; of albuminuria, 443; of cloudy swelling, 464, 465, 466; of diapedesis, 472; of urinary secretion, 293, 305, 508; of glaucoma, 654. Colloids, definition of, 38; nomenclature of, 38; lyophilic and lyophobic, 40; emulsion and suspension, 40; hydrophilic and hydrophobic, 40, 41; hydration of, 41; hydration of, in liquid state, 106; biological importance of, in water absorption, 151; syneresis in, 240; absorption of, from peritoneal cavity, 261, 272, 609; effect of, on urinary secretion, 292; intravenous injection of, 335; transfusion of, 346; in growth, 373; staining of, by dyes, 514. Coma. 497, 557, 601, 642. Consequences, of kidney disease, 482; oedema as one of, of kidney disease, 491; treatment of, of nephritis, 586, 591. Constipation, 269. Contraction, of muscle, 377; of catgut, 378; residual, in catgut, 381; interpretation of, in catgut, 385. Convoluted Tubules, staining of, 410. Convulsions, 497, 557, 601; in pregnancy, 529, 598. Corneal Opacities, nature of, 664. Corpuscles, swelling of red, 367; loss of color by, 367. Correlation, of morphological and clinical manifestations in nephritis, 448. Criticism, of osmotic theory of water absorption, 153; of lipoid membrane theory of water absorption, 159; of colloid-chemical theory of oedema, 220; of theories of urinary secretion, 319; of theories of muscle con- traction, 389; of solution theory of albuminuria, 445; of case histories of nephritis, 566; of sodium chlorid restriction therapy, 607; of colloid- chemical theory of nephritis, 633; of colloid-chemical theory of glau- coma, 660. Curves, of water absorption. 111. Cyanids, and nephritis, 425. SUBJECT INDEX 685 D Dead, tissues, 233; animals, absorption in, 266; muscle, 394. Death, and oedema, 198, 202, 246. Decapsulation, of kidney, 597, 599. Delirium, 601. Dextrose, and fibrin, 52, 53; and gelatin, 86, 101; and eye, 135; absorp- tion of, from peritoneal cavity, 264; and diuresis, 307; use of, in nephritis, 583, 600. Diabetes, and diuresis, 313. Diapedesis, hemorrhage by, 470; theories of, 470; of white blood cor- puscles, 472; coUoid-chemical explanation of, 472; model of, 473. Diet, in nephritis, 539; fruit and vegetable, in nephritis, 541; milk, in nephritis, 547. Diffusion, 274. Digitalis, and urinary secretion, 318; in heart disease, 590. Dissolved Substances, absorption and secretion of, by protein, 164; peri- toneal and alimentary absorption of, 257, 271, 273; secretion of, by kidney, 324, 510; secretion of, secondary to water secretion, 512. Distilled Water, toxic effects of, 332, 360. Distribution inequahties, 165; laws, 166; coefficient, 166; and solubility, 166; and adsorption, 168; and chemical differences, 170; importance of, in urinary secretion, 511. Diuresis, and salts, 295; and sugars, 305; and diabetes, 313; after sweating, 515. Diuretic Salts, see Saline Diuretics. Diuretics, of second order, 314, 317. Dryness, of skin, 495. Dtds, 168; staining of cells by, 514; in tests of kidney efficiency, 627. Dyspnea, 601. E Eclampsia, see Pregnancy Intoxication. Edema, see (Edema. Editorials, in Journal of the American Medical Association, 446, 633. Efficiency Tests, general principles governing, 621; of kidney, 621. Egg-white, absorption of, from peritoneal cavity, 261. Emulsoids, 40. Epilepsy, and nephritis, 423. Epithelial Cells, 356. Equilibrium, 274; osmotic, 276. Equimolar Solutions, 47. Ether, and urinary secretion, 317, 318; and nephritis, 425. Etiology of vascular disease, 499. Excretion, see Secretion. Explanation, see Interpretation. Exudates, 610. Eye, absorption of water by, 127; and acid, 128; and alkali, 130; and salts, 130; and non-electrolytes, 135. 686 SUBJECT INDEX F Fallacy, of sodium chlorid restriction in cedema, 518. Fasting, see Starvation. Fatalities, in nephritis, 595. Fat-like, see Lipoid. Ferments, proteolytic, and cedema, 200, 203. Feveh, and oedema, 196. Fibrin, swelling of, 41; and carmin, 368; solution of, 434; acid and solu- tion of, 436; salts and solution of, 437; staining of, 514; sodium chlorid retention by, 605. Filtration, r61e of, in absorption, 262; theory of urinary secretion, 505. Filtration Angle, obliteration of, 661. Fischer's Solution, see Sodium Carbonate-Sodium Chlorid Solution. Flea-bites, 602. Focal Infections and vascular disease, 621. Free Water, 262, 289. Fruit, citrus, 637; in nephritic diet, 541. G Gangrene, 199. Gastro-intestinal Contents, physical chemistry of, 253. Gastro-intbstinal Tract, absorption from, 267; absorption of water from, 268; absorption of salt solutions from, 268; absorption of sugars from, 269; absorption of blood from, 279. Gelatin, swelling of, 57; intravenous injection of, 335; solutions for intra- venous injection, 346; solution of, 439; acid and solution of, 440; salts and solution of, 441 ; sodium chlorid retention by, 604. Gels, change of, to sols, 433. Glands, changes in, during rest and activity, 329; cloudy swelling in, 330. Glaucoma, vascular changes as cause of, 497; nature and cause of, 653; colloid-chemical theory of, 654; reUef of, 655; sodium citrate injections in, 656; systemic measures for relief of, 657; nature of corneal opacities in, 664; as illustration of cloudy swelling, 668. Glomeruli, staining of, 410. Gluten, swelling of, 102; cohesiveness of, 105; solution of, 105. Gout, 621; infectious nature of, 621. Gram-molecular Solutions, 48. Growth, passive and active, 372; energy for, 373; colloids in, 373; water in, 374; osmotic forces in, 374; curvatures, 374 H Hard Work, and nephritis, 418, 535. Hat-pevbr, 601. Heart, hypertrophy of, 484, 488; failure of, in nephritis, 496; failure of, in vascular disease, 690. Heart Disease, and nephritis, 422; treatment of, 590; digitalis in, 590; alkali and salt in, 690. Hemoglobinuria, paroxysmal, 371, 423, 424. Hemolysis, 364; and adsorption, 368; inhibition of, by salt, 528. SUBJECT INDEX 687 Hemoeehage, effects of, 345; by diapedesis, 470; from kidney, 470. Horse-serum, intravenous injection of, 335, 338, 346. Hydration, of fibrin, 41; of gelatin, 57; of gluten, 102; of liquid colloids, 106; of muscle, 109, 359; of eye, 127; of nervous tissue, 13(5; bio- logical significance of, 161. See also, Absorption of Water. Hydremia, 234, 243, 271. Hydrocele Fluid, transfusion of, 346. Hydrogen Ions, in blood, 403; in urine, 404; estimation of, in urine, 631. Hydrophilic Colloids, 40. 41. Hydrophobic Colloids, 40, 41. Hypertonic Solutions, absorption of, 275, 361. Hypertrophy, of heart, 484, 488. Hypotonic Solutions, absorption of, 276, 361. Hysteresis, 54. Idiopathic (Edema, 540. Imbibition, 279, 363. Inanition, and oedema, 194. Indicators, in urine examinations, 635. Indigocarmin, see Sodium Indigosulphonate. Infarct, 198. Infections, of kidney, 453. Injection, rectal, of sodium carbonate-sodium chlorid solution, 550; of sodium citrate in glaucoma, 656. See also. Intravenous Injection. Injury, brain oedema following, 635. Insects, stings of, and oedema, 199, 602. Interpretation of associated manifestations in nephritis, 494; of experi- ments on nephritis, 533; of sodium chlorid retention in nephritis, 603. Intestinal Contents, physical chemistry of, 253. Intestine, see Gastro-intbstinal Tract. Intravenous Injection, of salt solutions, 334; of blood, 335; of gelatin, 335, 346; of blood-serum, 338, 346; of casein, 341, 343; of sodium carbonate-sodium chlorid solution, 559; of dextrose, 600. ISOSMOTIC, 155. Isotonic, 155; solutions, and absorption of, 276, 361. ISOTONICITY, 155. Joint Affections, 601. K Kaolin, 168. Kidney, passive congestion of, 225; osmotic behavior of, 232; physical chemistry of, 282; secretion of water by, 286; work of, 314; conditions affecting output of water by, 314; transition from physiology to path- ology in, 322; secretion of dissolved substances by, 324; staining of, in nephritis, 406; secondarily contracted, 449; small red, 449; primarily contracted, 450; infections of, 453; small gray, 454; cloudy swelling in, 459; hemorrhage from, 470; loss of substance in, 492; decapsulation of, 597, 599; efficiency tests for, 621. Kidney Efficiency Tests, 621; general principles governing, 622; strain in, 622. 688 SUBJECT INDEX Lactic Acid and oedema, 193. Laws of osmotic pressure, 155; of distribution, 166. Lead, and nephritis, 425. Lesions, anatomical, of vascular disease, 500. Leukemca, and oedema, 194; and nephritis, 423. Levulose, and fibrin, 52, 53; and gelatin, 84, 86; and diuresis, 309. Ligation, and oedema, 178; of renal vessels, 226. Lipoids, membranes of. 111, 159; and distribution coefficient, 166; and solubility, 166. Liquid Colloids, hydration and dehydration in, 106; viscosity of, 107. LivBE, passive congestion of, 225; ligation of vessels in, 229; osmotic be- havior of, 233; yellow atrophy of, 495. Living tissues, 233; muscle, 394. Local CEdemas, 198. Loss of kidney substance, 492. Lymph, role of, in absorption, 257; compared with blood, 258; composition of, 258; formation of, 327; salts and formation of, 328; and drugs, 328; why it remains in the lymph vessels, 333. Ltmphagogues, 3'28. Ltophilic Colloids, 40. Lyophobic Colloids, 40, M Magnesium Sulphate, as tissue dehydrant, 601 . Mania, 497. Marasmus, 601. Medical Patients, prophylaxis of nephritis in, 648. Membranes, osmotic. 111, 154; lipoid. 111, 159; precipitation, 154; semi- permeable, 154. Mercury Drop, and diapedesis, 473. Methyl Orange, 635. Methyl Red, 635. Milk Diet, in nephritis, 547. Mineral Water, in oedema, 195; in nephritis, 543. Model, of urinary secretion, 283; of growth curvatures, 376; of diapedesis, 473. Molar Solutions, 47. Molecular Solutions, 48. MoHPHiN, and oedema, 203; and urinary secretion, 317; and lymph forma- tion, 328; and nephritis, 425. Morphological Changes, in nephritis, 447; and clinical manifestations of nephritis, 448; catalogue of, in kidney in nephritis, 454. Mosaic Theory, of Nathansohn, 161. Mucous Colitis, 601. Muscle, swelling of, 109, 359; osmotic properties of, 110; and acid, 112; and alkali, 114; and salts, 117; and non-electrolytes, 122; contraction of, 377; analogy between contraction of, and catgut, 387; historical and critical remarks on contraction of, 389; r61e of acid in contraction of, 392; living and dead, 394. SUBJECT INDEX 689 N Nephrectomy, 492. Nephritis, and oedema, 196; and pulmonary oedema, 236; definition of, 399; common cause for, 400; thesis of, 400; abnormal production and accumulation of acid in, 401; urine in, 404; decreased alkalinity of blood in, 405; and staining of kidney, 406; and ammonia excretion, 413; low carbonic acid content of blood in, 413; abnormal acids in, 413; blood colloids in, 414; tolerance to alkali in, 414; after acid in- jection, 415; of hard work, 418, 535; of athletes, 419; in heart disease, 422; in respiratory disease, 422; in anemia, 423; in carbon monoxid poisoning, 423; in epilepsy, 423; in leukemia, 423; after exposure to cold, 423; after direct interference with kidney circulation, 424; in vascular disease, 425; consequent upon drugs, 425; consequent upon anesthetics, 425; consequent upon poisons, 425; of the newborn, 426; of salt starvation, 427; after excessive water consumption, 427; of other than acid causes, 428; due to alkali, 428; morphological changes in, 447; classification of types of, 448; parenchymatous, 448; correla- tion of morphological and clinical manifestations in, 448; chronic inter- stitial, 449; hemorrhage in, 470; alleged consequences of, 482; relation of, to vascular disease, 482; and oedema, 491, 534; and uremia, 493; reinterpretation of associated manifestations, 494; disturbances of secretion in, 501; secretion of water in, 503; secretion of dissolved substances in, 510; experimental foundations for treatment of, 518; asphyxial, 519; and Ringer solution, 522; relief of, by sodium ohlorid, 525; following clamping of renal vessels and relief of, by sodium chlorid, 529; interpretation of experiments on, 533; treatment of patients with, 537; use of alkali in, 638; use of salts in, 538; use of dextrose in, 538; general rules for treatment of, 538; diet in, 539; acids and treatment of, 539; water consumption in, 544; r61e of salts in, 547; milk diet in, 547; physiological salt solution in, 548; more aggressive treatment of, 549; bivalent metals in, 554; treatment of severe cases of, 556; treat- ment of alleged consequences of, 556, 591; toxic types of, 569; after phosphorus and metallic poisoning, 571; treatment of, due to preg- nancy intoxication, 574; treatment of chronic interstitial, 587; use of alkali in chronic interstitial, 588; clinical histories of fatalities in, 595; prognosis in, 614; significance of casts in, 615; and heart failure, 616; high blood pressure in, 618; brain oedema in, 618; kidney efficiency tests in, 621; acidity measurements of urine in, 629; ammonia deter- minations in, 640; acetone bodies in, 641; uremia of, 642; prophylaxis of, 644. Nerve Blocking, 647. Nerves, vasomotor and secretory, 327, 328, 329, 330. Nervous Tissue, swelling of, 136; and acids, 138; and salts, 139; and non-electroiytes, 146. Neutral Red, 615. Non-acid Causes of oedema, 220; of nephritis, 428. NoN-ELECTROLYTEs, and fibrin, 51; and gelatin, 75, 81, 82; and muscle, 122, 362; and eye, 135; and nervous tissue, 146; and oedema, 216; absorption of, from peritoneal cavity, 264. Normal Solutions, 48. 690 SUBJECT INDEX O Oat Diet, 543. CEdbma, problem of, 37; main discussion of, 175; osmotic theory of, 177; ligation experiments on, 178; cause of, resides in tissues, 178; nature and cause of, 190; colloid-chemical theory of, 190; abnormal acid ac- cumulation in, 190, 191; and circulatory disturbances, 191; and lactic acid, 193; in pernicious anemia, 194; in leukemia, 194; in chloro- sis, 194; in inanition, 194; and alkalinity of blood, 194; in starvation, 195; in fever, 196; of nephritis,. 196; of the dead, 198, 202; local, 198 of insect stings, 199; and proteolytic ferments, 200, 223; and poisons, 203; rehef of, 207, 218, 607; and salts, 211; and non-electrolytes, 216 and sodium chlorid, 220; of non-acid origin, 220; due to alkaU, 224 due to pjrridin, urea and amins, 224; blood pressure theory of, 228 of lungs, 233; and syneresis, 240; and transudates, 240; and serous accumulations, 240; of cachexia, 244; of nephritis, 491, 534; and uremia, 494; fallacy of salt restriction in, 518; reduction of, by sodium chlorid, 528; idiopathic, 540; of the brain, in sulphuric acid poisoning, 591; treatment of, with sodium carbonate-sodium chlorid solution, 601; angioneurotic, 601; as alleged consequence of sodium chlorid retention, 602. Opacities, corneal, 664. Optic Nerve, cedema of, 494. Oranges, 536. Osmotic, properties of muscle, 110; membranes. 111, 154; theory of absorp- tion, 153, 273; cells, 154; theory of cedema, 177; behavior of kidney, 232; behavior of liver, 23i5; equilibrium, 276; forces in growth, 374. Osmotic Pressure, laws of, 155. Oxalic Acid, formation of, in plants, 377. Oxygen, lack of, and oedema, 192; consequences of lack of, 193; lack of, induced chemically, 206; supply and secretion, 316. Paranitrophenol, 635. Parenchymatous Nephritis, generalized, 448; spotty, 448; cedema of, 492. Paroxysmal Hemoglobinuria, 371, 423, 424. Partition, see Distribution. Passive Congestion, of kidney, 225; of Uver, 225; explanation of, on colloid-chemical basis, 225. Peritoneal Cavity, absorption of water from, 265, 259; absorption of dissolved substances from, 257; absorption of blood from, 261, 609; absorption of colloids from, 261; absorption of salt solutions from, 262; absorption of non-electrolytes from, 264; absorption of acid and alkali from, 265; secretion of fluid into, 611. Peritoneum, see Perctoneal Cavity. Pernicious Anemia, 194, 423. Phenolphthalein, 635. Phbnolsulphonbphthalein, in alkali therapy, 542; as kidney test, 627. Phosphorus, and nephritis, 425, 571. Physiological theory of absorption, 273, 277; driving force, 277. Physioloqecal Salt Solution, 48, 361; in nephritis, 548; perfusion of kidneys with, 599. SUBJECT INDEX 691 Physiologischb Triebkbaft, 277. Phtsostygmin, and lymph formation, 328. PiLOCARPiN, and lymph formation, 328. Plants, proteins in, 106; growth curvatures in, 374; growth in, 374; pro- tection against water loss in, 376; formation of oxalic acid in, 377. Plasmolysis, 355. Plasmoptysis, 355. Plethora, 242. Polyuria Tests, 625. Postoperative Treatment of surgical patients, 647. Precipitation Membranes, 154. Pregnancy Intoxication, theory of, 574; clinical manifestations of, 574; treatment of, 574; convulsive seizures in, 598. Preparation, of sodium carbonate-sodium chlorid solution for rectal use, 550; of sodium carbonate-sodium chlorid solution for intravenous use, 556; of surgical patients, 644. Pressure Bottle, 561. PrbssureTheory, of oedema, 176; of secretion, 315; of urinary secretion, 505. Primarily Contracted Kidney, 450. Prognosis, in nephritis, 614; in uremia, 629. Prophylaxis, of nephritis, 644. Protein, sweUing of, 41; plant, 106; analogy of, with protoplasm, 109; solution of, 433; in nephritic diet, 540. Proteolytic Ferments, and oedema, 200, 223. Protoplasm, analogy of, with protein swelling, 109; absorption and secre- tion of dissolved substances by, 164. Pulmonary (Edema, 233; due to circulatory disturbances, 234; due to acid, 236; in nephritis, 236; in vascular disease, 236; in excised lungs, 237. Pyridin, and fibrin, 55; and gelatin, 94, 95, 96; and oedema, 224. Q Quantitative Aspects, of water absorption in muscle, 126; of water absorp- tion in gastro-intestinal tract, 270. Quellungswasser, 363. R Reaction, after intravenous injections, 564. Rectal Injection, of sodium carbonate-sodium chlorid solution, 550; amount of, 553. Relief, of oedema, 207, 218. See also, Treatment. Removal of kidney substance, 492. Respiratory Disease, and nephritis, 422. Retention, of water, 289; of sodium chlorid and oedema, 602. Retinitis, albuminuric, 496. Reversibility, of water absorption in fibrin, 53; of water absorption in gelatin, 91; of water absorption in muscle, 123; of water absorption in eye, 135; of water absorption in nervous tissue, 148; of changes in nephritis, 448. 692 SUBJECT INDEX RrGOB, 389, 391, 396. Rigor Moetis, 396. Ringer Solution, and catgut, 384; in treatment of nephritis, 522. Saccharose, and fibrin, 52, 53, 56; and gelatin, 82, 86; and muscle, 124; and eye, 135; absorption of, from peritoneal cavity, 264; and diuresis* 311. Saline Cathartics, 269, 280. Saline Diuretics, 295; mode of action of, 295. Salines, effect of, on blood pressure, 339. Saliva, reaction of, 638. Salivary Glands, changes in, during rest and activity, 329. Salts, and fibrin, 47; and gelatin, 63; and muscle, 117; and eye, 130; and nervous tissue, 139; and oedema, 211; absorption of, from peritoneal cavity, 262; diuretic, 295; and lymph formation, 328; and catgut, 382; and solution of fibrin, 437; and solution of gelatin, 441; and cloudy swelling, 461, 463; use of, in nephritis, 538, 547; starvation in nephritis, 548. Salt Solutions, fate of, after intravenous injection, 334. Salvarsan, 601; avoidance of after effects from, 648. Scarlet Fever, and nephritis, 570, 571. Scurvy, 601. Seaweed, 270. Secondarily Contracted Kidney, 449. Second Order of diuretics, 314, 317. Secreting System, 282; physical chemistry of, 282. Secretion, general problem of, 249, 281; of water by kidney, 286; surface tension theory of, 305; influence of circulation on, 315; blood pressure theory of, 315; and oxygen supply, 316; of dissolved substances by kidney, 324; disturbances of, in nephritis, 501; of water in nephritis, 503; of dissolved substances in nephritis, 510, 512; into cavities, 240, 611. See also, Urinary Secretion. Secretory, theory of absorption, 277; work of kidney, 314; nerves, 327, 330. Selective, absorption, 275; secretion by kidney, 326. Semipermeable Membranes, 154. Serous Cavities, fluid accumulations in, 242. Serum, see Blood and Blood Serum. Shock, cause of, 345; cases of, 347; toxemic, 598; protection against, 647. Small Gray Kidney, 454, 496. Small Red Kidney, 449, 496. Sodium Carbonate, 549; equivalents of different kinds, 550. Sodium Carbonate-Sodium Chlorid Solution, preparation of, 549; rectal injection of, 550; amount of, to be injected, 553; preparation of, for intravenous use, 556; intravenous injection of, 559; quantity and time interval for intravenous injection, 562; reaction following use of, 564; as general treatment for oedema, 601; use of, in acute infections, 601. Sodium Chlorid, and oedema, 220, 528; effects of, on urinary secretion, 287, 523; and water retention, 289; starvation and nephritis, 427; restriction of, in nephritis, 518, 602; relief of asphyxial nephritis by. SUBJECT INDEX 693 619; effects of, on urinary secretion in nephritis, 523; reduction of oedema by, 528; inhibition of hemolysis by, 528. Sodium Chlorid Retention, 602. Sodium Chlorid-Sodium Citrate Solution, 601. Sodium Citrate, injections of, in glaucoma, 656., Sodium Indigosulphonate, staining of kidneys by, 409, 515. Solubility, and distribution, 166. Solution, of gluten, 105; of protein, 433; of fibrin, 434; of gelatin, 439. Solution, Fischer's, see Sodium Carbonate-Sodium Chlorid Solution. Solutions, molar, gram-molecular, and molecular, 47, 48; normal, 48; physiological, 48, 361; absorption of colloid, 261, 272; effect of colloid, on urinary secretion, 292. Solution Theory, of albuminuria, 432. Spermatozoa, absorption of water by, 356. Spring Water, see Mineral Water. Starvation, and oedema, 195; salt, and nephritis, 427, 548; acidosis, 542. Stomata, 470. Strain Tests, 622. Strychnin, and oedema, 203; and nephritis, 425. Stupor, 497, 557, 601. Sugars, and fibrin, 53; and gelatin, 75; and muscle, 123; and eye, 135; absorption of, from peritoneal cavity, 264; absorption of, from gastro- intestinal tract, 269; diuretic action of, 305; use of, in treatment of nephritis, 600. Sugar Diuresis, 305. Surface Tension Theory of secretion, 305. Surgical Patients, preparation of, 644; acid intoxication in, 645; post- operative treatment of, 647. SUSPENSOIDS, 40. Sweating, effects of, 555. Swelling, of fibrin, 41; of gelatin, 57; of gluten, 102; of muscle, 109, 359; of eye, 127; of nervous tissue, 136; cloudy, 455. Stneresis, 240. Tartaric Acid, and nephritis, 543. Testicle, 446. Tests, efficiency, of kidney, 621 ; strain, 622. Theory, osmotic, 153; lipoid, 159; mosaic, 161; colloid-chemical, 162; of oedema, 175; increased permeability, of cedema, 177; colloid-chemical, of cedema, 190; blood pressure, of oedema, 228; of absorption, 271; filtration, of absorption, 272; osmotic, of absorption, 273; physiological secretory and vitalistic, of absorption, 273, 277; surface tension, of secretion, 305; blood pressure, of secretion, 315; of urinary secretion, and criticism thereof, 319; of muscle contraction, 389; of albuminuria, 431; solution, of albuminuria, 432, 443; of cloudy swelling, 455; of diapedesis, 470; coUoid-chemical, of diapedesis, 472; of urinary secre- tion, 504; coUoid-chemical, of urinary secretion, 508; of pregnancy intoxication, 574; colloid-chemical, of glaucoma, 654. Therapy, see Treatment. Thoracic Duct, peritoneal absorption after ligation of, 258, 694 SUBJECT INDEX Thtmolphthalbin, 635. Tissue Spaces, 242. TiTRATioN.AciDiTY, of uHne, 630. TOLUIDIN Bltjb, 514. Toxemic Shock, 698. Toxic, effects of distilled water, 332, 360; types of nephritides, 569, 572. Tbanspusion, of blood, 346; of hydrocele fluid, 346; of ascitic fluid, 346; of colloid solutions, 346. Transition, from physiology to pathology of kidney, 322. Transudation, 240, 607. Trauma, see Injury. Treatment, of oedema, 207, 218, 607; of low blood pressure, 345; of nephritis, 518; more aggressive, of nephritis, 549; of nephritis with bivalent metals, 554; of severe cases of nephritis, 556; of pregnancy intoxications, 574; of chronic interstitial nephritis, 587; explanation of good results following sodium chlorid restriction, 611; of high blood pressure, 619; postoperative, 647; of glaucoma, 655. Tropisms, 374. Turgor, 154, 355. Twitching, 601. U Uranium, and oedema, 203; and nephritis, 425. Urea, and fibrin, 55; and gelatin, 94, 95, 96; and muscle, 122; and eye, 135; and nervous tissue, 146; and oedema, 224; absorption of, from peritoneal cavity, 264. Uremia, 493, 497; not secondary to kidney loss, 493; as a brain oedema, 494; periodic character of, 498; prognosis in, 629; and nephritis, 642. Urinary Secretion, model of, 283; conditions for, 286; effect of sodium chlorid on, 287, 523; effect of blood injections on, 292; effect of colloid solutions on, 292; blood pressure theory of, 315; effect of drugs on, 317, 318; critical remarks on, 319; theories of, 504; pressure theory of, 505; colloid-chemical theory of, 508; effect of sodium chlorid on, in nephritis, 523. Urine, physical chemistry of, 282; in nephritis, 404; acidity of normal, 404; acidity of abnormal, 404; acidity measurements of, 629; titration acidity of, 630; hydrogen ion acidity of, 631; use of indicators in exam- ination of, 635. Urticaria, 199, 602. Varicose Veins, 500. Vascular Disease, and pulmonary oedema, 236; and nephritis, 425, 589; x-rays of, in kidney, 451; definition of, 451; relation of, to nephritis, 482; relation of heart hypertrophy to, 484, 488; etiology of, 499; anatomical lesions of, 500; varicose veins in, 500; heart failure in, 690; labored breathing in, 601; foci of infection in, 621. Vasomotor Nerves, 327, 328, 329. Vegetables, in nephritic diet, 541. Velocity, of blood and secretion, 316. Viscosity, of liquid colloids, 107; of blood, 490. Vitalistic Theory of absorption, 273. SUBJECT INDEX 695 W Water, of condensation, 240; free, 262; retention, 289; in growth, 374 loss of, in plants, 376; excessive consumption of, and nephritis, 427 consumption of, in nephritis, 544; advantages of, in nephritis, 544 objections to, in nephritis, 546. Water Absorption, by iibrin, 41, 53; by gelatin, 57; by gluten, 102; by muscle, 109, 359; by eye, 127; by nervous tissue, 136; osmotic theory of, 153; from peritoneal cavity, 256, 259; from gastro-intestinal tract, 269; by spermatozoa, 356; by epithelial cells, 356; by white blood corpuscles, 356. Water Secretion, by kidney, 286; conditions affecting, 314; in nephritis, 503; effect of sodium chlorid on, in nephritis, 523. See also Secretion. Wet Dressings, 601. Wheals, 199. Work, of kidney, 314, 507; demands on heart, 488. X ^-rays of vascular disease in kidney, 451.