(Ufllumliia UniurrBity ^ ^^ iiiJhrUIitgof Npui ^ork l&tfnmtt ICtbrarg PUBLIC HEALTH LABORATORY WORK Digitized by tine Internet Arcinive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/publichealthlabo1914kenw PUBLIC HEALTH LABORATORY WORK BY HENRY R. KENWOOD, M.B., F.R.S. Edin. D.P.H.. F.C.S. CHADWICK: fROFESSOR OF HYGIENE AND PUBLIC HEALTH, UNIVERSITY OF LONDON ; .MEDICAL OFFICER OF HEALTH AND PUBLIC ANALYST FOR THE METROPOLITAN BOROUGH OF STOKE NKWINGTON ; EXAMINER IN PUBLIC HEALTH TO THE ROYAL COLLEGES OK PHYSICIANS AND SURGEONS, LONDON, ETC. SIXTH EDITION, WITH ILLUSTRATIONS PAUL B. HOEBER 67 & 69 EAST 59TH STREET NEW YORK 1914 [Printed in England. | PREFACE The first edition of tiiis book, while dealing mainly with chemical matter, contained a useful resume of public health bacteriological work, and this contribution was continued in subsequent editions; but it is felt that the time has now come to exclude all but occasional references to bacteriological matters, in order to keep the volume within the compass of a handy laboratory guide to the chemical branch of public health laboratory work; for the needs of the public health student in bacteriology are now fully provided for by several excellent laboratory works. The book does not describe a large number of methods to the same end. It has always been the aim of the writer to simply and sufficiently describe selected processes which experience has proved to meet satisfactorily the requirements of the public health worker. In the section on Food, while prominence is given to adul- teration which raises the presumption of a danger to health, it has not been judged wise to exclude all reference to other matter; for the public health student is still required at some examination centres to show a knowledge of laboratory methods for the detection of sophistication which has no bearing on health, such as, for instance, the methods which serve to distinguish between butter-fat and other fats used as substitutes. I have to acknowledge gratefully my indebtedness to Mr. F. Marchant for assistance rendered in the preparation of this new edition; and I have also to thank Messrs. Townson and Mercer, Messrs. Baird and Tatlock, and others, for the loan of blocks. H. R. K. London, 1914. CONTENTS PAGE INTRODUCTORY NOTES ----- i PART I THE CHEMICAL, MICROSCOPICAL, AND PHYSICAL EXAMINATION OF WATER FOR PUBLIC HEALTH PURPOSES CHAPTER I. THE COLLECTION OF SAMPLES INFORMATION REQUIRED AS TO SAMPLES QUANTITATIVE EXPRESSIONS - 19 II. THE PHYSICAL CHARACTERS OF WATER - -2^ in. CHLORINE - - - - - - 31 IV. HARDNESS - - - - - "37 V. THE POISONOUS METALS - - - - 44 VI. CALCIUM AND MAGNESIUM SALTS SILICA SULPHATES PHOSPHATES - - - - - 54 VII. THE SOLID RESIDUE - - - - - 60 VIII. THE EXAMINATION OF SUSPENDED AND DEPOSITED MATTER IN WATER - - - - - 64 IX. ORGANIC MATTER IN WATER - - " ' 75 X. WANKLYN'S PROCESS - - - - "77 XI. THE OXIDIZABLE ORGANIC MATTER E. FRANKLAND's PROCESS - - - - - - 86 XII. OXIDIZED NITROGEN (NITRATES AND NITRITES)- - 92 XIII. THE GASES IN WATER ----- loi XIV. COMPOSITION OF WATER FROM VARIOUS SOURCES THE OPINION ON WATER SAMPLES - - I ID XV. SEA WATER ------ 130 XVI. ALKALIMETRY AND ACIDIMETRY ICE MINERAL WATERS ANALYTICAL SCHEME . - _ _ 136 PART II THE ANALYSIS OF SEWAGE AND OF SEWAGE EFFLUENTS - - - i43 PART III SOIL EXAMINATION - - 157 Vlli CONTENTS PART IV AIR ANALYSIS CHAITER TAGK I. THE NORMAL CONSTITUENTS OF AIR — OXYGEN — EUDIO- METRY - - - - - - 173 II. CARBONIC ACID - - - - - "179 III. THE ORGANIC MATTER IN THE AIR - - - I92 IV. AMMOXI.\ — MARSH GAS — CARBON MONOXIDE — SULPHUR COMPOUNDS — NITRIC, NITROUS AND HYDROCHLORIC ACIDS — PHOSPHURETTED AND ARSENIURETTED HYDRO- GEN ------- 195 V. OZONE — PEROXIDE OF HYDROGEN - . - 206 VI. SUSPENDED MATTER IN THE AIR - - - 2IO VII. THE CHARACTERS OF THE AIR COLLECTED FROM VARIOUS SOURCES — BACTERIOLOGICAL NOTE - - - 216 VIII. SCHEME FOR THE DETECTION OF GASES WHEN PRESENT IN LARGE QUANTITIES . . - . 221 PART V FOOD EXAMINATION I. COMPOSITION OF COW'S MILK AND OF OTHER MILK - 225 II. THE ANALYSIS OF MILK . . - - 229 III. THE SOPHISTICATION OF MILK — MILK PREPARATIONS ■ MILK STANDARDS BACTERIOLOGICAL NOTE - - 242 IV. BUTTER — CHEESE LARD . - - - 253 V. CORN — WHEAT-FLOUR ----- 266 VI. BREAD ------- 278 VII. THE AVERAGE COMPOSITION OF OTHER FLOURS AND MEALS THE MICROSCOPIC CHARACTERS OF THE DIF- FERENT STARCH GRANULES - - - - 284 VIII. MEAT PARASITES OF FLESH — POISONING BY FOOD MEAT PREPARATIONS ----- 293 IX. ALCOHOLIC BEVERAGES - - - - "S^S X. VINEGAR — LIME AND LEMON JUICE — MUSTARD — PEPPER SUGAR — HONEY ----- 332 XI. COFFEE — COCOA CHOCOLATE . - - - 343 XII. TEA INFANTS' FOODS ----- 35O XIII. PRESERVED AND TINNED PROVISIONS - - - 358 XIV. CHEMICAL ANTISEPTICS AND COLOURING AGENTS IN FOOD 366 XV. ARSENIC IN FOOD — ARSENIC IN WALL-PAPERS, ETC. RAG FLOCK ------ 387 PART VI THE EXAMINATION OF DISINFECTANTS - 399 INDEX _._--- 413 LIST OF ILLUSTRATIONS Plates I. -IV. — Objects found in Impure Water (Variously Mag- nified) . 1. Water-Bath and Drying-Oven - - - - - 2 2. Chemical Balances (Townson and Mercer) - - - 3 3. Water-Bath with Constant Water-Level - - - - 6 4. A Specific Gravity Flask ------ 7 5. The Westphal Balance ------ 8 6. Soxhlet's Fat-Extraction Apparatus - - - - 10 7. The Spectroscope - - - - - -11 8. The Polariscope -- - - - - -I3 9. A Burette filled up to the 10 c.c. Mark - - - - 15 10. Wynter Bl3rth's Tube for collecting Sediments - - - 64 11. Showing the Sediment of a Pond-Water, a Sample of which was collected in the Early Spring - - - - - 65 12. Vegetable Tissue - - - - - -67 13. Ciliated Embryo and Cercaria Form of Distoma hepaticum - 73 14. Apparatus for Wanklyn's Process - - - - 79 15. Apparatus for Thresh's Process - - - - -105 16. Collecting-Bottle and Tin - - - - 125 17. Ice-Box - - - - - -- - 125 18. Apparatus for Adency's Process . - - - 150 19. Arrangement for Registering the Varying Levels of the Ground- Water - - - - - - - 158 20. Knopp's Soil-Washing Cylinder (Townson and Mercer) - - 161 21. Fraenkel's Borer ------- 170 22. Hempel's Gas Burette and Absorption Apparatus - - I75 23. Hempel's Double Absorption Pipette (Townson and Mercer) - 177 24. The Flexible Bellows-Pump employed by Angus Smith to draw out Air from the Air-Jar . . - - - 181 25. Store Bottle for Baryta Water ----- 183 26. The Apparatus for the Lunge and Zeckendorf Process - - 186 27. Haldane's Apparatus for the Estimation of CO2 - - - 188 28. Apparatus for collecting the Organic Matter in Air - - 193 29. Showing the Characteristic Disposition of the Absorption Bands in the Spectroscopic Picture of Oxy- and Reduced Haemo- globin - - - - - - - -198 X LIST OF ILLUSTRATIONS FIG. I'AOK 30. The Ozone Cage (Negretti and Zambra) - - - - 207 31. M. Marie-Davy's Modification of Pouchet's Aeroscope - - 211 32. Hesse's Apparatus for the Collection of Suspended Matters in Air 212 33. Hesse's Apparatus for collecting Ground Air - - . 219 34. The Cream Tube (Townson and Mercer) - - - - 232 35. Stokes' Tube for the Werner-Schmidt Process (Townson and Mercer) ....... 235 36. Centrifugal Apparatus for Milk-Testing - - - - 238 37. Milk, showing the Large Colostrum Corpuscles - - - 240 38. Del6pine's Milk-Collecting Apparatus . - . - 251 39. Apparatus for the Keichert-Wollny Process . . - 257 40. Aspergillus glaucus ...... 263 41. Mucor mucedo .....-- 264 42. The Cheese Mite {Acarus domesticus) .... 264 43. The Corn Weevil (Calandra granaria) .... 266 44. Vibriones tritici - ...... 266 45. A Wheat Spikelet with Ear-Cockle .... 267 46. The Wheat Mite {Acarus farina) .... 267 47. Ear of Rye with Ergot (A), and a Section of Ergot (B) - - 267 48. Ergot - - - - - - - - 268 49. Smut Spores {Uredo segetum) ..... 269 50. Bunt {Uredo fastida) ...... 269 51. Tilletia caries (Bunt) and Tilletia IcBvis - - ~ - 269 52. Puccinia graminis - - - - - -270 53. Section through Branny Envelope and Outer Portion of Endo- sperm of Wheat Grain - - - - - 270 54. Penicilliiim glaucum -_.._- 280 Plates V. and VI. — Illustrations of Starch Grains - facing 286 55. Bruchus Pisi (of the Pea, Bean, etc.) .... 288 56. Section of Wheat Grain (Outer Coat) ... - 289 57. Wheat Tissue from the " Testa " of the Grain . - - 290 58. Barley Tissue from the " Testa " of the Grain - . - 290 59. Rye Tissue from the " Testa " of the Grain . . - 291 60. Oats Tissue from the " Testa " of the Grain . . _ 291 61. Coccidium oviforme - - - - - -297 62. Head of Tesnia solium -....- 299 63. Head of TcBnia mediocanellata ----- 299 64. Brood Capsule of an Echinococcus . . - - 300 65. Trichina spiralis, encysted in Muscle .... 301 66. One of Rainey's Capsules ..... 303 67. Distoma hepaticum __...- 304 68. Apparatus for estimating Fusel Oil by Rose's Process (Baird and Tatlock) .--.... 324 69. Torula cerevisics (Yeast Plant) .... - 326 70. The Cuticle of the White Mustard-Seed - - - - 336 71. Pepper - - - - - - - - 337 72. The Sugar Mite {Acarus sacchari) - . . . 339 73. Cofiec: Cells of Testa and Cellular Structure - - - 3^5 LIST OF ILLUSTRATIONS XI FIG. I'AGE 74. Chicory: Dotted Ducts and Cellular Structure - - - 34O 75. Lacteal Vessels of Chicory - - - - - 347 76. Cocoa Starch Cells _..... ^48 77. The Elder Leaf (after Bell) - - - - -351 78. The Willow Leaf (after Bell) - - - - - 351 79. The Sloe Leaf (after Bell) . _ . . . ^52 80. The Tea Leaf (after Bell) ..... ^52 81. The Epidermis of the Under Surface of the Tea Leaf - - 353 82. Section of a Tea Leaf, showing Idioblasts ... ^53 83. Crystals of Arsenious Acid . _ . _ . ^BS 84. Marsh's Apparatus for Testing for Arsenic (from the Analyst) - 389 85. Arsenical Mirrors (W. Thomson) - - . . - 391 86. Delepine's Apparatus for the Oxidation of Arsenic and Sublima- tion of Arsenious Acid from Deposits of Arsenic on Copper - 393 87. Apparatus employed in the Rideal-Walker Method - - 408 PLATE Paramaecmn NavLcida jEpii/tyd/xiJ. CeUs Lepiolhnx _ A FUamenlm sheath ' B. » willwulshealh Cotton Fibres. IyLn,en,J^ibres Miol Fiire . Sa/c Fibre. v Foj-ticZe- Cff/catfiej- Clalhrocyslis Aslenonella A^Uenrice o/'jn^&ct^ OBJECTS FOUND IN IMPURE WATER (VARIOUSLY Magnified). PLATE II. A Tl^zier-- Bear J^ava, cfJ^i^eci I'upa, ofln^eci OBJECTS FOUND IN IMPURE WATER (variously Magnified). PLATE III ,^^A °im^ q/m Ad!yce^uufZ' FbwiaiUis Pollen. Spores of Funffi A Con/erroid celi. Sacte/'Ucm, iefmo. ADiatom' A Ue-STmd Piffetailet Spore ^ Sitrireiia- Fun^z^ Jiil^i/ of PJfiUry^lanfain frocococc-cu. PUiMiali^ J^ella! Lcb fioccv^a. A uqleri-a. Vert di^ CoTi/erva, ffoTTZpTioncTTia, OsaUaj-ia OBJECTS FOUND IN IMPURE WATER (variously Magnified). PLATE IV. VoLvox globator 1 1 Diiriu-gia Amirea cochleans Stawashinn crenulatum Spirogyra Egg or Q Dolhrwcephulus Latas AnabsTicu CUosleruim Diana; Egg of Tapnia Sohxum, Egg of Ascaris LunzbricohcLes Egg of T Medio canellata Egg of Egg of Egg of Tricocephalvis Dbspar Ankyloslomumdiwdenale Oxinms Vermiciilaris OBJECTS FOUND IN IMPURE WATER (variously Magnified). PUBLIC HEALTH LABORATORY WORK INTRODUCTORY NOTES The Collection and Weighing of a Precipitate. The substance is precipitated from a known bulk of liquid, and the precipitate is collected on a filter-paper which has been folded and placed inside a glass funnel. The filter - paper should never project beyond the funnel, and the fluid should be conducted by a glass rod on to the filter-paper, to prevent loss. Special filter-papers are sold which yield an ash which is generally quite insignificant; and the amount of ash furnished by them is a definite and known quantity for each paper. Such papers should be always employed for collecting pre- cipitates which have subsequently to be ignited and weighed. The precipitate on the filter-paper is next washed with dis- tilled water from a wash-bottle. A fine jet of distilled water, either hot or cold, is generally employed for this purpose. The process of washing is complete if a drop of the last washing yields no residue when evaporated on a platinum spatula. The precipitate on the filter-paper is then dried in a drying- oven. Fig. I represents Wills' combined water-bath and drying- oven. It consists of a hot-water bath with openings for evaporating dishes, a spacious hot-air chamber, a pair of hot- water funnels for filtering fatty substances which tend to solidify when cool, and a hot-air box for drying test-tubes, etc. The thermometer, in situ, registers the temperature in the interior of the oven. 2 LABORATORY WORK The filter-paper should then be folded up, placed in a small porcelain crucible (previously weighed), and covered by a lid; the filter-paper and precipitate are next ignited to dull redness, at first gently so as to obviate spurting and loss, but the lid should be removed after a little so as to permit free access of air. It is also desirable to keep the porcelain dish on the slant during ignition, since this favours air draught. When the filter- paper has been entirely destroyed, the capsule and its contents are allowed to cool under a desiccator, and weighed. The weight found, minus that of the crucible and the ash of tlie filter-paper, FIG. I. WILLS WATER-BATH AND DRYING-OVEN. represents the weight of mineral precipitate. Care must be taken to remove the dish from the flame immediately all evidence of charring or discoloration has disappeared, and not to conduct the i icineration at a higher temperature than is found absolutely necessary, or there may be a considerable loss in the mineral r3sidue. Such loss is most generally from ammonia salts (by volatilization), from nitrates and nitrites (by loss of oxygen), frjm certain chlorides, such as sodium and potassium chlorides (by volatilization), from combined carbonic acid, and from the INTRODUCTORY NOTES 3 water of hydratcd salts (such as calcium sulphate), which thercljy become anhydrous. A desiccator is simply a glass shade inside of which there is a vessel containing some agent which will free the air from moisture (such as strong sulphuric acid or solid calcium chloride). A residue completely dried by heat will absorb a little of the vapour from the atmosphere while coohng and thus increase sHghtly in weight, unless the precaution is taken to place it inside a desic- ^. FIG. 2. CHEMICAL BALANCES. cator during the coohng process. In the desiccator a perforated tray or a tripod supports the substance to be cooled, and the rim against which the cover closely fits is greased with tallow so that the desiccator is hermetically sealed. The balances shown in Fig. 2 will be found to be suitable to all weighing purposes. They consist of a short beam which supports two pans, the ends of the beam being constructed with straight knife-edges of agate, upon which the pans are suspended by agate planes. The case is fitted with a sHding window in 4 LABORATORY WORK front, which, even when closed, admits of the working of the scales by means of turning a screw which projects externally. The balances must be kept in a dry room, away from any fireplace or door, and placed on a perfectly firm and level surface. The operation of weighing consists of first lifting the beam off its support by turning the screw, and then noting, by the long indicator which hangs down in front of the central vertical support of the balances, whether the two pans exactly counter- balance each other; if not, the balance must be adjusted by means of a small mechanism situated on the top of the centre of the cross-beam, which can be moved to the right or left, accord- ing as it is necessary to increase the weight in either of these diiections. After thus seeing that the scales are accurately equipoised, the material is then placed upon one of the trays, and the weights are added to the other. After each alteration made in the weights, the result must, of course, be tested; and before any further addition or removal is made the scales must be brought to rest upon their supports, or the apparatus may be put out of gear. Each of the weights is marked. The larger brass weights (i to 50) represent grammes, the next in size decigrammes (o-i to 0-5), the next centigrammes (o-oi to 0-05); and small forceps are used for picking up and applying them to the pan. The milligrammes are added by a little piece of bent wire (the " rider "), which is carried by means of a shding-rod moving just above the level of the cross-beam, which beam bears markings numbered from i to 10. By sliding the rod, the " rider " may be carried to, and placed upon, an}^ one of these marks, when that number of milligrammes of weight will have been added. Each milligramme division is further subdivided to i parts of a milligramme. Example. — A small platinum dish is placed on the left-hand pan. A 5-gramme weight is placed on the other pan. The beam is raised by means of a half-turn of the screw; when the platinum dish is found to be heavier than the 5 grammes. The scales are put at rest by reversing the screw to its original position, and a 2-gramme weight is added to the 5. This is INTRODUCTORY NOTES 5 also carried up by the greater weight of the platinum dish. Another gramme is added; and being found to be too much, is removed. The dish therefore weighs between 7 and 8 grammes. A 5-decigramme weight {i.e., 0-5 gramme) is added. The platinum dish is still slightly the heavier ; therefore another deci- gramme is added, with the result that the weights now slightly overbalance the dish. The i-decigramme weight is removed. The dish therefore weighs 7-5 grammes, but not 7-6 grammes. A 5-centigramme weight (0-05) is added. This is not enough; but a 3-centigramme weight further added so extremely nearly establishes the required equihbrium that the addition of another centigramme is found to be too much. Therefore the dish weighs 7-57 grammes, but not 7-58 grammes. Three milhgrammes, added by means of the little " rider,' make the long indicator oscillate quite evenly on either side of the central mark on the piece of porcelain, where it would ulti- mately come to rest. The weight, therefore, of the platinum dish is : 7 grammes = 7 5 decigrammes = 0-5 7 centigrammes =0-07 3 milligrammes = 0-003 Total = 7 '573 grammes. The dish need not be reweighed on every occasio-n of using if it is thoroughly cleansed; the reweighing is only necessary at intervals of every few days. The Collection and Weighing of a Solid Residue AND Mineral Ash. A given weight of the liquid is placed in a clean weighed platinum dish. (A platinum dish is cleansed after use with a little dilute hydrochloric acid; then well washed in pure water; and finally heated to redness in the Bunsen flame. It should be allowed to cool under the desiccator prior to weighing.) The dish and its contents are then placed upon a water-bath. A water-bath is a receptacle which holds water, and admits of this water being heated to a certain temperature. When vessels containing liquids are made to stand over the heated water, evaporation of their fluid contents may be effected at a O LABORATORY WORK temperature which can never quite reach that of the boiling-point of water. The water-bath must not be allowed to boil dry. Fig. 3 shows an arrangement by which this may be guarded against, by the maintenance of a constant water-level in the bath. When the contents of the dish have evaporated to dryness, the dish is placed in the desiccator for half an hour to cool. It is then weighed, and the solid residue is the weight obtained less the weight of the platinum dish. The dish is then held by a pair of crucible tongs (which may, with advantage, be platinum-pointed) in the flame of a Bunsen FIG. 3. WATER-BATH WITH CONSTANT WATER-LEVEL. burner, until nothing but the mineral ash remains. Fletcher's burners are an improvement upon the common type of Bunsen burner, when it is required to employ a very small flame. The mineral ash is allowed to cool in the desiccator and then weighed. Specific Gravity or Relative Density. The relative density or specific gravity of a solid or hquid is generalty referred to water, taken as unity or as a thousand. Where possible, the test must be applied at the temperature of 15-5° C. ; but in the case of fats a higher temperature is necessary in order to obtain them in a liquid state. The student is already famihar with the float instruments or hydrometers which are commonly in use for obtaining specific gravities, and it is only necessary to point out the importance of verifying all these instruments prior to use, by comparing their indications with the results obtained by more delicate methods. The most accurate estimates of specific gravity are obtained by actual weighings in the specific gravit}^ bottle. The method may be illustrated by indicating how the specific gravity of butter-fat would thus be obtained. INTRODUCTORY NOTES 7 1. A quantity of the butter is heated to, and maintained at about 65° C. in a water-bath made by standing a small beaker containing the butter in a larger beaker containing water. 2. The fat slowly separates and forms an upper stratum, which rests upon a lower stratum of the water, curd and salt. 3. In the course of time the upper layer of butter-fat gets clearer and clearer, until at last, all the water, curd, and salt having separated, it becomes clear and transparent. Immedi- ately this has taken place the fat is decanted on to a fine dry filter, in order to guard against the presence of traces of curd and salt ; and the filtrate of pure butter-fat is collected and poured into a specific gravity bottle. The specific gravity bottle is a small vessel of thin glass, fitted with a thermometer which also forms a FIG. 4. A SPECIFIC GRAVITY FLASK. stopper to the bottle and which registers the temperature of the contained liquid, so that this may be known at the moment of weighing. This bottle must be accurately filled and then stoppered, care being taken that no air-bubble or empty space is allowed to remain between the stopper and the liquid. 4. The temperature at which the fat is poured into the specific gravity bottle should be a fraction above 38° C, when the bottle and its contents are transferred to the balance and weighed. The precise weight must be taken when the thermometer registers exactly 38° C, the flask is entirely filled with the fat, and there is no evidence of air-bubbles. The weight of the specific gravity bottle when completely filled with distilled water and closely stoppered at the temperature of 38° C, has been previously taken. By a comparison of the respec- 8 LABORATORY WORK live weights of the two fluids when occupying the flask at the same temperature, the specific gravity' of the butter-fat is obtained, that of distilled water being taken as i,ooo — i.e., S.G.= The weight of the fat at 38° C. The weight of the water at 38° C. X 1,000. 38° C. is here selected as the temperature for weighing because it is the lowest temperature to which it is quite safe to reduce FIG. 5. THE WESTPHAL BALANCE. the contents of the bottle without any solidification ensuing, all the fats (animal and vegetable) used as adulterants of butter remaining hquid at that temperature. The Westphal balance registers the specific gravity on the principle that a body immersed in a liquid loses a part of its weight which is equivalent to the weight of the displaced liquid. The apparatus (Fig. 5) has a swinging arm, which rests on a knife-edge, and the upper surface of a part of the arm is notched and graduated. At the free end of the graduated part of the arm is a hook, by which a glass plummet is suspended by means of fine platinum wire. Three different-sized riders (or weights) are provided, of which the largest indicates hundreds, the next INTRODUCTORY NOTES 9 tens, and the smallest units. At the other end of the arm is a metal pointer, and the balance prior to use must be so adjusted that, with the plummet immersed in distilled water, and the largest rider placed on the hook (which represents the tenth notch, or i,ooo), this pointer rests vertically opposite a small projection on the frame. The adjustment is made by means of the small screw shown on the vertical support to the frame. The liquid is placed in a glass cylinder and the plummet just completely immersed in the hquid, and by placing the riders on various notches the two pointers are again brought opposite to each other. If, for instance, in order to obtain this result, the largest rider is on the ninth notch, the next largest on the seventh, and the smallest on the fifth, the specific gravity would be 975. A correction for temperature is necessary for an exact observa- tion by float hydrometers, since all such instruments are originally graduated by water at the temperature of i5'5° C, and the specific gravity varies with the temperature. Within the ordinary ranges of temperature in a laboratory it is sufficient to add 1° of specific gravity for every 3° of temperature above i5'5° C, and to subtract 1° for every 3° below I5"5° C. The Extraction of Fat by Soxhlet's Apparatus. Soxhlet's apparatus is shown in Fig. 6. A is the small flask which has been thoroughly dried and weighed and then about half filled with ether ; the extraction apparatus is shown attached to the flask between it and the condenser (K), the latter being fixed in a very slanting position. F represents a piece of fat- freed paper containing the substance to be extracted; this is placed in D, care being taken that it is entirely below the level of the smiall siphon E, so that it may be completely immersed in the solvent, and also that it does not close the opening to the siphon. The weighed flask of the Soxhlet should have a capacity of about 150 c.c, and contain about 75 c.c. of ether. The dish on which the flask stands is partially filled with water, and this is cautiously heated; the ether vapour then ascends G, passes into the condenser, and is at once condensed and drops on to F; the ether goes on accumulating, rising the while in the ascending arm of E, until it reaches the level of the upper bend. 10 LABORATORY WORK and overflows, when siphonage takes place, and the ether passes out of D back to the flask. Thus the circulation of the ether is completed every few minutes. Immediately after a siphon dis- charge has returned all the ether to the flask, the latter is removed, the ether driven off over the water-bath at a temper: - ture sufficient to make the ether boil, after which the flask and its contents are dried at ioo° C. until a constant weight is obtained. Of course, there must be no doubt as to whether the extrac- tion has been complete; this may be tested by fixing a second FIG. 6. — soxhlet's fat-extraction apparatus. small flask containing more ether, and after about half an hour evaporating off the ether and drying at ioo° C. ; it can then be noted whether there is any material increase over the original weight of the flask. It is well to place a small plug of blotting-paper in the mouth of the open tube at the top of the condenser so as to limit the access of air, the moisture of which would otherwise condense and slightly wet the ether. INTRODUCTORY NOTES II The vSpectroscope. A knowledge of the spectroscope is useful to the public health worker, and for those unacquainted with the use of this instru- ment a brief description is given: If a compound light, such as sunlight, is made to pass through a glass prism, the different coloured rays of which it consists are unequally refracted (or bent out of their original course), so that beyond the prism they form, upon a white surface, a continuous line of colours called the " spectrum "; and the spectrum of the compound white light will be seen to consist, in order from right to left, of red, orange, yellow, green, blue, indigo, and violet. A number of dark hues — called " absorption bands " or " Fraun- hofer's hues " — are also seen to cross the image of the solar FIG. 7.— THE SPECTROSCOPE. spectrum. These lines indicate the absence of raj^s of certain refrangibilities from the beam of solar hght; each occupies a definite position, and therefore affords a means of accurately localizing the parts of the spectrum. In other lights the spectrum will only show a few bright bands (that of the sodium flame only one), and the remainder of the spectral image is almost — or quite — invisible, by comparison. If we transmit solar light through different coloured solutions, we then get different absorption bands. If a solution of fresh blood, for instance, be taken, and a small colourless cell containing it is placed before the slit in the instrument which admits the hght, two distinct and characteristic dark stripes or absorption bands appear in the yellow and green parts of the solar spectrum. 12 LABORATORY WORK Fig. 7 will serve to show the manner in which a spectroscope is constructed. A firm iron stand is seen to support at its upper end a brass plate carrjing the glass prism; laterally, a cjdinder is also fastened to the brass plate, and in the end of this cylinder which is nearest the prism a lens is placed, the other end being closed by a plate with a vertical slit in it (the width of which can be regulated by a screw to meet requirements) ; through this slit the light is admitted to the prism, the rays first passing through the lens and thereby being rendered parallel and condensed. The spectroscopic appearance is then viewed through a small tele- scope (with a low magnifying power), and this (the tube on the right as seen in the figure) is fitted on to the cast-iron foot so as to be movable in a horizontal plane about the axis of the foot. The telescope is made to move over a scale which can be read ^^^th a vernier. All foreign light must, of course, be cut off; and this may be done by a black cloth, which is thrown over the prism and the tubes. The sht may be furnished with a reflecting prism, by means of which two spectra can be compared at the same time. Thus, by a spectroscopic examination, the colour, number, and position of the bright lines on the spectroscopic scale may be carefully observed and noted. If it is desired to distinguish metals by means of their spectral lines, the substance is dissolved in a drop of the purest hydrochloric acid; a piece of recentty ignited platinum wire is then dipped in the solution and held in a Bunsen flame. A convenient method of performing spectroscopic observations is by means of the Sorby- Browning micro-spectroscope, which consists of a small spectroscope placed in connection with a microscope in such a way that the former fits into the tube of the latter, similar to an eyepiece. The Polariscope. A simple form of half-shadow polariscope consists of a hori- zontal brass tube mounted on a vertical stand, and having a Nicol prism at each end, one being the " polarizer," and the other the " analyzer." A monochromatic hght, such as the yellow sodium flame (which may be obtained by placing a INTRODUCTORY NOTES 13 platinum cup containing sodium chloride in the flame of a Bunsen burner) is admitted to the polarizer. At the opposite end of the brass tube an eyepiece is fitted just in front of the analyzer. In the brass tube can be placed a clean and dry glass tube con- taining the solution under examination. Light consists of vibrations of ether in all planes, and its trans- mission occurs in waves; but the monochromatic light consists of light of a single wave-length. The polarizer allows only the vibrations taking place in one plane to pass, others being inter- cepted. Now, when the analyzer is placed parallel to the polar- izer, all the vibrations pass through the analyzer also, and equal illumination is seen on both sides of a sharply defined vertical middle hne when looking through the eyepiece, this point of FIG 8. THE POLARISCOPE. equal illumination being called the " zero-point." The slightest rotation of the analyzer will then produce a difference in the illumination of the two sides. In using the instrument, the zero-point is first obtained, with the glass observing-tube filled with distilled water and placed in position; then if some sugar solution (or other optically active liquid, which has the property of rotating the plane of polarized light) be placed in the glass tube, the rays will no longer pass through the analyzer, and the equal illumination is disturbed. If the analyzer be turned round, it is possible to obtain an equal illumination, or, in other words, to compensate for the optical disturbance of the rotating substance; but the direction and the angle through which it has been turned (as indicated on a dial fitted with a vernier) vary with the amount and nature of the rotating substance examined, the number of degrees being termed 14 LABORATORY WORK the " index of refraction," from which the so-called " specific rotary power " of the substance may be calculated. The polariscope is used to find the percentage adulteration of butter with other fats (refractometer), and also the strength of saccharine solutions (saccharimeter). With pure butter an equally distributed light can be obtained, but with butter con- taining fat which has been melted (margarine) this is impossible, since such fats rotate the plane of polarization. Glucose in honey and added sugar to milk may also be detected by this instrument, for while most sugars have the property of deflecting the ray of polarized light to the right (dextro-rotary, indicated by the sign +), others deflect to the left (levo-rotary, indicated by the sign — ), and this affords a means of distinguishing between them. If the nature of the substance is known, one can, moreover, estimate its quantity, since i gramme of a particular optically active substance has its own specific rotary power. But for the work demanded of the public health worker it is not necessary to determine specific rotary powers, useful as these may be in some of the work which a public analyst may be called upon to perform. Indeed, from the standpoint of the public health worker the micro-polariscope (in which the polari- scope is adjusted to an ordinary microscope) will generally suffice. In this instrument one of the Nicol prisms (the analyzer) is inserted in the brass tube of the microscope imme- diately above the objective, and the other (the polarizer) is fitted beneath the stage of the microscope, so that the specimen examined on the slide stage of the microscope is now between two Nicol prisms, the lower one of which is the polarizer. Such an instrument will be found of assistance in distinguishing between certain starches, some of which polarize better than others, in detecting the addition of starchy matter (such as rice) to pepper or mustard (which do not polarize in the mass), and in distinguishing between pure butter and margarine. When, for instance, a specimen of pure pepper is examined, it is possible to obtain, by rotating the analyzer, a completely darkened field; whereas this is impossible when ground rice is the article under examination. Hence the addition of ground rice to pepper can readily be detected from the circumstance that it is not possible to obtain a completely obscured field. Similarly, with pure butter a completely dark field cannot be obtained, whereas with INTRODUCTORY NOTES 15 margarine from fat which has been melted it can; and in the case of mixtures it is impossible to completely obscurr; tiio field by rotating the analyzer. Graduated Burettes. In working with delicate standard solutions it is best to employ a mounted burette fitted with a stopcock at the bottom, rather than an unmounted one controlled by the finger; as in the former case the possibihties of contamination are reduced, and there is no risk of any loss from the burette while the operator is mixing or colour-matching in the intervals of the addition of further quantities of the standard solution. When a hand- burette is employed the index-finger which controls its delivery must be quite dry. Its: 10 9 FIG. 9. A BURETTE FILLED UP TO THE lO C.C. MARK. Unmounted burettes should not be blown out, but allowed to drain, and the drop at the dehvery end removed by touching the side of the vessel into which the contents are emptied. In judging the height to which fluid stands in a burette, always take the level of the convex lower border of the meniscus which forms upon its upper surface, and make this rest upon the Hne to which the fluid is required to reach. Water standing to the level of 10 c.c. in a burette will appear, therefore, as in Fig. 9. The eye must always be on a level with the upper surface of the liquid when a reading is made. The burette just holds 10 c.c. of water if at a temperature of about 15° C. the water weighs 9-99 grammes. Similarly, with a 100 c.c. measuring flask the graduation is correct if the 100 c.c. of water, at about 15° C, weigh 99-9 grammes. i6 LABORAtORY WORK For cleaning glass burettes, etc., and porcelain apparatus, especially from fatty matter, the commercial trisodium phosphate is useful. International Atomic Weights (1914). 0=i6. Aluminium Arsenic Barium Boron Bromine Calcium Carbon Chlorine Chromium Copper Fluorine Hydrogen Iodine Iron Lead Magnesium Manganese Mercury Nitrogen Oxygen Phosphorus Potassium Silicon Silver Sodium Sulphur Tm Zinc (Al) = 27-1 (As) = 74-96 (Ba) = 137-37 (B) = II'OO (Br) = 79-92 (Ca) = 40-07 (C) = I2-00 (CI) = 35-46 (Cr) = 52-00 (Cu) = 63-57 (Fl) = 19-00 (H) = 1-008 (I) = 126-92 (Fe) = 55-84 (Pb) = 207-10 (Mg) = 24-32 (Mn) = 54-93 (Hg) = 200-60 (N) = 14-01 (O) = i6-oo (P) = 31-04 (K) = 39-10 (Si) — 28-30 (Ag) = 107-88 (Na) = 23-00 (S) ^ 32-07 (Sn) = 119-00 (Zn) = 65-37 Weights and Measures upon the j\Ietrical System. The metrical system is founded upon the " metre," which is divided or multiplied by ten to represent different measures, as follows : Length. inVu part of a metre, ji- part of a metre. jV part of a metre. I millimetre I centimetre I decimetre I metre I decametre I hectometre I kilometre* 39-37 inches. 10 metres. 100 metres. 1,000 metres. * The Latin prefix therefore indicates division, the Greek multiplication. INTRODUCTORY NOTES Capacity. I cubic centimetre 28'35 cubic centimetres 1,000 cubic centimetres or I cubic decimetre I litre = 35-3 ounces I pint 1,000 litres O'oGi cubic inch. I fluid ounce. I litre. 1-765 pints. 568 cubic centimetres, 1 cubic metre. 17 One c.c. of distilled water at 4° C, and 760 millimetres baro- metric pressure, weighs i gramme, which is the standard of weight. Weight. I milligramme = I centigramme = I decigramme = I gramme = I decagramme = I hectogramme = I kilogramme = I ounce = I pound (16 ounces) = I gallon of water = I litre of hydrogen at 0° C, and O'oSgS gramme. I litre of oxygen at 0° C, and O'oSqGx 16 grammes. tttVo part of a gramme. jItj part of a gramme. ^^ part of a gramme. 15-432 grains. 10 grammes. 100 grammes. 1,000 grammes. 28-35 grammes = 437-5 grains. 453*6 grammes =7,000 grains. 4-536 litres = 10 pounds. 760 millimetres pressure, weighs 760 millimetres pressure, weighs Thermometer Scales. Centigrade. Reaumur Fahrenheit Freezing-point = o 3- Boiling point =100 Centigrade Reaumur Fahrenheit — 32 5 4 9 ■ To convert Centigrade to Fahrenheit, x 9-^5, and add 32. Fahrenheit to Centigrade, subtract 32, -=-9x5. Reamur to Fahrenheit, -f 4 x 9, and add 32. grains to grammes, x 0-0648. cubic feet to cubic metres, x 0-0283. cubic feet to litres, x 28-3. PART I THE CHEMICAL, MICROSCOPICAL, AND PHYSICAL EXAMINATION OF WATER FOR PUBLIC HEALTH PURPOSES CHAPTER I THE COLLECTION OF SAMPLES— INFORMATION REQUIRED AS TO SAMPLES— QUANTITATIVE EXPRESSIONS The sample should always be collected for analysis just as it is ordinarily obtained for drinking purposes. It is obvious, since our object is to discover all the possibilities of danger, that an endeavour should be made to ascertain the maximum amount of pollution to which the water is liable. For instance, in the case of streams, lakes, etc., the point of entrance of any drains, should only be avoided to the same extent as it is by those who may come to collect their drinking-water. When there is a general system of water-supply, an effort must be made to meet the same ends by choosing samples from the street fountains and street mains, rather than from storage, etc., reservoirs. But since impurities may gain access during domestic storage and distribution, it would not be fair in all cases to judge a public supply from the tap-water of any particular dwelling. With regard to shallow wells from which the water is removed by pumping, it is advisable to continue the process for some time, but no longer than it is judged that the water may be pumped during any one day, under the prevailing circumstances of demand. This is done because the last " pumpings " will often furnish the maximum evidence of any pollution present. To ascertain whether the water may have been contaminated during its domestic storage and distribution, the sample should be taken from the lowest draw-off tap (gener ally the scullery sink 19 20 LABORATORY WORK tap), as then the water will have run the maximum risk of contamination. When the fact is borne in mind that the water from many shallow wells is materially influenced both as to quantity and quality by the rainfall, it will be understood how samples from the same well may vary in purity according as to whether a long dry period may have preceded the collection, or a heavy rainfall, which may be the means of conveying to the well, water impreg- nated with surface washings, or water which may have washed accumulated impurities out of the interstices of the soil. These facts as to rainfall should always be ascertained ; and it may be desirable to examine a further sample after prolonged and heavy rain. The fact as to whether a cesspool or drain contaminates a well can readily be decided either by introducing a considerable quantity of sodium chloride, followed by plenty of water, and estimating the chlorine in the well-water every morning and evening for several days; or by introducing a strongly alkaline solution of fluorescine, and endeavouring to detect the green colour in the well-water. Water is customarily collected for analysis in a large glass- stoppered bottle, called a " Winchester quart," which holds about twice the amount which is implied by its description {i.e., about half a gallon). Stout wicker covers are made to protect them in transit by parcel post or rail. These bottles have become gener- ally adopted because, in addition to holding an amount which meets all the requirements of an ordinary analysis (even though it be necessary to repeat some of the estimations), they are strongly made; but obviously any stout glass bottle of similar dimensions, fitted with a glass stopper, will serve the same end. If a mineral analysis should be required, it is necessary to have quite 2 gallons of the water. It is well to avoid the employment of stoneware bottles. The bottle must be thoroughly cleansed by first well rinsing with a little dilute hydrochloric acid, and then by washing in good water until the washings are no longer acid. In collecting a sample the bottle is first quite filled with the water, and then emptied; it is again almost completely filled up (in a manner which will not favour the aeration of the water), and the glass stopper, having been found to fit accurately and tightly, is well rinsed in the water before it is inserted, when it is THE COLLECTION OF SAMPLES 21 tied down firmly on to the neck of the bottle and the knots are protected with seahng-wax. Care is taken to keep the sample cool and unexposed to light until the analysis is commenced; and under no circumstances should some of the estimations be un- necessarily delayed, as important chemical changes may occur — i.e., organic matter may suffer a very slight reduction, free ammonia may increase or decrease in amount, nitrates may be reduced or even increased, calcium or magnesium carbonates and iron, which were held in solution by carbonic acid, may, owing to the escape of the carbonic acid, be partially deposited. Therefore the figures of the two ammonias, the oxidizable organic matter and of the oxidized nitrogen, together with the physical characters, should always be ascertained as soon as possible (and certainly within forty-eight hours) after the sample has been collected. Information required as to Samples. It is often difficult to form a correct opinion upon the purity of a sample without the knowledge of some of the circumstances of its source; and if the water is held to run risk of harmful pollution this should suffice for its condemnation, although the chemical analysis at the time may prove satisfactory; natural agencies may suffice to purify water for a time, but there is always a possibility of their purifying powers being exhausted at any moment, and the danger of drinking such water is a constant one. Thus, information as regards the risks of pollution may be of great value as indicating possibilities of danger, when such danger may not be manifest at the time by analysis; it is also of value to ask, in every instance, the motive for requiring an analysis. Information bearing upon the constitution of the strata through or over which the water has passed is most valuable, since the soluble mineral constituents of certain strata are similar to those which may result from previous organic pollution. Anyone is able to furnish information as to whether the surface consists of such familiar substances as clay, gravel, sand, chalk, or vegetable mould, and whether the subsoil, exposed as it is by railway or road cuttings, is of chalk, sandstone, etc. It is very desirable that labels should be given to those col- lecting samples, and that these should be affixed to the bottle. The subjoined label, when filled in, would convey all necessary information to the analyst : 22 LABORATORY WORK Sample of Water for Analysis. Name and address of sender Place, date, and hour of collection Source of sample and method of collection If from well, give approximate depth and geological characters of the soil and subsoil of the district If from shallow well, give the rainfall during the previous week, in such terms as " nil," " small," or " great " in amount Nature and distance of any evident or possible source of pollution Reason for tlcsiring an analj'sis. The following is the usual form of report upon the chemical examination of a sample of water: Report on the Analysis of a Sample of Water received on. from and labelled . . . Number of sample . . Date of examination . . . . . . ■ . . Physical characters Reaction . . Saline and free ammonia Organic (or " albuminoid ") ammonia. . Oxygen absorbed from permanganate in two hours at 27° C. Chlorine Nitrogen as nitrates Total solid matter . . (a) Volatile (fe) Fixed Appearance on ignition Total hardness {a) Temporary.. (b) Permanent.. Poisonous metals Nitrites Phosphates Sulphates Microscopical examination of the sediment Opinion (Signed) Date SAMPLE OF WATER FOR ANALYSIS 23 Where a series of analyses are to be brouf^ht into comparison, the following form of report is to be preferred : Results of Analysis expressed in Parts i^kr 100,000. If I) Bi c 2 Organic or Albuminoid Ammonia. Oxygen absorbed in Two Hours at 27° C. .S 'u _o Nitrogen as Nitrates and Nitrites. Hardness. Solids. B V >> 2 a, 6 4) H c > T3 rt H The result of every analysis should be carefully entered in a book kept for the purpose, for such a record becomes most valuable for reference purposes and for making comparisons with future samples of water from the same locality. The results of the estimations made in water analysis are still variously returned in terms of grains per gallon and parts per 100,000. It seems desirable that uniformity of expression should be established. Parts per 100,000 is the most common return made in this country, and it is, moreover, in general use in France, Germany, etc. It is easy to convert grains per gallon to parts per 100,000, or vice versa. Supposing a report reads " chlorine 2-8 grains per gallon," how many parts per 100,000 will this represent ? There are 70,000 grains in a gallon. Therefore there are 2-8 grains in 70,000 grains, or 2-8 parts per 70,000 parts, or 4 parts per 100,000. It is only necessary, therefore, to multiply results returned in " grains per gallon," by ten, and to divide by seven, in order to convert them into " parts per 100,000," since grains per gallon are parts per 70,000; and so, to convert "parts per 100,000" to " grains per gallon," the returns must be multiplied by seven and divided by ten. Where the results of a quantitative test are returned in terms of " grains per gallon," it is convenient to work with 70 c.c. of the sample. The reason for this is that 70 c.c. represent "a miniature gallon," so called, and the results can at once be 24 LABORATORY WORK expressed in terms of an imperial gallon. The relation between the so-called " miniature gallon " and the imperial gallon depends upon the following facts: One c.c. of water is taken to weigh i gramme. Therefore 70 c.c. (" the miniature gallon") of water weigh 70 grammes, or 70,000 milligrammes. Therefore, since there are 70,000 component parts in each case, the milhgrammes in " the miniature gallon " are equivalent to the grains in the imperial measure. The various tests emploj^ed in the chemical and ph^^sical ex- amination of water for public health purposes have now to be considered. But it must be fully reahzed that organic pollution will give evidence of its presence in many of the steps which form a complete analysis, and that it is tliis collective evidence which determines the opinion to be formed, and not the evidence which any one special test may appear to offer. CHAPTER II THE PHYSICAL CHARACTERS OF WATER Whereas polluted shallow-well waters are notoriously often clear, sparkling, and pleasant to the palate, these characters are precisely those of our purest and best waters. For this reason, and from what follows, it will be seen that the evidence of purity furnished by the senses may be very misleading. Such physical tests, therefore, are not worthy of lengthy consideration. The Physical Characters. The sample is first well shaken, and then a thin, colourless glass tube, 24 inches long, is filled with the water, and from the appearance of this, in the " 2-foot tube," as it is called, the physical characters of clearness or turbidity, colour and the degree of aeration, are noted. 1, Clearness. — Though the best waters are always bright and clear, these qualities cannot be considered as evidence of purity, for a polluted well water may also possess them; and, on the other hand, any slight haziness or turbidity — which is, of course, furnished by minute particles of suspended matter — ^may by chemical and microscopical examination be proved either in- nocuous or harmful. Opacity may be caused by clay, iron, chalk, lead, or vegetable matter in suspension. It is difficult to satisfactorily estimate the amount of turbidity. Generally the transparency of water is measured by ascertaining through what depth of liquid a black-and-white figure can be seen with a given intensity of light. The degree of any such turbidity may be expressed as " very slightly turbid," " slightly turbid," and " turbid." 2. Colour. — To detect this the tube is fixed vertically so as to stand upon a white porcelain slab, and the observer looks down through the depth of the column of water on to the slab, which 25 26 LABORATORY WORK then forms a background. It is only in this matter that the faintest degrees of coloration are best appreciated, if even they be detected at all; and when thus examined it is rare that the water is not seen to possess a colour, however faint. In a good water it is generally of an extremely faint greyish-blue or greenish tint. Filtration of the water would show whether any colour is due to suspended or dissolved matter. The colour, which is most marked in water from reservoirs, lakes, and rivers, tends to a greenish hue in the spring and a brown in the autumn and winter. The various hues of j^ellow and brown will denote either the presence of animal or vegetable pollution {i.e., sewage or peat), or mineral contamination such as iron or clay; but if such colour is due to iron or clay a sediment will form. Clear waters from a depth which turn to a brownish-yellow colour on standing contain iron, the soluble ferrous salts becoming oxidized to insoluble ferric salts. The amount of sewage pollution must be 'enormous to furnish any colour, and a brownish tint in waters used for drinking purposes is generally due to peat or iron. A brown or red colour associated with turbidity and odour is often due to the growth of Crenothrix polyspora, a microscopic vegetable growth consisting of massed zooglceae and slender cobweb-like filaments. The organisms form a great number and diversity of spores, and hence its specific name of polyspora. The growth is rich in iron. A marked green denotes the presence of vegetable matter containing chlorophyll, which will generally be found to consist of the harmless algae. A blue-green tint in water, associated with a iloating bright green scum, has been found to be due to a species of Anabcena in association with a monad (Garrett). Red rain — in which the colour was produced by dust (prob- ably of cosmic origin) — fell in some parts of Europe in igoi. Rain water has also been coloured by volcanic ash and by desert sand; and " bloody " snow, produced by growths of Palmella sanguinea, has been described. A red colour has been found to be due to an alga named Oscillatoria rubescens. Colour alone affords no justification for condemning a water as unfit for drinking purposes until the nature of the material THE PHYSICAL CHARACTERS OF WATER 27 furnishing it is known; peat, for example, present to quite a harmless extent, will often colour a water markedly. The importance of the test does not seem to warrant any attempt at definite measurement when isolated examples are examined; but this is of service as a rough indication of the working efficiency of filtration processes in the case of some waters. If such a test is desired it is best performed by the method of Crookes, Odling, and Tidy. In this method an empty tube, exactly similar in every respect to one containing the water to be compared, is employed; this has two hollow glass wedges behind it, the one filled with | per cent, sulphate of copper solution, and the other with a mixture of ferric chloride (0-7 gramme per Htre) and cobalt chloride (0-3 gramme per litre), with a very slight excess of hydrochloric acid. These wedges are made to slide across one another in front of a circular aperture in a metal sheet; and thus any desired combination of brown and blue can be obtained. They are pushed over the empty tube until the colour, on looking down it, appears to be identical with that of the water-tube. Each prism is graduated from i to 50, the figures indicating millimetres in depth of the solution at that particular part of the prism, and the degree of colour is expressed as equivalent to so many millimetres of blue and so many of brown solution. 3. Taste. — The pleasant taste of good water is furnished by the gases dissolved in it; but since water must contain large quantities of any ingredient for its presence to be detected by the sense of taste, as an indication of dangerous contamination the test is useless. One-quarter grain to the gallon of iron will impart a faint chalybeate flavour, and this amount should not be exceeded in a drinking-water. Chloride of sodium (common salt) may be present in enormous quantities (80 to 90 parts per 100,000) without causing a brackish taste; and waters foully polluted with organic matters are often so palatable that they have frequently been preferred by the public to much purer waters. Certain vegetable growths (Anahcena and Tahellaria) may be the cause of unpleasant tastes. It is not advisable, as a rule, to taste samples sent for analj^sis. 4. Odour. — This is best detected by nearly filling a 200 c.c. glass-stoppered bottle (itself odourless), almost completely im- mersing this in hot water at about 60° C. for a few minutes, and 28 LABORATORY WORK then noting any odoni^ immediately on removing the stopper. The prehminary addition to the water of a httle strong potassic hydrate solution helps to make the test more sensitive. For most practical purposes, however, it will suffice to smell the sample after it has been thoroughly shaken; and it is only necessary to resort to the plan of heating the water when a sus- picion remains after this procedure, or when there is strong reason for considering that odoriferous gases may be present in small quantities inappreciable except when disengaged by heat. The test of odour is unreliable, and none may be evident in waters which are grossly polluted by sewage; it must be borne in mind, moreover, that many of the noxious materials which may gain access to a water have little, if any, odour originally. . The variety of odours which may be given off from a water defies description; many of them, though quite peculiar and distinct, it is impossible to describe, and any comparisons made with other odours are not always appreciated; but under any circumstances the analyst should attempt to describe the odour in his own words. Many odours are due to the growth of minute organisms, more especially algae. Anahcena and Tahellaria may give rise to un- pleasant odours, also Cryptomonas and other protozoa. To favour a large growth of algae the water must be stagnant and quiescent (as in ponds and reservoirs) in situations sheltered from the wind. Such terms as " musty," " horse-pond-like," " pig-odour," "fishy" "cucumber-like," "grassy," "earthy," etc., have been employed to denote the odour in certain pond waters. Some of these odours are related to essential oils, and are produced by living organisms; others result from the decay of the bodies of these organisms. A distinctly putrid odour is characteristic of large quantities of decomposing animal or vegetable matter, and a urinous odour is sometimes perceptible when large quantities of fresh sewage have gained access to water. The rotten-egg odour of sulphu- retted hydrogen and that of coal-gas are both peculiar and dis- tinctive. The presence of any of these odours would condemn the water, except in the case of waters from a depth naturally charged with sulphuretted hydrogen and free from animal pollution (as at Harrogate, etc.). The fishy odour is generally due to infusorians {Uroglena and THE PHYSICAL CHARACTERS OF WATER 29 other protozoa, Volvox, etc.), but it may arise from decomposing algae. Apparently it may also be due to small water-snails; and the introduction of trout to reservoirs thus affected has, by keeping down their numbers, led to a cure. Crenothrix, a fungoid plant which grows in the presence of protosalts of iron and decomposing organic matter, has often given rise (hke Beggiatoa) to disagreeable odour (SH2) and taste in public water-supplies, especially in the late summer and autumn. Small water-eels in water-pipes have, by their decomposition, occasioned evil odours in waters. Sulphuretted hydrogen may mask other odours; the addition of a little copper sulphate solution will prevent this and produce a brownish discoloration (due to CuS). As a rule the most frequent and objectionable odours are developed in surface waters. 5. Aeration. — Evidence of this is afforded by the amount of lustre the water possesses. Coarser degrees of aeration are noted by minute air bubbles collecting at the sides and bottom of the 2-foot tube, and rising up occasionally through the water to the surface. Waters from the chalk and limestone formations con- tain much carbonic acid, and little of any other gas; and where the water from these strata has been subjected to conditions of high pressure and low temperature, carbonic acid may even be in such quantities as to give a white turbidity to the water when this is first exposed to the air. The degree of aeration is of no value by itself in the estimation of the purity or impurity of a water, though a good water, to be palatable, must be well aerated. Many deep-well and spring waters, of great purity, are poorly aerated, and many foul waters are particularly bright and sparkling (chiefly from carbonic acid derived from organic decomposition). 6. Reaction. — This is important, not so much as affecting an opinion upon the wholesomeness of the water as from the circum- stance that it determines the effect which the water may have upon lead, iron, and zinc surfaces, and because in some steps of the analysis it becomes necessary to neutralize any acidity in the sample. Most waters markedly polluted with animal matter are decidedly alkaline from the carbonate of ammonia furnished by urine decomposition. The maj ority of pure waters in this country are faintly alkahne, the alkahnity being most generally given by calcium carbonate, and less often by sodium carbonate, etc. 30 LABORATORY WORK Some waters will be found to be neutral in reaction, and acid waters are bj- no means uncommon in this country; the latter most frequentl}' obtain their acidit}^ from the peaty acids (the humic, ulmic, and geic) taken up from the decaying vegetable matter encountered in their surface flow, and such acidity is a characteristic feature of so-called " peaty " waters. It is most marked after long periods of dr}^ weather. The reaction may be obtained by partially immersing in the water, pieces of delicate blue and red litmus-papers, and noting the change after a few minutes. If the water gives an acid reaction, it should be boiled, allowed to cool, and tested again. If the acidity has been lost, it was due to free carbonic acid. Any free acid in water is most gener- ally carbonic acid or organic acids, but it may also be sulphuric acid. The sulphuric acid may gain access to water from the oxidation of iron pyrites (sulphides) in the soil, or, like other acids, from the waste of factories; or in small amount from the sulphur in the coal burnt when rain-water washes the atmosphere over a town. Alkahnity that disappears on boihng is due to free ammonia. It is not generally necessary for hygienic purposes to test the degree of alkalinity or acidity of the water, but it is sometimes useful to make such an estimation, and the method is set forth in Chapter XVI. 7. The Sediment. — ^The presence of this, together with its macroscopic appearance, should be noted at this stage, but any opinion of its nature must be reserved until a microscopic and chemical examination have been made {vide Chapter VIII.). Certain waters deposit calcium carbonate or iron when allowed to stand exposed to the air; this is due to the escape of the carbonic acid which held these substances in solution. 8. Temperature. — Sometimes spring water from a great depth is warm or even hot. This is due to the fact that, below the level at which variations due to atmospheric alternations of temperature cease to be recognizable, the temperature of the earth increases with the depth when the measurement commences a few feet from the surface, the water temperature rising about 1° F. for every 50 to 60 feet of depth, on an average. The temperature of water may furnish a useful indication of its source as well as the depth from which it issues. CHAPTER III CHLORINE Chlorine exists in most waters as chloride of sodium, potassium or calcium. Rarely free chlorine has been found in waters polluted by industrial waste products. This combined chlorine is present in all waters to a small extent, even in rain water. In the rain water of country districts the chlorine varies from o-2 to 0-5 part per 100,000. The presence of combined chlorine in water in excess of this amount is due to one of the following causes : (a) The water has previously percolated strata which yield chlorides, such as greensand, sandstone, the London clay, chalk, etc. In the districts of salt deposits and in certain districts near the coast the well water may contain very large quantities of chlorides. (b) Pollution by animal organic matter, and chiefly urine (which contains nearly i, per cent, of chlorides). (c) Admixture with sea water, as in tidal rivers, and occasion- ally deep wells by the sea-coast. (d) Open reservoirs and other expanses of fresh water stored near the coast take up chlorides in appreciable amount from the atmosphere. It may also happen that large collections of rain water show a chlorine figure con- siderably above that in the rainfall of the district, even when they are not situated near the coast. This cir- cumstance is accounted for by the concentration which the water is always undergoing on account of the evaporation from its surface. {e) Effluents from alkali and other industrial works. Thus where it is possible to exclude sources {a), (c), and (e), the presence of chlorine in excess of i part per 100,000 may be taken 31 32 LABORATORY WORK to indicate organic contamination (past or present), and that of an animal nature, since vegetable pollution fer se furnishes no such excess; hence clilorine often serves to indicate whether a vitiated water is polluted by animal or by vegetable matter. It remains now to be seen, apart from other information which is available, what chemical means there are of excluding causes {a), (c), and (e) as furnishing chlorine in excess of the amount in pure rain and surface waters. When the excess is derived solely from the strata, there \\dll be no evidence of organic pollution furnished by the other steps of the analysis. If, on the other hand, the excess be due to animal pollution, there will be further evidence of this contamination in all those steps of the analysis which serve to indicate such pollution. The amount of chlorine originally taken up from animal pollution is not reduced by subsequent filtration of the water through subsoil or strata. Finally, if the excess of chlorine be due to admixture with sea water, there will also be present large quantities of magnesium salts, and this fact will at once indicate its source. Qualitative Test and Quantitative Estimation. Special Apparatus required : A white porcelain dish, A loo c.c. graduated flask. A burette graduated to iVths of c.c.'s. A glass stirring-rod. Special Chemical Reagents required : 1. a cold saturated solution of the yellow chromate of potassium, which must be free from chlorine ; this may be proved by acidulating a little of the solution by dilute nitric acid, and then adding a drop of nitrate of silver solution; in the absence of chlorine the solution will remain perfectly clear. 2. A standard solution of silver nitrate, made to the strength that I c.c. is capable of precipitating i milligramme of chlorine. This is made by dissolving 4-79 grammes of pure recrystallized silver nitrate in distilled water, and then making up to a litre. The solution should be kept in a brown-coloured bottle. The reason why 4-79 grammes of silver nitrate are required to the litre of water is as follows: CI (35-46) combines with AgNOj (i69-89). There- fore I part of chlorine combines with ( = ) 4-79 parts of AgNO^. \ 35'4" / Thus a litre of distilled water containing 4*79 grammes AgN03 will pre- cipitate I gramme of chlorine, and i c.c. will precipitate i milligramme of chlorine- GHLORlNit 33 Qualitative Test. The presence of chlorine (as chlorides) may be best detected by the addition of a few drops of a solution of silver nitrate and of dilute nitric acid to the water in a test-tube, when a white haze, turbidity, or precipitate of chloride of silver will appear, according to the amount of chlorine present : AgN03 + NaCl=AgCl + NaN03. Quantitative Estimation. 1. Measure out lOO c.c. of the water in a graduated flask, and pour into a white porcelain dish. 2. Add a few drops of the solution of yellow chromate of potassium until a distinct yellow colour is furnished to the water. The object of adding this reagent is to make it serve as an " indicator," which shall denote at once the stage when all the chlorine present in the water has combined with the silver employed in the estimation. 3. The burette is charged with the standard solution of nitrate of silver, and this is added drop by drop to the water; a dull reddish precipitate of the red silver chromate forms, which when stirred up in the water by means of a glass rod at once dis- appears, owing to the chlorine in the water displacing the chromic acid and itself combining with the silver to form a white pre- cipitate of the chloride of silver. As the addition of the standard solution is continued, the water, though it retains the yellow colour, becomes turbid, owing to the accumulation of this pre- cipitate of silver chloride. At length a point is reached at which, there being no longer any chlorine which is not already combined with the silver, the chromic acid holds undisputed possession of this metal, and the red silver chromate remains permanently present; the first evidence of this is afforded by the yellow colour changing into a permanent orange colour (K2Cr04 + 3AgN03 = 2KN03 + Ag,Cr04). Without the " indicator " there would be no means of knowing when just the amount of silver nitrate necessary to precipitate all the chlorine had been added; for it would be impossible to judge of the exact stage when the maximum amount of white precipitate of silver chloride had been created. 3 34 LABORATORY WORK 4. The first evidence of any red colour remaining permanent is the clue for withholding any further addition of the silver nitrate; or the amount of chlorine, estimated as it is from the amotint of the solution of the silver salt used, will be overesti- mated. Example. — Five c.c. of the standard solution of silver nitrate were required to combine with all the chlorine in 100 c.c. of water, and to furnish a reddish tint to the water. But I c.c. of the solution = i milligramme of chlorine. .•. 5 c.c. of the solution = 5 milligrammes of chlorine. .•. there are 5 milligrammes of chlorine in 100 c.c. of water. But 100 c.c. of water = 100 grammes = 100,000 milligrammes. .-. there are 5 milligrammes of chlorine in 100,000 milligrammes of water; or 5 parts per 100,000. Conclusions to he Drawn from the Amount Estimated. — It must always be borne in mind that chlorine can only be attributed to animal organic pollution when other figures of the analysis point to the probability of such an origin. Some pure chalk and red sandstone waters furnish chlorine up to 5 parts per 100,000. In the coal-measures and certain chalk and sandstone formations, low -lying and near the coast, the water may contain up to 50 parts of chlorine per 100,000; or, on the other hand, very little indeed. Pure deep-well waters from greensand deposits may y'\t\6. as much as 15 parts per 100,000, or more. Upland surface waters free from animal pollution rarely furnish more than i part per 100,000. Thresh is of opinion that as much as 100 parts per 100,000 of salt should condemn the water for drinking purposes. This amount imparts a distinct saline taste. Notes. — Some workers deduct o-i c.c. for the excess of silver solution required to indicate the change of colour. It is highly important that neither the water nor the standard solution of silver nitrate should be acid, or the results will be incorrect, since red silver chromate is soluble in an acid medium. In these cases the smallest quantity of precipitated calcium carbonate that will suffice should be added to effect neutrality. In estimating very small quantities of chlorine it is advisable to first concentrate the water before titrating, as the results are otherwise very slightly in excess. CHLORINE 35 Those who have not a keen appreciation of colour change may prepare a second lOO c.c. of water, to which a precisely similar amount of the chromate has been added; if this is placed alongside the water under examination, it serves as a comparison whereby to judge the commencement of the colour change. Coloured waters may be bleached in acid solution by means of potassium permanganate; the water is then neutralized and fil- tered before the chlorine is estimated. The chlorine is sometimes expressed in terms of the common salt it is equivalent to. This is readily calculated by a com- parison between the atomic weight of chlorine and the molecular weight of NaCl (common salt). The atomic weight of chlorine is 35-46, and that of sodium is 23-00; .-. the molecular weight of NaCl= (35-46 + 23-00)= 58-46. •••Cl=J|^ofNaCl. 58-46 Now, the chlorine in the example taken amounted to 5 parts per 100,000; then 5"^^^ NaCl, orNaC1^5^5^8..4. .-. there would be present 8-24 parts of NaCl per 100,000 if all the chlorine present were furnished by sodium chloride. Thus the weight of chlorine x ( ^ = j i -65 = the weight of sodium chloride. In the State of Massachusetts it is found that the chlorine in the surface waters and streams decreases in amount with the distance from the sea-board. The normal chlorine in waters free from all risks of pollution has been ascertained for each district, and these amounts are entered on a map of the State. Lines have been drawn connecting the districts the waters of which contain similar quantities of chlorine, and these are termed " isochlors." If the chlorine in any water is found to exceed the normal of the district from which it has been obtained, the pre- sumption is that the water is polluted with sewage. But a chlorine test is by no means delicate enough to indicate the lesser degrees of dangerous animal contamination. 3^ LABORATORY WORK The percentage admixture of sea water with fresh water may A-C be calculated from the formula x^ r — T\ ' '^^^^^ ;v=the number of volumes of fresh water to i volume of sea water; A=the chlorine in sea water (wliicli may be taken as 1,850 parts per 100,000); B= the chlorine in the local fresh water; C = the chlorine in the mixture of fresh water and sea water. CHAPTER IV HARDNESS The " hardness " of water is of economic rather than of hygienic importance, and the main object of the estimation is to decide whether the amount of hardness is such as to render the water unsuitable for washing, cooking, and trade purposes. A hard water entails in its use a great waste of soap, for considerable difficulty is experienced in procuring a lather (i grain of calcium carbonate will use up 8 grains of soap before a lather forms) ; it does not extract the same amount of strength from coffee, tea- leaves, and substances used for making soups, stews, and gravies, as softer water; and meat and vegetables boiled in it lose much of their flavour and colour, become slightly hardened and less digestible. On the other hand, moderately hard waters are always more palatable than very soft ones. It must not be thought, however, that " hardness " in a water is a factor which can be altogether disregarded from a health standpoint, for gastro-intestinal derangement, of a degree vary- ing with the constitution of the salts which form the " hardness," may arise among those who are constitutionally susceptible, and unaccustomed to a very hard water. The " permanent hard- ness " is generally mainly due to sulphates of the alkaline earths, and these have a marked aperient action, when they exist in large amounts. Such waters are obtainable at Epsom, Leam- ington, Scarborough, and Cheltenham. Finally, hard waters form a deposit on boilers and in pipes; and this is sometimes the cause of explosions and demands occasional removal. It is calculated that ^ inch of the incrusta- tion — which is a bad conductor of heat — requires the use of 45 per cent, of extra coal. For trade purposes generally — apart from the waste- of fuel, damage to boilers and danger occasioned by the " crust " from 37 38 LABORATORY WORK hard waters — it is of great importance to the process itself that the water should be moderate^ soft. The factors which commonly cause the total " hardness " in water are the following: Calcium and magnesium salts; iron, silica, and alumina; free carbonic acid, or free mineral or vegetable acids. The " total hardness " in most of the drinking-waters of this country is largely furnished by calcium and magnesium salts, and free carbonic acid; and more especially to calcium salts. Of the calcium and magnesium salts, the carbonates very greatly predominate as the cause of " hardness." These car- bonates of calcium and magnesium, almost insoluble in pure water, are held in solution b}^ carbonic acid, in the form of bicarbonates. If the water be well boiled, some of the salts forming the total hardness usually become precipitated, and being no longer in solution, they cease to add to the "total hardness"; the amount of hardness thus removed is termed " temporary," and that remaining " permanent." By boiling, the carbonic acid which held the carbonates of calcium and magnesium in solution is driven off, so that these salts precipitate. Any other con- stituent which may have been held in solution by the carbonic acid present, such as iron, would also be precipitated. Phosphate of lime, sihca, and the sulphate of lime (if present in large quan- tity) maj' also in part be precipitated. The " permanent hardness" results from what still remains in solution — i.e., calcium and magnesium sulphates, phosphates, chlorides and nitrates, an}' iron which was not held in solution by CO2, silica, alumina, etc. A little of the magnesium carbonate thrown down by the boiling will, moreover, become redissolved by the time the water cools, and thus may add to the " per- manent " hardness. Although the amount of mineral solids which the water con- tains generally forms an index to the extent of " hardness," 3"et this is by no means always the case; and some sahne waters yielding considerable quantities of mineral matter are " soft," a large quantity of sodium salts determining the softness. The salts causing temporary hardness tend to furnish a loose deposit in boilers; those causing permanent hardness, a hard deposit. HARDNESS 39 Quantitative Estimation, Special Apparatus required : A small glass-stoppered bottle of about 150 c.c. capacity. A burette with c.c.'s graduated to J c.c. A glass beaker. Filtering apparatus. Iron tripod, wire gauze, and triangle lined with pipeclay. Special Chemical Reagents : A standard solution ol potassic soap or of good undried Castile soap, made to such a strength that i c.c. will exactly precipitate either i milli- gramme of calcium carbonate or those other soap-destroying agents in the water to an extent which is equivalent to i milligramme of calcium carbonate. Fourteen grammes of Castile soap are dissolved up in a litre of a mixture of equal volumes of methylated rectified spirit and warm distilled water; it is then filtered and standardized (and, being unstable, should be re- standardized every few days) by means of a standard solution of calcium chloride. The calcium chloride solution is made by dissolving 0-2 gramme of pure crystaUized calcite (CaCOg) in dilute hydrochloric acid. When this is completely dissolved, evaporate to dryness on a water-bath; then add a little distilled water and again evaporate to dryness, and repeat this treat- ment several times to insure that all the acid has been driven off. The calcium chloride is then dissolved in a litre of distilled water. Such a solution will then contain the equivalent to 0*2 milligramme of calcium carbonate in every cubic centimetre, or 20 milligrammes per 100 c.c. The soap solution must then either be fortified by adding a little strong solution of soap, or weakened by a mixture of water and rectified spirit (in the proportion of 3 volumes of water to 5 of spirit), until the soap solution registers hardness equivalent to 20 milligrammes of CaCOg in 100 c.c. of the calcium chloride solution. Supposing the soap solution registers 19 milligrammes, then it is too strong, and must be weakened so as to register 20 parts — i.e., if the total filtered soap solution is 990 c.c, it must be made up to ta of 990 c.c.= 1,042 c.c. with extra water and rectified spirit. The rationale of the process is as follows: The soap employed is a combination of an alkali with a fatty acid. When it is added to water which contains calcium and magnesium salts in solution, then the fatty acids (oleic mainly in this case) will combine with the lime and magnesia to form insoluble calcic and magnesic oleate; and when the soap is added until there is no longer any lime and magnesia left to combine with, the fatty acids remain- ing in solution form a lather on shaking. Hence the more cal- cium and magnesium salts present, the larger the amount of soap 40 LABORATORY WORK solution required, and, in consequence, the longer is the produc- tion of a lather delayed. 1. One hundred c.c. of the water are placed \\dthin the small glass-stoppered bottle. 2. A graduated burette is then filled up to the lo c.c. mark with the soap solution, of which 2 c.c. are run into the bottle, when a cloudy precipitate of insoluble calcic and magnesic oleate, etc., is formed. The bottle is then briskly shaken to see if its contents will produce a lather. 3. The solution is afterwards added in cubic centimetres, and the bottle well shaken up after each fresh addition, until even- tually a certain definite amount of lather forms. The air should be sucked from the bottle (with a glass tube) from time to time, so as to remove any carbonic acid which has been liberated. Sufficient soap solution has been added when, with the bottle placed on its side, the lather presents a thin, unbroken surface after the lapse of five minutes. It is helpful to know that when the requisite quantitj^ of soap solution has been added, the con- tents of the bottle on being shaken give only a faint, dull, soft sound; and, after shaking, small particles of the lather cling to and slowly descend the sides of the bottle. 4. From the number of cubic centimetres of soap solution required, the amount of calcium carbonate, or its equivalent (in soap-destroying power), in the 100 c.c. of water, is ascertained. But a deduction of i c.c. from the amount of soap solution used must be made in every case, since this amount is required to create a similar lather in the same bulk of distilled water — which is free from an}' of the ingredients which are considered as furnishing " hardness." Example. — One hundred c.c. of water required 15 c.c. of the soap solution to furnish the characteristic lather. Deduct the i c.c. which would be required for 100 c.c. of dis- tilled water, and 14 c.c. of soap solution indicate the total hardness. But I c.c. of the soap solution =1 milligramme of calcium carbonate, or its equivalent. Therefore 14 c.c. = 14 milligrammes of calcium carbonate, or its equivalent. Therefore the " total hardness " in 100 c.c. of the water is equivalent to 14 milligrammes of calcium carbonate; and 14 milli- grammes in 100 c.c. (or 100,000 milligrammes of water) = 14 parts per 100,000. HARDNESS 4^ Conclusions to he Drawn from the Amount Estimated. — If the " total hardness " of a water reaches 30 parts per 100,000, it becomes unsuitable for washing and cooking purposes; and if it reaches 40 it is practically useless in these respects. A " soft " water may contain up to 10; a " hard " water from 15 to 25; a " very hard " water from 30 and upwards. Notes. — Where the hardness exceeds 25 parts per 100,000, so much precipitate of calcic and magnesic oleate, etc., is created that it interferes with the formation of a characteristic lather, and leads to an error of overestimation of the " hardness." In these cases it is necessary to dilute the water with an equal amount of distilled water — i.e., 50 c.c. of distilled water are added to 50 c.c. of the sample, and in the estimation of the hardness i c.c. is still deducted from the soap solution used. When the results are expressed in " degrees " upon Clark's scale, 1° (Clark) is equivalent i.n this country to i grain of calcium carbonate per gallon — i.e., to i part per 70,000. In France, however, a degree signifies i part of calcium carbonate in 100,000 and in Germany i part of lime in 100,000. The " total hardness " having been found, the next step is to ascertain the " temporary " and the " permanent " hardness. 1. One hundred c.c. of the water are measured out and poured into a glass beaker, which is placed on an iron tripod. To protect the glass against direct contact with the flame, the beaker is placed upon a triangle hned with pipeclay, which itself rests upon a piece of iron gauze. The water is boiled until only about two-thirds of its original volume remain. 2. The mouth of the flask is covered, and its contents allowed to cool, when all the calcium and magnesium carbonate, and often the bulk of any iron present, will be contained in the pre- cipitate noticeable at the bottom of the beaker. It is the super- natant fluid which contains the " permanent hardness," the " temporary hardness " which has been separated being repre- sented by the deposit. 3. From the beaker the cooled water is decanted into the measuring flask, care being taken to disturb the precipitate (which is left behind) as little as possible. The water is then made up to its original bulk by filling up to the 100 c.c. mark with recently boiled distilled water; or a reflux condenser may be used while the water is boiling. 4. The 100 c.c. of water is then filtered through a fine filter- 42 LABORATORY WORK paper and its " hardness " is estimated as previously described, and the result furnishes the " permanent hardness." 5. If the " permanent hardness " be subtracted from the " total," the difference will represent the hardness separated by the boiling — i.e., the " temporary hardness." Assuming that the permanent hardness is represented by (7- i) 6 c.c. of the soap solution, it is thus equivalent to 6 parts per 100,000 CaCOg. The total hardness was 14 parts per 100,000; .-. the "tem- porary hardness" =1^ — 6, or 8 parts per 100,000 of CaCOg, or its equivalent. Notes. — If it is desired to know the proportion of hardness due to magnesium salts in a water, where the " total hardness " is known to be due entirely to calcium and magnesium salts, it is necessary to first precipitate and remove all the calcium salts in the manner described in Chapter VI., when the hardness remain- ing will be due to magnesium salts. Wanklyn has pointed out that whereas lime reacts imme- diately upon the solution of soap, magnesia requires the lapse of time; and that one equivalent of magnesia consumes as much soap solution as one and a half of lime. If magnesium salts contribute materially to the hardness, a thin, fine, dirty scum, somewhat similar to a lather, forms upon the surface of the water as the soap solution is added. This scum finally breaks up, and is replaced by the genuine pure white lather. In such cases the water must be diluted considerably with distilled wate.. Bearing in mind the longer time taken for magnesia to react, the presence of this film or scum will warn the operator that as he adds the soap solution he must proceed slowly and shake thoroughly. In Clark's softening process lime is added, in quantity depend- ing upon the amount of carbonic acid in the water, in order that it may combine with this acid which holds the calcium and mag- nesium carbonates in solution (CaC03,C02 + CaH202=2CaC03 + H2O). When the lime is added in excess, some of it remains in solution in the water in an uncombined state, and since this is undesirable in drinking-water, a water treated by Clark's process should be frequently tested for uncombined Hme. A ready and simple method of detection is by adding a few drops of a solution of silver nitrate to some of the water, when, if free lime be present, the cloudiness created, instead of being white and clean HARDNESS 43 (silver chloride), becomes dirty and brown (an oxide of silver being formed). The Rivers Pollution Commissioners in their Sixth Report give the following classification of waters as to their softness: (i) Rain water; (2) upland surface water; (3) surface water from cultivated land; (4) river water; (5) spring water; (6) deep- well water; (7) shallow- well water; and they found that the follow- ing formations almost invariably furnish hard waters: (i) Cal- careous Silurian; (2) calcareous Devonian; (3) mountain lime- stone; (4) calcareous rocks of the coal-measures; (5) new red sandstone; (6) conglomerate sandstone; (7) lias; (8) oohte; (9) upper greensand; (10) chalk. CHAPTER V THE POISONOUS METALS Those poisonous metals for which it is commonly necessary to test a water are lead, iron, and zinc. Water most generally takes up these metals either from pipes through which it has been made to flow, from receptacles in which it has been stored, or from materials used in making or repairing the joints of pipes or cisterns; but, in addition, such metals may gain access from trade processes carried on by river- sides, or from metalliferous mines within the district, or, in the case of iron, from ferruginous soil or strata. Lead may be taken up from the pipes and cisterns made of this material. The action of water upon this metal is primarily an oxidizing one, and in the presence of dissolved oxygen a loose coating of oxj^hydrate of lead may form. The lead oxide is practically insoluble in those waters which do noi contain some free acid, but when this is the case (as notably in peaty waters and those containing much free COg) the lead salt is carried away in solution ; in other cases a relatively small quantity is removed in suspension. Ackroyd finds that plumbism due to the solvent action of peat}^ waters does not occur when the acidity of the water is equivalent to less than o"5 part of sulphuric acid in 100,000 parts of water (phenolphthalein being used as indicator). Plumbo-solvency is diminished by the presence of carbonates, sulphates, and chlorides in water which is not acid, but nitrates favour the oxidation of the metal to oxyhydrate. The action of the above salts is ascribed to the varying solubility of the lead salts of the corresponding acids, the nitrate being the most soluble, and the sulphate and carbonate the least so. Since non- acid waters containing carbonate of calcium provide a coating of carbonate of lead to the surface of the metal, and this coating is insoluble in such waters, unless there is free COg over and above 44 THE POISONOUS METALS 45 the amount necessary to form bicarbonate, it follows that soft waters are the great lead-carriers. Soluble phosphates in the water will also protect the metal to a marked degree. As a general rule, then, soft waters attack and hard waters protect lead; but in certain districts hard waters containing free COg, but a small amount of carbonate, are capable of dissolving appreci- able amounts of lead. " Soda-water " is particularly liable to take up large quantities of lead if it is allowed to come into contact with that metal. Waters containing a mere trace of lead often present to the trained eye a faint haziness. This disappears on adding nitric acid. Iron. — A chalybeate water generally contains its iron in the form of ferrous carbonate held in solution by an excess of carbonic acid; on prolonged exposure to air, or by apply- ing heat, hydrated ferric oxide, or " rust," is thrown down (4FeC03 + 02 + 2H20=2Fe203H20-f 4CO2), since it is insoluble in water containing no free acid. Upland, moorland and some other waters (as those from the greensand and new red sand- stone) generally contain traces of iron, which are taken up from the soil or strata permeated. The solution of iron from soils is generally due to organic matter removing oxygen, and thus converting the iron to the ferrous condition, in which form it is soluble in water containing carbonic acid. Copper is rarely found in drinking-water; but it is sometimes given to water by culinary utensils made of this metal, for a small amount of copper is dissolved when water which contains common salt, acid (vinegar, etc.), fatty or oily material, is boiled in contact with it. The writer has recorded an instance where the practice of placing a penny into the saucepan in which vegetables are boiled, in order to improve their colour, gave rise to symptoms of copper-poisoning in two children of a household. Zine is most generally taken up from galvanized iron cisterns and pipes or zinc surfaces. All kinds of water attack zinc in the presence of air ; even hard waters with an alkaline reaction. But it generally exists in water in the form of carbonate held in solu- tion by COg, and is not present in more than traces, except in soft acid waters. Galvanized iron must not be held to entail danger in its use, unless the water contains much free carbonic acid, since under common conditions the zinc oxide or basic carbonate form and protect the metal from further action. Zinc, as sul- 46 LABORATORY WORK phate, has been observed in considerable quantit}' in certain springs in the South of France, New Zealand, and America. Chromium. — Chromium may possibly get into water from colour and dye works, etc., but it is extremely rare that this very poisonous metal ever gains access to drinking-water. Tin, arsenic, and barium are rarely found in water; but man- ganese is not uncommonly found in some parts of the Continent. Qualitative Tests and Quantitative Estimation. Special Apparatus required : White porcelain basins. Boiling flask. Burette with c.c.'s graduated to tenths of c.c. Nessler glasses. Filtering apparatus. Ignition crucible and crucible tongs. Desiccator. Drying oven. Chemical balances. Wash-bottle. Marsh's apparatus. Special Reagents required : 1. A standard solution of lead acetate — i c.c. = i milligramme of lead — made by dissolving 1-83 grammes of crystallized acetate of lead in a litre of distilled water. 2. A standard solution of copper sulphate — i c.c. = i milligramme of copper — made by dissolving 3*927 grammes of sulphate of copper in a litre of distilled water. 3. A standard solution of ferric chloride — i c.c.= i milligramme of iron — made by dissolving 1-004 grammes of iron wire in nitro-hydrochloric acid, precipitating with ammonia solution, washing and redissolving the ferric oxide in a little hydrochloric acid, and then diluting to a litre. Each of these standard solutions may be diluted, in some cases with advantage, so that each c.c. contains o-i milligramme of the metal. Solutions of — Ammonium sulphide. Cyanide of potassium. Ferrocyanide of potassium. Yellow chromate of potassium. Ammonia. Ammonium chloride. Mercuric chloride. Peroxide of hydrogen. Dilute hydrochloric acid. Solid potassium nitrate and sodium carbonate. Granulated zinc. Metallic copper. THE POISONOUS METALS 47 Qualitative Tests. Though the metals lead, copper, and iron, even when existing in faint traces, may generally be detected by testing the original water, it is sometimes desirable to reduce the bulk of the water — and thus concentrate their solutions, as it were — by evaporation before testing. In the case of zinc, it will always be necessary to thus considerably reduce the original bulk of the water. A litre of water may be evaporated to 200 c.c. by previously marking a narrow beaker at the precise level to which 200 c.c. of water reaches, and, after acidulating with a drop or two of hydrochloric acid, in order to keep the metals in solution,* boil- ing the litre of water until it is reduced to this level. Lead, Copper, and Iron. — i. To about loo c.c. of the sample of water, placed in a white porcelain dish, apply a little of the ammonium sulphide solution by means of a glass rod which has been dipped in the solution. By drawing the rod gently through the water, and noticing any discoloured streak imme- diately adjacent to the track of the rod, faint quantities of poison- ous metals will be more readily detected than by allowing a drop of the reagent to fall into the water and then stirring. The reason for this is that the reagent itself imparts a slight colour, and it is therefore advisable to add as little of it as possible to commence with, otherwise a faint discoloration caused by a metal may be lost in that created by the reagent. Any evidence of a dark colour appearing in the water denotes the formation of the sulphides of either lead, iron, or copper, and the ammonium sulphide should then be further added until the maximum amount of darkening has been produced. It must be borne in mind that iron when in faint traces may only impart a slight dirty green colour to ammonium sulphide at first, but after a while the black colour of the sulphide forms. If there is any colour present in the original water, a com- parison must be made with a similar quantity of the water placed in another porcelain dish, before it is decided whether any additional colour has been furnished by the ammonium sulphide. Where, however, the colour originally present is marked (as by peat, etc.), it may well obscure, even with these precautions, a trace of lead, iron, or copper. It must then be decolorized as follows : 100 c.c. are acidiiied with hydrochloric acid and heated to * If the presence of lead is suspected, the water should not be acidulated. 48 LABORATORY WORK boiling; a crystal of sodium chlorate is added to the liquid, which is next boiled until the excess of chlorine is expelled. Ammonia is then added to the cold solution until it is just alkaline, and the whole diluted to its original volume with distilled water. 2. If the water darkens, pour half of it into a second porcelain dish. To one part add a drop or two of dilute hydrochloric acid, when if the colour disappears it is due to iron ; or if it diminishes perceptibly iron is present. 3. A confirmatory test should then be applied to some of the water in a test-tube — i.e., a drop or two of HCl is followed by a few drops of a solution of the ferroc^^anide of potassium, when the colour of Prussian blue (ferrocyanide of iron) is produced. Another very delicate test is to boil the water with a few drops of nitric acid, cool, and add a little potassium sulphocyanide; when a blood-red or sherry colour results, due to ferric sulphocyanide. 4. If, after adding the hydrochloric acid, the colour does noi disappear, the metal is either lead or copper. To the other half of the darkened water add a few drops of a solution of potassium cyanide; the PbS will be unaffected, but CuS will be completely dissolved. An excellent conlirmator}^ test for lead is to add to some of the water in a test-tube a few drops of a solution of the yellow chromate of potassium, when if lead is present an opacity appears in the water (due to the formation of lead chromate). The reaction, is, however, difficult of appreciation with faint traces of lead, which will often be missed by this test unless a careful comparison is instituted with another test-tube containing a similar amount of lead-free water and reagent. Iron may possibly be present along with lead, and may con- tribute to the darkening created by the ammonium sulphide. If so, this may be detected by adding a drop of dilute hydrochloric acid to the water, for this has been seen to remove any darkening furnished by an iron salt. Or the iron may be separated by adding nitric acid, evaporating to a small bulk, and precipitating the iron with excess of ammonia and warming; the precipitate of ferric oxide may be separated on a ftlter-paper, washed, dis- solved in nitric acid, then reprecipitated with ammonia, and again filtered and washed; the filtrate should be boiled until the ammonia is driven oft", and then tested for lead. 5. As a confirmatory test for copper, a drop or two of a solu- tion of the ferrocyanide of potassium should be added to some THE POISONOUS METALS 49 of the water after it has been acidulated with a drop of dihite hydrochloric acid. If copper is present, a bronze coloration and precipitate of cupric ferrocyanide appears. A faint colour will often be missed, unless it be looked for through the depth of the water on to a white background. 6. When no darkness is created and it is judged desirable to test for zinc, concentrate the water; render slightly alkaline with ammonia ; add a few drops of ammonium chloride solution ; then boil; add to some of the further concentrated water, after filtration, a few drops of ammonium sulphide. The white pre- cipitate (of hydrated sulphide) formed in the presence of zinc when once seen will always be recognized, since it is of a floc- culent, curdled, or gelatinous nature. As a confirmatory test, render the water slightly alkaline with ammonia; further concentrate by boiling; filter; add a few drops of the ferrocyanide of potassium with excess of dilute hydrochloric acid, and note a white gelatinous precipitate of zinc ferrocyanide, insoluble in dilute acids. Potassium ferricyanide furnishes a rusty yellow precipitate of zinc ferricyanide, soluble in hydrochloric acid and ammonia. 7. The presence of arsenic has extremely rarely to be tested for, but when it is desirable to do so Marsh's test is the most delicate. A litre of water is rendered alkaline by solid sodium carbonate (free from arsenic); evaporated nearly to dryness; and the residue introduced into Marsh's apparatus. For a full description of the application of Marsh's process, see "Arsenic in Food." 8. Tin. — A litre of water should be evaporated to a solid residue, and the tin dissolved out from the ash by warming mth a little strong hydrochloric acid ; then dilute a little and boil for a long time with metallic copper to make certain that the tin exists in a stannous condition; decant and add excess of a solu- tion of mercuric chloride, when a silky-looking cloud of mercurous chloride appears (2HgCl2 + SnCl2= SnCl4 + 2HgCl). If a mixture of ferricyanide of potassium and ferric chloride be added to a solution containing stannous oxide or chloride and hydrochloric acid, a precipitate of Prussian blue results from the reduction of the ferri- to the ferro-cyanide. If no other reducing agents are present, this is very delicate. Sulphuretted hydrogen yields a dark brown precipitate with stannous salts, soluble in potassic hydrate. 50 .LABORATORY WORK 9. Chromium. — A good test is to collect the residue from a litre of water: fuse the ash with solid potassium nitrate and sodium carbonate, so as to produce the yellow chromate of potassium; tliis in a neutral solution yields a purple precipitate with excess of silver nitrate. Slight traces may be detected by concentrating the water to a very small bulk, and then letting it drop upon a thin layer of ether floated on a dilute solution of peroxide of hydrogen acidified with sulphuric acid; the blue colour that forms in the lower solution passes over to the ether upon slight agitation. 10. Manganese. — Occasionally this metal is found in water, and its presence has been noted more particularly in America and German^^ A delicate test (Wanklj-n) is to evaporate a litre of water to a small bulk; nearly neutralize with hydrochloric acid; and treat with a few drops of peroxide of hj'drogen solution, when a brown precipitate forms in the presence of manganese. Having thus detected the presence of a poisonous metal, it becomes necessary to estimate its amount. OUAXTITATIVE ESTIMATION. The estimation of lead, copper, and iron may be performed by a colorimetric or colour-matching process. 1. Measure out 100 c.c. of the concentrated water which has been found to contain lead, and pour into a Nessler glass (a glass cyhnder graduated to 50 c.c). 2. Place a similar amount of lead-free water into three other Nessler glasses, to which different amounts (from o-i to i-o c.c.) of a standard solution of the metal have been carefully added. 3. To each glass add one drop of the ammonium sulphide solution, and well stir with a glass rod reserved for each basin. 4. Note which of the standard waters forms a match with the water under examination, and therefore contains the same amount of Pb. Example. — The amount of standard solution added to the particular 100 c.c. of distilled water which matched the brown coloration in the sample of lead-polluted water was 0*4 c.c. But I c.c. = I milligramme of lead. .-. o"4 c.c. = 0-4 milligramme of lead. .-. there is 0-4 milligramme of lead in 100 c.c. of the concen- trated water. But the water was concentrated to one-fifth of its original THE POISONOUS METALS 51 bulk; .-. the 0-4 milligramme of lead represents 0*08 milligramme in 100 CO. of the original water, or o-o8 part per 100,000, or 0-056 grain per gallon (about y\ grain). It is often better to work with a weaker standard solution made up to one-tenth of the strength of the solution here referred to ; and as it is only very light shades of brown which can be matched with great precision, if there is much metal present it may be necessary to lessen the degree of concentration, or even to deal with the original water. Similarly, a quantitative estimation of copper and iron may be made by employing standard solutions of these metals. Iron may be estimated gravimetrically, if in appreciable amount, in the following manner, and the colorimetric estimation thereby checked: The ash from the residue of 500 c.c. of the water is digested in strong hydrochloric acid; after filtration, the filter-paper and its contents are well washed with boihng distilled water; then add two drops of nitric acid to the filtrate, boil and add ammonium chloride solution and a slight excess of ammonia. Collect the precipitate on a Swedish filter-paper; wash with boiling water; dry in hot-air oven at 105° C. The filter-paper should next be folded up, placed in a small porcelain crucible (previously weighed) and covered by a lid; then ignite to dull redness, at first gently so as to obviate spurting and loss, and the lid should be removed after a while so as to permit free access of air. When the filter-paper has been entirely destroyed, let the capsule and its contents cool under a desiccator, and weigh. The weight found, minus that of the crucible and the ash of the filter-paper, represents the weight of FegOg, and this x 07 = Fe. The quantitative estimation of zinc can be conveniently made, gravimetrically, by taking a measured quantity of the concen- trated water (which is found to be free from other poisonous metals) and collecting the precipitate of zinc sulphide (obtained as described above) on a filter; this is then well washed with dilute ammonium sulphide, dried, ignited in a weighed capsule at a bright red heat, allowed to cool, and finally weighed as oxide (ZnO). The weight obtained x 0-8 = Zn. Or an approxi- mate estimation may be made by preparing a standard solution of zinc (4-4 grammes of the sulphate to i litre of water; i c.c. = I milligramme Zn), and matching, on the lines indicated above, the turbidity produced in 100 c.c. of the concentrated water 52 LABORATORY WORK after a small measured quantity of potassium ferrocyanide has been added. For the determination of small quantities of manganese in drinking-water, the colorimetric method of Volhard and Tread- well, modified for drinking-water, is recommended — that is, oxidation of manganese to permanganic acid \vith nitric acid and lead peroxide, and comparison of the coloured liquid with nitric acid solutions containing known amounts of permanganate. A very approximate quantitative estimation may be made, colorimetrically, as follows (Haas) : One hundred c.c. of the water under examination are acidified with 5 c.c. of 20 per cent, sulphuric acid, i gramme of potassium persulphate is added, and the solution is heated until a reddish- violet coloration is obtained, or until a brown colour, due to manganese dioxide, develops. The solution is now cooled, a trace of sodium hydrogen sulphite is added, and the oxidation with persulphate is repeated. The coloration obtained is then compared with that exhibited by j§^ potassium permanganate solution. Co7iclusions to he Drawn from the Amount Estimated. — Opinion is somewhat divided as to the amounts of the poisonous metals which may be considered dangerous in drinking-water. In the case of lead, ^ of a grain to the gallon is accepted as the limit by many authorities ; for in the case of a poisonous metal which is cumulative in its action, a very small trace should be con- sidered sufficient to render condemnation of the water justifiable. There is evidence that the system becomes habituated to copper salts, but this metal, if allowed at all, certainly never ought to exceed yV grain to the gallon in drinking-waters, for it is cumulative — though to a less degree than lead. Zinc rarely exists but in traces; and since it is not a cumula- tive poison, the possibihty of danger from this metal is remote. Therefore with regard to zinc, a trace (not exceeding yV grain to the gallon) may perhaps be allowed, for waters containing such a trace have been drunk continuously without apparent harm; but the faintest trace of arsenic — an exceptionally poisonous and somewhat cumulative metal — would suffice to condemn the water. Iron is not harmful to the same extent; and since it gives indication of its presence when in such amounts as would make its ingestion undesirable, by imparting a distinct taste to the THE POISONOUS METALS 53 water, its powers for evil are small; for people will not, as a rule, drink water that is unpleasant to the palate. A quarter of a grain per gallon is an amount which is just appreciable by taste. Iron in a general water-supply should under no circumstances exceed I grain per gallon, as a slight excess of this quantity may after a time provoke dyspepsia, headache, etc., in some people. To determine the solvent action of a particular water on lead or zinc surfaces, a piece of the metal may be submerged for twenty-four hours in a known quantity of the water in question. CHAPTER VI CALCIUM AND MAGNESIUM SALTS— SILICA— SULPHATES - PHOSPHATES Special Reagents required : Solutions of — ■ Ammonium chloride. Ammonia. Ammonium oxalate. Sodium phosphate (saturated). Barium chloride. Hydrochloric acid. Nitric acid. Molyhdic Solution. — Dissolve i part of pure molybdic acid in 4 parts ot NH3 (S.G. 0-960); filter, and pour with constant stirring into 15 parts of nitric acid (S.G. 1-2); let stand in the dark for a few days; carefully decant, and keep in the dark. Magnesium Mixture. — Fifty-five grammes of crystallized magnesium chloride are added to 70 grammes of ammonium chloride, and the whole dissolved in i litre of 2^ per cent, ammonia. About 15 c.c. of the mixture should be used to precipitate o-i gramme P20g. Special Apparatus required : Filtering apparatus. Platinum dish. Desiccator. Ignition crucible and tongs. Glass beakers and stirring rods Water-bath and drving oven. Calcium Salts. The presence of calcium salts, which mainly exist as the bicar- bonate and sulphate in water, may be indicated as follows : Add a solution of ammonium chloride and sufficient ammonia to furnish a slight ammoniacal odour. If there is any opacity or precipitate (due to ferric hydroxide, etc.), filter; then add am- monium oxalate solution to the filtrate, when a white precipitate 54 MAGNESIUM SALTS 55 of calcium oxalate forms. The ammonium chloride serves to hold any magnesium oxalate in solution, as soluble ammonio- magnesium oxalate. For a quantitative estimation, a measured quantity of water (say 200 c.c. concentrated from a htre of water previously acidi- fied with a drop or two of HCl) must be thus treated and set aside in a warm place for several hours; the precipitate carefully filtered (until the filtrate is quite clear) through a Swedish filter- paper; the oxalate of calcium precipitate remaining on this filter- paper is thoroughly washed, with hot distilled water, and afterwards dried in the hot-air oven at a temperature of 105° C. ; it is then ignited, the capsule and its contents allowed to cool under the desiccator, and the weight taken. The weight found, minus that of the crucible and the ash of the filter-paper, repre- sents the weight of the calcium as carbonate — to which form the oxalate has been reduced by ignition. Magnesium Salts. These generally exist in water in the form of the bi-carbonate and sulphate, and chiefly in water collected from sandstone deposits in the neighbourhood of the coast and from the dolomite strata. If these salts exceed 10 parts per 100,000 they may cause dyspepsia and diarrhoea in those unaccustomed to the use of such waters. The presence of magnesium salts may be best ascer- tained by precipitating all the lime present in the water by means of a solution of ammonium oxalate, ammonium chloride and ammonia; filtering until the filtrate is perfectly clear and free from lime, as shown by ammonium oxalate solution furnishing no opacity; the filtrate, slightly acidified with hydrochloric acid, should next be concentrated by boiling, and a few drops of a saturated solution of phosphate of sodium added, with sufficient ammonia to create strong alkalinity ; the whole is well stirred up with a glass rod and then set aside for several hours when a crystalline precipitate of a double phosphate of magnesium and ammonium (ammonium-magnesium phosphate) is formed. In those cases where the magnesium salts are present only in minute traces no definite precipitate results, but the points where the stirring-rod has touched the glass appear as white streaks, readily soluble in hydrochloric acid. The above precipitate of ammonium-magnesium phosphate may for the purpose of a quantitative estimation be collected on 56 LABORATORY WORK a filter ; washed with dilute ammonia; dried; ignited at a red heat; and weighed when cold as pyrophosphate (MgaPaO,), to which the red heat reduces the double salt. MggPaO? x 0-219= Mg. The amount present can also be approximately estimated from the hardness wliich magnesium will create when a water, pre- viously freed from calcium salts, is tested by the soap solution. Supposing, for instance, 5 c.c. of the soap solution are required to satisf}' the hardness remaining in 100 c.c. of such water; deduct I c.c. (the amount of solution required to produce a similar lather in an equal bulk of distilled water) = 4 c.c. But I c.c. of soap solution = I milligramme of calcium car- bonate. .■.4 c.c. = 4 milligrammes of calcium carbonate. .'. the hardness due to magnesium salts in 100 c.c. of water is equivalent to 4 milhgrammes of calcium carbonate. But i part of calcium carbonate is equivalent to 0*56 part of magnesium carbonate (Wanklyn), since i equivalent of magnesia consumes as much soap as i| equivalents of lime; for CaCOg is to MgCOg not as their respective molecular weights {i.e., 100 to 84), but as 150 : 84, or as I : 0-56. .-. the magnesium would be equivalent to 4 x 0-56= 2-24 parts of magnesium carbonate in 100,000 of water, or 1-56 grains per gallon. Magnesium carbonate has been estimated as high as 9 grains per gallon by Wanklyn and Playfair, in a Sunderland water. Silica. The estimation of silica may become of importance, having in view the fact that its presence diminishes the plumbo-solvent action of water. For the quantitative estimation a measured quantity of water, say 500 c.c, is slightly acidulated with hydrochloric acid, and then evaporated to a solid residue; this is treated with strong hydrochloric acid, and afterwards well washed with boiling distilled water; the residue collected on a filter is dried, ignited, and again treated with the acid and washed as before; any residue ultimately left will consist of most, if not all, of the silica origin- ally present in the water, and the white gritty powder of siUca (SiOa) may be dried, ignited, and weighed. SULPHATES 57 Sulphates. Sulphates exist in most waters, especially those which have been in contact with selenitic* deposits. They are either derived from the soil or strata over or through which the water has passed, or from the sulphur contained in organic pollution (urine, etc.). The rain water collected in large towns yields small amounts, originally derived from the sulphur in the coal burnt. Sul- phates in water sometimes result from the oxidation of metallic sulphides (chiefly iron pyrites), which exist as such in certain deposits. They mainly exist as calcium and magnesium sul- phates, and less generally as sodium sulphate; and either of these salts, if present in- large amount, would tend to cause diarrhoea and dyspepsia in those unaccustomed to the use of the water. Waters collected from the limestone and dolomite formations always contain a marked amount of sulphates. The sulphates in limestone may reach to 20 parts per 100,000, and consist mainly of calcium sulphate ; while those in magnesium limestones and dolomite consist partly of magnesium sulphate. Chalk waters are relatively poor in sulphates. The varieties of growth found capable of fixing sulphur and flourishing when sulphuretted hydrogen and sulphates are abundant in water (as in the case of gross sewage contamination and acid waste waters gaining access), are: Leptomitus lacteus, SphcBrotilus natans, Beggiatoa alba, and certain zooglea masses which may assume a branching appearance and simulate the other forms microscopically. Each form presents to the naked eye the appearance of long dirty white tufts, which are attached to stones, etc., in the bed and on the banks of streams below the level to which the water reaches. Qualitative Test.— To some of the water previousl3/ acidified with dilute HCl and placed in a test-tube, a few drops of chloride of barium solution are added; this is then left to stand for a few minutes, when an opacity or precipitate of the sulphate of barium is created with even very small quantities of sulphates. (H2S04+BaCl2=BaS04 + 2HCl.) For the quantitative estimation, 100 c.c. of water (concentrated from a litre) should be strongly acidified with hydrochloric acid, heated to boiling, and an excess of a hot 3 per cent, solution of * Selenite is a natural foliated or crystallized sulphate of lime 58 LABORATORY WORK barium chloride cautiously added, with constant stirring, until the maximum turbidity is furnished; the precipitate formed is collected on a small Swedish filter-paper, washed, ignited at a moderate red heat, and weighed as barium sulphate; the washing of the precipitate is continued until the filtrate no longer gives a turbidity with silver nitrate. If there is any doubt as to whether sufficient of the barium chloride solution has been added, let stand until the barium sulphate has settled, then add more barium chloride solution to the clear supernatant water, and note if any further turbidity occurs. If not, sufficient has been added. To express the result in terms of sulphuric acid (SO3) it is neces- sary to multiply the weight of barium sulphate by 0'343. In drinking-waters the amount of SO3, as sulphates of the alkalies and of magnesium, should not exceed 10 parts per 100,000. Phosphates. The phosphates found in water are commonly those of potas- sium, sodium, and ammonium, and their double salts. Their presence affords corroborative evidence of animal contamination (especially urine) ; but they may only exist in small amounts in waters dangerously polluted, for phosphoric acid is eagerly retained by the soil which the water may have subsequently come in contact with. When this point is considered in con- junction with the facts that traces of phosphates may also have their origin in strata — chiefly sandstone — permeated, and that they have also been found to be present in some marshy waters unpolluted with animal matter, it will be realized that they do not often furnish evidence of value to the analyst. But when in marked amount they may be taken as a certain sign of dangerous organic pollution. Their complete absence, on the other hand, is no guarantee of a water's freedom from such pollution. In every case before a test is applied for phosphates the water should be reduced from a large bulk to a ver}^ small one by evaporation, and it is even preferable to dissolve the phosphates out from the ash of the water. Qualitative Test. — Five hundred c.c. of water are acidified with a little nitric acid and evaporated to a solid residue ; the residue is dried over a water-oven for two hours, to render any silica PHOSPHATES 59 insoluble; then dissolve in 3 c.c. of dilute nitric acid and filter; mix the filtrate with 3 c.c. of molybdic solution, gently warm, and set aside for fifteen minutes at a temperature of about 26° C. If " traces " of phosphates are present, a faint greenish- yellow turbidity will be noted; if "heavy" traces, a marked yellow precipitate falls. The quantitative estimation may be made by comparing the colour results obtained, with standards made by diluting vary- ing quantities of a standard solution of sodium phosphate (i c.c. :=o-i milligramme of P2O5); or if the precipitate (which consists of yellow phospho-molybdate) is appreciable, it may be collected from 500 c.c. of water, washed with distilled water, dissolved in ammonia, and precipitated with magnesium mixture. This precipitate is then collected and washed with 2| per cent, ammonia, ignited, and weighed as MgaPgO, (magnesium pyro- phosphate). The MggPaOy X 0-64= P2O5. If this is the amount in 500 c.c, of water, one-fifth of this will represent parts per 100,000. More than 0-05 part of P2O5 per 100,000 should always be regarded with suspicion (Hehner). CHAPTER VII THE SOLID RESIDUE Special Apparatus required : Water-bath. Platinum dish. Drying oven. Crucible-tongs. Chemical balances. By " the solid residue " of water is generally implied the sub- stances which are held in solution, and which, when the water is evaporated to dryness, remain behind; and such a significance must be attached to the expression " total sohds " throughout Part I. of this book. The solubility of much of the matter taken up by water may be determined by soil micro-organisms, and the amount of sohd matter in water collected from a depth will depend upon the geological characters of the locality from which it has been collected. The Process. The suspended matters are first allowed to subside, or are separated by filtration either through a clean porcelain filter or through several large filter-papers, ribbed, and packed rather loosely into a large funnel. These filter-papers must be pre- viously well washed with distilled water. 1. Fifty c.c. of the water are placed in a previously weighed platinum dish; this is then put upon the water-bath, and it may be protected from dust by means of a small glass cover, one side of which is raised a little by inserting a glass rod beneath it, so as to allow the condensed moisture to escape. 2. When the water is evaporated to apparent dryness, the dish is removed and placed for half an hour in the hot-air oven, in 60 THE SOLID RESIDUE 6t order that the " solid residue " may be finally dried at 105'^ C. ; the object being to remove all adventitious moisture, but not the water which is an essential constituent of the substance as " water of crystallization." 3. The dish is removed from the oven, and then allowed to cool under a desiccator. 4. In fifteen minutes the dish and its contents are weighed, and the difference between the weight found and that of the clean and empty dish represents " the total solids " in 50 c.c. of water. 5. By means of a pair of platinum-pointed crucible-tongs the dish is next held in the flame of the Bunsen burner and slowly heated to dull redness, when any organic matter will give evidence of its presence by charring. If in small amount, this charring only causes an evanescent brown shade of coloration to spread over the residue; but if large quantities are present, the organic matter during incineration shows blackened specks or patches, which slowly disappear and give off dark fumes which may possess an odour of burnt hair or horn when due to nitrogenous animal matter, or of burning sugar if the material is vegetable. When a large amount of oxidized compounds of nitrogen exist, they may give rise to an evolution of red fumes of nitrogen dioxide. Marked scintillation is sometimes also perceptible — that is to say, tiny sparks are emitted. Eventually nothing remains but clear white or grey mineral ash, except where iron is markedly present and imparts a reddish tint to the ash. 6. The dish is allowed once more to cool under the desiccator and is reweighed; then the excess of weight over that of the clean and empty dish consists solely of mineral ash, and repre- sents the " non- volatile solids." 7. The weight of the total solids; less the weight of the " non- volatile solids " represents the " volatile solids." Example. — ^The clean platinum dish weighs 44-225 grammes. The dish + the total solids weighs 44-245 grammes. .-. 44-245 -44-225 = 0-020 gramme of total solids in 50 c.c. of water, or 0-040 gramme in 100 c.c. (100 grammes). .-. there is 0-040 gramme of total solids in 100 grammes of water Or 40 parts per 100,000 of total solids. After ignition the dish -j- contents weigh 44-240 grammes. .-. the " non- volatile soHds " in 50 c.c. water =44-240 -44-225 = 0-015 gramme. 62 LABORATORY WORK .-. there is 0-030 gramme in 100 grammes of water. Or 30 parts -per 100,000 of " non-volatile solids." Titus the total solids amount to 40 parts and the non-volatile solids to 30, and the difference of 10 parts per 100,000 will represent the volatile solids. Notes. — It may be pointed out that a few drops of dilute hydro- chloric acid will, by creating little or much effervescence, roughly indicate the amount of carbonates present, and will generally dissolve out everything but silica and the sulphates of calcium and magnesium. In the case of mineral medicinal waters and those used for some trade purposes, a detailed and complete analysis of the mineral ash ma}' be required; but for public health purposes, in view of the information which is acquired in other steps of the analysis, a complete anatysis of the ash is not necessary. This matter is therefore beyond the scope of this work ; but it may be stated that in regard to the principles which guide chemists as to the association of the different acids and bases to form the saline matter in w^ater, it is assumed that the combinations are governed by their respective affinities — that is to say, the strongest acid is assumed to be combined with the strongest base, due attention being also paid to the greater or less degree of solubility of the salts, since it is well known that this exercises a considerable influence on the manifestations of the force of affinity. Thus it is assumed that the chlorine is combined with sodium; any excess being allotted to potassium, or, in the absence of potassium, to calcium. If there is excess of sodium, it and the potassium are assumed to be in combination with sulphuric acid, any excess of which is allotted to calcium and magnesium; and calcium and magnesium, if not combined with sulphuric acid, nitric acid, or chlorine, are in the form of bicarbonates. Nitric acid is held to be combined with ammonia, and when there is excess it is allotted to either soda or lime or magnesia (as circumstances indicate) in waters which are found to be com- paratively free from organic matter; otherwise it may be allotted to fixed organic bases. The other constituents of the mineral residue, being in such small amounts, are generally not grouped as salts ; the silica, iron, alumina, nitrous acid, and phosphoric acid being returned as SiOg, FegOg, AI2O3, N2O3, and P2O5, respective!}-. THE SOLID RESIDUE 63 To give a simple example: A mineral water is analyzed and found to contain: Sulphuric acid Soda . . Magnesia Chlorine Lime . . Carbonic acid Parts per 100,000. 186-07 6672 5276 13-40 6-68 2-12 32775 These constituents would be expressed in combination as follows : Sulphate of magnesium 158-28 Sulphate of sodium . . 132-86 Sulphate of calcium 9-69 Chloride of sodium . . 22-11 Carbonate of calcium 4-81 32775 The ignited residue may be retained, as a routine practice, for the estimation of phosphoric acid. A loose white light residue indicates the presence of magnesium. Most good waters furnish a sohd residue, which, when ignited, shows no darkening ; but the sohds of potable peaty waters may show marked charring. Surface waters generally furnish from 5 to 20 parts of total solids, according to the nature of the surface ; well waters from 20 to 60, or over. The total solids have been estimated even above 300 parts per 100,000 in certain deep-well waters. A high amount of mineral solids, consisting as it so frequently does of harmless salts (such as calcium carbonate), is not neces- sarily injurious; but if a goodly proportion of the mineral residue is found to be contributed by sulphates, the water ma}- be productive of digestive disturbances; and, generally speaking, the mineral matter in a domestic water-supply should not exceed 100 parts per 100,000. * • CHAPTER VIII THE EXAMINATION OF SUSPENDED AND DEPOSITED MATTER IN WATER Chemical. — The estimation and examination of the " total solids " have only included those solids which are in solu- tion, but some samples of water contain suspended matter and sediment. The most simple method by which the matter which will deposit can be collected and esti- mated is the following: After well shaking, remove a litre of the sample; place it in a large conical flask; cover up and set aside for twenty- four hours; decant or syphon off as much of the supernatant water as it is safe to do without running any risk of disturbing the deposited matter; the deposit should then be shaken up with about 200 c.c. of distilled water, and after deposition that remaining behind in the flask should then be washed by distilled water into a platinum dish, dried, and weighed. The sediment may then be carefully in- cinerated at as low a temperature as possible, and the volatile and non-volatile matter esti- mated. Wynter Blyth's tube is a convenient instru- ment for collecting water sediments; as seen in Fig. 10, it is similar in appearance to a large pipette capable of holding a litre of water, A small glass cell fits over the small lower ex- tremity of the tube, and into this the deposited matter gradually collects. After the insertion of the long rod-shaped stopper, which plugs the outlet to the tube, the cell can easily be removed 64 fig. 10. wynter blyth's tube for collecting sediments. SUSPENDED AND DEPOSITED MATTER 65 without the sediment being disturbed. This may then be trans- ferred to a platinum dish, dried and weighed. Recently washed and ignited fine quartz sand may be cm- ployed to filter off the material for microscopic examination; all the filtered material from a large bulk of water thus collected may be subsequently washed out by a little distilled water and examined. Where a sediment forms from a sample of water advantage should be taken of the valuable evidence as to its nature which FIG. II. SHOWING THE SEDIMENT OF A POND-WATER, A SAMPLE OF WHICH WAS COLLECTED IN THE EARLY SPRING (X 250). DRAWN BY A. E. EVANS, M.B. I, A desmid; 2, Tabellaria floccosa (Diatomaceae) ; 3, actinophrys ; 4, a confervoid growth; 5, a vegetable spiral vessel; 6, silicious particles; 7, conferva; 8, gomphonema. Various forms of minute unicellular plants are seen scattered about the " field." a. microscopic examination affords. Some of the sediment maj^ be taken up by means of a small pipette, and transferred to several glass slides and cover-glasses applied, any excess of water upon the shde being removed by clean blotting-paper; or, better still, the sediment may be collected by means of centrifugalization, and the deposit from the centrifuge may be mounted and examined. When suspended matter is so light that it will not settle, drops of the water must be examined under the microscope. If a quantitative estimation is desired, the water should be 5 66 LABORATORY WORK allowed to rest for twenty-four hours, so as to exclude matter which will deposit, and the total solids in the supernatant water are then estimated; the solids remaining should be ascertained after the same water has been filtered through a Pasteur filter; and the difference in the two estimations will represent the amount of suspended matter retained by the filter. The various forms of animal and vegetable life (exclusive of bacteria), and of inanimate organic and inorganic matter, are best sought after by commencing with the ^-inch power, and next passing on to the :^-inch power. The presence of living organisms in abundance, bearing as they generally do a ratio to their food-supply, will be often sufficient in itself to condemn the water as containing a considerable amount of organic (mainly vegetable) pollution. The fact that animal and vegetable life have powers of purifying the water is beside the question; their very presence denotes impurity, and with the attainment of purit}? they mostly disappear. The higher and macroscopical types of animal life, such as water-fleas and other Crustacea, broadly speaking, denote less danger than the lower and' more minute forms (bacteria, amoebae, infusoria). The former are generally associated with suspended matter in waters that are not likely to be used for drinking purposes on account of their contamination being obvious to the senses, while the latter are often found to be associated with dissolved organic matter in waters which may possess excellent physical characters. More especially do large numbers of bacteria and the presence of fungi, infusorians, and anguillulse, suggest harmful pollution. But the most suspicious elements which may be detected by a microscopical examination of suspended or deposited matter are those which point directly to sewage contamination, and those which point indirectly to human contamination. The latter will be found in hair, wool, cotton and linen fibres, epithelial scales, etc. The former include objects which can rarely gain access to water save in actual sewage, such as: (i) Substances which from their indigestibihty commonly leave the body in the faeces; (2) substances which ma}^ do so when digestion is interfered with ; and (3) eggs, etc., of animal parasites which infest the human gastro-intestinal tract. Under the first heading would be em- braced such substances as — Various connective-tissue elements, fat globules and crystals, muscle fibres, starch cells, etc. Under SUSPENDED AND DEPOSITED MATTER 67 the second heading, shreds of mucous membrane, epithehal cells, gall-stones, particles of various kinds of foods in a semi- digested state, etc. The third heading would include T. solium, T. medio- canellata, Bothriocephalus laius (either as eggs or segments), Ascaris lumbricoides, Oxyuris vermicularis, Tricocephalus dispar (ova or mature forms), Paramoecium colt. It must not be thought that this direct evidence of human contamination is obtainable except on rare occasions. Matter which may be found by a Microscopical Examination of Water. 1. Inanimate. (a) Mineral. — ^This may be examined by the ^-inch power, and then chemical tests applied under the cover-glasses. Sand and flint particles generally have a sharp and angular c V FIG. 12. — VEGETABLE TISSUE. A, Vegetable parenchyma ; B, a pitted vessel ; C, a scaliform vessel ; D, a spiral vessel. outline, though often somewhat rounded from rolling and attri- tion ; a drop of hydrochloric acid let to run under the cover-glass has no effect upon them. In clay and marl (silicate of alumina) the particles are amor- phous and very minute, and unaffected by hydrochloric acid. Chalk particles are likewise amorphous, mostly somewhat larger than those of clay and marl, and generally rounded in outline. A drop of hydrochloric acid causes them to disappear with effervescence. Iron peroxide forms a yellowish-brown amorphous debris, soluble (Hke the chalk^particles) in hydrochloric acid, and blued 68 LABORATORY WORK when a drop of HCl and of a solution of ferrocyanide of potassium are allowed to run under the cover-glass. Mica forms thin, fine, scale-like films of very irregular outline, and insoluble in hydrochloric acid. {b) Vegetable. — Parenchyma; the dotted ducts and spiral vessels or spiral fibres; pieces of the cuticle with the vegetable " hairs " still adhering; pollen. Woody fibres; fragments of leaves, etc. ; starch cells; macerated paper; linen and cotton fibres; dark particles of soot. Vegetable matter often appears as dark flattened structureless particles, or frequently only as debris, when the classification of the deposit is attended with great difficulty; if, however, any spiral vessels, or fibres, or dotted ducts can be distinguished, these will always point to the probable nature of any obscure debris, etc., with which they are associated. (c) Animal. — Hairs; feathers; down; wool or silk fibres. Striped muscular fibres; fat globules and crystals; connective tissue ; epithelial scales ; shreds of mucous membrane. Scales and wings, legs, etc., of insects. Reddish-brown globular masses are sometimes found in asso- ciation with grosser sewage pollution. Their nature is somewhat obscure. The following particulars will serve for the microscopic identi- fication of textile fibres {vide Plate I.). Linen. — Cylindrical jointed fibres, with minute branching filaments at intervals. Cotton. — Flattened t\visted fibres, with no joints or nodes and no branching filaments or transverse markings. The apex of the fibre is blunt. Wool. — Rounded fibres, with fine cross-markings and indenta- tions on the border at the site of the cross-markings. A central longitudinal canal exists, but this is generally obliterated. The fibres stain yellow with a saturated aqueous solution of picric acid, warmed and then allowed to cool — not so those of linen. Silk. — Long cylindrical fibres with a well-defined central canal, but no cross-markings or indentations. The fibres dis- solve rapidly in concentrated H2SO4 — not so those of wool. Hemp. — Fibres faintly striated, with central canal, and trans- verse and oblique lines cross the fibres. (These are rendered more distinct by chloral hydrate.) A transverse section of fibres is irregular!}' rounded. The ends of the fibres are usually blunt. SUSPENDED AND DEPOSITED MATTER 69 Flax. — Closely similar to hemp, but the ends of the fibres taper to fine points. The fibres are polygonal in transverse section. 2. Animate. Vegetable. —Winnie forms of vegetable life mostly belong to the class of cryptogamous {i.e., non-flowering) plants, and con- tain chlorophyll. They may be divided into — I. Small and microscopic fungi, which represent some of the lowest forms of vegetable growths. These may be present as spores, sporangia, or my cell a. Both Bacterium termo and Sarcina ventriculi present familiar instances of these forms. Beggiatoa alba has been badly named " the sewage fungus," but any water containing a high amount of sulphuretted hydrogen or sulphate is capable of supporting this fungus, quite indepen- dently of the source from which such sulphur compounds are derived, and certain other growths are found even more frequently in association with sewage pollution. In Beggiatoa alba the sheathless cylindrical filaments contain roundish particles of sulphur, which are highly refractile to light. Leptothrix presents under the microscope a similar appearance to that of Beggiatoa alba, but the cylindrical cells, connected in threads and surrounded by a sheath, exhibit no sulphur granules. The empty sheaths may form large brownish deposits in water containing iron. Leptomitiis lacteus forms soft white or dirty tufts attached to stones or channels. In the absence of oxygen the growth darkens and putrefies. Under the microscope long branching filaments, constricted at regular intervals and bearing zoospores on the terminal segments, are seen. Sphcerotilus nutans is a similar growth, presenting under the microscope chains of long undivided filaments. Cladothrix may be found to be abun- dant in iron waters; the threads are composed of rod-shaped cells surrounded by a thin sheaf. An aquatic plant known as crenothrix gives much trouble at times, because of its tendency to develop in the water-mains and to clog water-pipes. Under the microscope this plant presents cylindrical filaments transversely divided into cells, and these filaments are surrounded by a gelatinous sheath which is coloured by a deposit of ferric oxide. The cells may, by division and production of viscous matter, escape and form zooglota. A 70 LABORATORY WORK drop of dilute hydrochloric acid let under the cover-slip dissolves up the iron, and enables the plant structure to be more clearly defined. Iron is requisite for its growth, and it is absent from waters which are but slightly ferruginous. It is often discovered in mains quite unexpectedly, and its long rusty filaments have been sometimes taken for horse-manure, with a consequent poor opinion of the character of the water-supply. Removal of the iron by oxidation and filtration is the best guarantee against the occurrence of this growth. Mucor, Aspergillus, and PeniciUium are moulds which may often be found in stagnant water. These are illustrated on pp. 263, 264, and 280. 2. Nitmeroiis forms of algcB, ranging in size from those visible only at high microscopic powers to those visible with the naked ej^e. Of these there are man}^ families: The volvocineae, of which volvox is the type, include the lowest vegetable forms of minute organisms; the oscillatoria exhibit a pendulum-like motion; the confervaceas are ver}^ numerous. Volvox glohator forms a green colony of cells. The colony is spherical, and the individual cells live in the wall of the sphere, each having two flagella and a nucleus. The ball, or hollow colom% continually rotates by means of the flagella of each indi\idual. This organism gives rise to a fishy taste and odour in water. Protococcus phwialis is an interesting instance of an algoid plant which can live in the atmosphere, and which may be found in rain-water. Raphidium and Scene desmus are not uncommon. Clathrocystis and Microcystis are widely distributed in surface waters. Of diatoms, Asterionella and Navicula are familiar types. The algae, when in considerable quantities, may furnish a dark green, repulsive appearance to the water, and may give rise to diarrhoea; when they die and decay the water acquires an offensive taste and odour. In winter comparatively few of such growths are found. In spring various diat-oms appear; these will disappear in a few weeks, and in their place will come the green algae; in the fall these will disappear, and the diatoms develop again — in turn to disappear with the onset of winter (Whipple). They {tahellaria, asterionella, etc.) occasionally grow in large numbers on the surface of stagnant water, or even on filter-beds. SUSPENDED AND DEPOSITED MATTER 7I Animal. — i. Protozoa, (a) Rhizopoda. .4 m«&a will be recog- nized by its characteristic amoeboid movement. AmrehcB coli are found in the mucus of the discharges of persons suffering from dysentery. Actinophrys, the body of which is surrounded by stiff radiating pseudopodia, is another common and famiUar form; and polyps shows a very low type of structure. In Spongilla fluviatilis (the fresh- water sponge) the animal substance is spread over a network of spicules; it grows in green masses. Difflugia has a coating formed by sand which it has taken into its body, and with which it makes a globular shell inside of which is the soft protoplasm of the body substance; this flows out at the mouth of the shell and forms anastomosing threads, which, acting as a net, catch food. Cercomonas and Euglena viridis are also common. Euglena is actively motile and green from chlorophyll. {b) Infusoria. Paramcecium, vorticella, and coleps are all common types. Stentor is among the largest of this class, and is so named from the trumpet-Hke shape of the body. It is covered with small ciha, and possesses long cilia in front. Coleps is also covered with ciha, and has an opening surrounded by short spines at both ends. Paramcecium is flattened and covered with ciha. Paramcecium coli are sometimes associated with diarrhoeal discharges in human beings, and Cercomonas intestinalis (a protozoan) may be present in the mucous discharges of children. Paranema is one of the simplest of the infusoria; it possesses one flagellum. Vorticella possesses a long contractile stalk. The body has a lid, and both the hd and the opening into the gullet are ciliated. 2. Ccelenterata. — Hydra is a common type of this sub-kingdom. 3. Annulosa.^ — This sub-kingdom embraces — {a) Crustacea, including the amphipoda,* isopoda,t and branchiopoda. % * The amphipoda are sessile-eyed malacostracans. Their bodies are compressed laterally, the eyes are immobile, and the feet are directed partly forwards and partly backwards. f The isopoda possess sessile eyes and a depressed body, and the feet are of equal size and move in the same direction. X The branchiopoda are so called because their branchias or gills are situated on the feet. The head is not distinct from the thorax, which is much reduced in size. 72 LABORATORY WORK Cyclops qiiadricornis, Gammarus pulex, and Daphnia pidex are familiar types of this class. In the latter the antennae act ,as oars and propel the little animal through the water by a series of short springs or jerks; the}^ assume a red colour in summer, and when in swarms they give a bloody tinge to the water. (b) Arachnida, including the microscopic tardigrada or " water- bears." (c) Insecta. Either in the larval, pupal, or adult forms. 4. Annuloida. — This sub-kingdom embraces the scolecida, and includes turbellaria, rotifera (or wheel-animalcules), tseniadea, naematoidea, anguillulse (water- worms). Rotifer vulgaris. — Gives a red or green colour to gutter-water. It is often called the " wheel " animalcule on account of its circular, oval disc, which is fringed with ciha; it is motile, and by its movements conveys an appearance of rotation; the cilia serve to propel the animal and to set up food currents. Small water-leeches {HirudinidcB) found in fresh water and on wet grass may be distinguished by their two suckers, one at either extremity; the mouth, armed with three teeth, is set in the middle of the anterior sucker. 5. Mollusca. — Including polyzoa, siphonida, etc. The various human parasites which may be conveyed through the medium of water are — The segments and eggs of tape- worms {Tcenia solium, T. medio- canellata, T. echinococcus, and Bothriocephalus latiis) ; the Guinea- worm {Dracunciilus medinensis); the round- worm [Ascaris lum- hricoides); the threadworm [Oxyuris vermicularis); Bilharzia hamatobia ; Ankylostomwn duodenale ; Tricocephahis dispar (the whip-worm); Filaria sanguinis hominis ; the filarial stage of Distoma hepaticimi (the liver-fluke of sheep). Several of these may be found in either embr^'onic or adult stages of development. The ova of bothriocephalus are developed in fishes, and man can only be infected by eating the latter. The larva of Ankylostomum duodenale is developed from ova in foul waters, and man may become infected by handling and drinking such water. The smooth oval egg of Ankylostomum duodenale possesses a seg- mented yoke when it leaves the female. After being discharged in faeces, under suitable conditions of moisture and temperature an embr3^o forms, which after a few days passes a sluggish exist- SUSPENDED AND DEPOSITED MATTER 73 ence usually in damp mud or earth. It may thus live for weeks or months, and if it gains access to the small intestine of the human being the worm reaches sexual maturity in about a month. The adult worm is a small nematode with a mouth furnished with four strong projecting hooks and two conical teeth, and the tail of the male has a large umbrella-like expanded bursa from which two long thin spicules project. Man is very rarely infected by the larvse of Distoma hepatictim. The ova develop in water into cihated embryos, and these undergo in small water-snails {Limnceus truncatulus) a further development to larvae; these ultimately change into little organisms (cercaria) resembhng tadpoles, which either remain encysted in water-snails or leave them and become attached by their suckers to grass. Bilharzia hcematobia. — The male is a grey flattened trematode worm, -|- inch in length, which inhabits the portal, splenic, FIG. 13. — -CILIATED EMBRYO AND CERCARIA FORM OF DISTOMA HEPATICUM (MAGNIFIED). mesenteric veins and the inferior vena cava of man and of some species of apes, especially in Africa and India; posteriorly the sides of the parasite curve towards each other and meet to form a channel (the gynecophoric canal), in which part of the long slender female (f inch in length) lies during fecundation. The spindle-shaped ova possess a beak, which generally projects from one end, but sometimes laterally ; these ova may be hatched before the parasite leaves the tissues of the original host, but the embryos are not born until afterwards. If the ova find their way into water, their walls swell up and rupture and the minute embryos escape, armed with cilia which serve to project them through the water. Apparently the embryo becomes attached to some fresh-water mollusc (or possibly fish), and develops into a cercaria form. Probably man is not infected by drinking water, but by protracted immersion in contaminated water, when 74 LABORATORY WORK the parasite enters the urethra or anus. The parasite is in- capable of reproducing itself within the human body. The embrj^o of Filaria medinensis, or the Guinea-worm, is also aquatic in habit — indeed, its first stages of development occur in a fresh-water crustacean. From these facts the inference that the parasite is transferred to man bj^ drinking water is justifiable. Of the Filaria sanguinis, the embryo of one of these {Filaria nocturna) has been traced through the body of the mosquito, and so into water, by which it may enter the human body — although it is usually inoculated directly by the bite of the insect. Tricocephalus dispar is also transmitted by water, for the egg develops only in water or upon some very damp medium, and the liberated embryo may thus find its way and attach itself to the mucous membrane of the caecum. There are other human parasites which infect their host fre- quently through the medium of drinking-water, although their life-histories do not even include a temporary residence in water. Thus the ova of Ascaris Imnhricoides and of TcBiiia echinococcus are often washed into water or blown into it as dust. The ova of the female ascaris are discharged with the faeces of the host, when, but not before, they are capable of furnishing embryos; these probably have an independent existence (possibly in water or some intermediate host, such as worms or insects) before again entering the human body and completing their development. It may be stated that, as a general rule, with one or two excep- tions — such as Oxyuris vermicularis, Trichina spiralis, and the tape-worms — contaminated water is the principal means by which the entozoa of man pass into his system. The ova of 0. vermicularis, unlike those of A. lumhricoides, contain embryos prior to their discharge, but probably these are incapable of further development until they have been passed with the faeces, when they may reinfect the same individual or others occupying the same bed, etc. ; or may pass into water, or get deposited upon vegetables and fruit and thus get ingested. The animal parasites of the lower animals also supply many instances of transmission by water. CHAPTER IX ORGANIC MATTER IN WATER Organic pollution may be of animal and vegetable matter; and since the danger of these two forms of organic contamination differs very materially (animal matter being far more dangerous than vegetable), it is important to learn the nature as well as the amount of any organic matter which is fouling water. Organic matter, as is well known, becomes, under suitable conditions of temperature, air, and moisture, resolved into simpler parts, by fermentation, putrefaction, and slow oxidation. As the ultimate result of these processes, the carbon appears as carbonic acid, the hydrogen as water, and the nitrogen as ammonia, nitric and nitrous acids. When putrefaction sets in, odorous gases are evolved, which mostly consist of compounds of sulphur. The effort to estimate the organic matter from the amount of oxygen of which it will deprive the permanganate of potassium was practised almost universally for a long period, and it remains as an auxiliary test for organic matter to this day. The facts that potassium permanganate in solution is so very unstable, that other substances in the water — apart from organic matter — are capable of reducing it, that it will part with its oxygen readily to the less dangerous {i.e., vegetable) as well as to the more dangerous {i.e., animal) pollution, and that the oxidizable organic matter bears an unknown and inconstant ratio to the total organic matter, all conduced to some dissatisfaction with the test, and endeavours have been made to find others which are of greater service. E. Frankland devised a most ingenious process to meet the want, but it is quite unsuited to the bulk of health officers, and there is scope for some error of experiment to creep in even with practised hands. In this process a measured volume of water 75 76 LABORATORY WORK is evaporated to a solid residue, and this is collected in a hard glass combustion tube, mixed with oxide of copper, and burnt in a. furnace. The oxide of copper parts with its oxygen to the organic matter, which is completely destro^^ed, and the carbonic acid, nitric oxide, and nitrogen which result are collected, measured, and expressed in terms of " organic carbon " and " organic nitrogen." A method superior in the facility of its execution, and equally as valuable for the purpose at issue, is that known as " the Wanklyn, Chapman and Hall Process." By it an endeavour is made, after computing the amount of " free and saline " ammonia originaUy present in the water, to estimate the amount of nitro- genous organic matter present from the amount of ammonia which can be derived from the breaking up of such matter by strongly alkaline permanganate of potassium at the boiling temperature. The organic matter which gains access to water is largely nitrogenous, and a very delicate indication of its presence and amount may be obtained from the nitrogen which it furnishes. Great importance is also attached in this process to the amount of the " free and saline " ammonia originally in the water, for it is, generally speaking, a product of recent animal contamina- tion. It will be recalled that one of the chief nitrogenous sub- stances in sewage is urea; and this urea, by the action of the Micrococcus urecR, is rapidly converted into saline ammonia, thus: CO(NH2)2 + 2H20= (NHJ^COg. It is obvious that no chemical process can determine as to whether the organic matter is Hving or dead, or whether in the former case it is harmful or not ; but while considerable quantities of the germs of disease cannot by themselves appreciably affect the amount of " albuminoid ammonia," since they always gain access to water along with other organic matter, this latter often furnishes by chemical analysis the evidence of danger. CHAPTER X WANKLYN'S PROCESS Special Reagents required : 1. A standard solution of chloride of ammonium, made to the strength that I c.c. contains o-oi milligramme of ammonia. The solution is made by dissolving 3-14 grammes of pure chloride of ammonium in a litre of distilled ammonia-free water; if some of this is diluted a hundredfold with distilled ammonia-free water, it is of the required strength. 2. Nessler's reagent. This consists of a saturated solution of the per- iodide of mercury in distilled ammonia-free water, the whole being rendered strongly alkaline with caustic potash. When this reagent is applied to a solution containing ammonia, it imparts a colour varying from a faint yellow to a reddish-brown, or even a precipitate, according to the amount of ammonia present. This reaction, which is due to the formation of ammonio-mercuric iodide (2(2KI,Hgl2)4-NH3-|- 3KHO=NHg2lH20-^ 7KI-I- 2H2O). is not shared by organic matter as such. The solution of the reagent should have an extremely faint yellow colour, which indicates that it is saturated with the periodide of mercury, and is therefore " sensitive "; should it be colourless and non-sensitive, this can be corrected by the addition of a drop or two of a saturated solution of corrosive sublimate. Any precipitate of mercuric iodide which settles should not be disturbed when the reagent is being used. The following has been found the best method of preparing Nessler's reagent: Dissolve 13 grammes of corrosive sublimate in about 250 c.c. of water, and 35 grammes of iodide of potassium in another 250 c.c. of water — by boiling; mix the two hot solutions, when a precipitate of the red per- iodide of mercury forms, which redissolves in the excess of iodide of potassium present; add a cold saturated solution of corrosive sublimate until a precipitate of red periodide just begins to remain permanently; raise to the boiling-point, and the precipitate will possibly be dissolved; allow the solution to become cold, or cool under the tap in a suitable vessel, and decant from any precipitate ; then dissolve 120 grammes of caustic potash in about 400 c.c. of water; and cool this solution; mix the two cold solutions, and make up to i litre of Nessler's reagent with water. Ammonia- free water must be used throughout. 77 78 LABORATORY WORK The reagent should be kept in a tight-fitting glass-stoppered store bottle, and small quantities emptied out into a smaller one for use from time to time. It is most sensitive after it has been kept for some time. 3. A strongly alkaline solution of the permanganate of potassium; which should alwaj's be boiled for a few minutes prior to use, in order to get rid of any traces of ammonia. The amounts recommended to be used in making up the stock solution are — Caustic potash, 200 grammes. Permanganate of potassium, 8 grammes. Ammonia-free distilled water, to i litre. Ammonia-free distilled water may be made by distilling tap water, after first fixing the ammonia present by the addition of a drop or two of dilute sulphuric acid ; the distillate from the bulk of a litre of the water can then be collected as " ammonia-free." As the previously ammonia- free water is very liable to take up traces of ammonia, it should always be carefully tested prior to use, and any trace of ammonia present must be distilled off. Special Apparatus required : Six Nessler glasses. These are narrow cylinders, each marked off at a point which indicates the level to which 50 c.c. of water will stand in them; they should be made of thin colourless glass, and of precisely siniilar diameter. Condensing apparatus, as shown in Fig. 14. The small tube perforating the stopper of the boiling-flask is seen to be surrounded in the greater part of its length by a larger tube. A constant circulation of cold water in the space between these tubes causes a condensation of the vapour which arises from the boiling water, this distillate being received into Nessler glasses. A white porcelain slab, 6 inches by 4 inches. A mounted graduated burette. A 2 c.c. pipette The Process. The amount of "free and saline ammonia " is first estimated — i.e., that ammonia which exists in solution in the water, or in combination with acids (carbonic, nitric, etc.), or in some other easily decomposable form. The Nessler reagent will create the faintest possible evidence of a yellow colour in 50 c.c. of the sample when this contains only a very small amount of " free ammonia." It is well to make it a practice to test the water in this waj' before com- mencing Wanklyn's process, in order to know whether the sample contains Httle or much " free and saline " ammonia. If more than a faint yellow tint forms, the water should be diluted, as it may otherwise be difficult to get the large amount of wanklyn's process 79 8o LABORATORY WORK ammonia over and to match it. For instance, in the case of extremely foul waters, the degree of colour due to the ammonia in the first 50 c.c. of distillate is too intense to be matched by the standard solution, for in many cases a copious precipitate appears and it is impossible to make a comparison. In such cases, smaller quantities of the original water should be diluted with an equal bulk, and sometimes even with five or ten times its amount, according to the depth of yellow colour obtained, of distilled ammonia-free water, prior to distillation. 1. The condenser is hable to become contaminated with ammonia. Therefore first distil some clean water through the apparatus until the distillate gives no colour with Nessler's reagent. 2. Five hundred c.c. {i.e., half a litre) of the water are placed within a boiUng-flask. If the water is acid or even neutral, a little pure anhydrous sodium carbonate should be added so as to ensure alkalinity. The motive for tliis is to enable the free ammonia to come away readily, since any acidity would exert a fixing influence upon it; it also decomposes ammonium sulphate, 3. The boiling-flask is then tightly connected to the con- denser, so that no uncondensed vapour can escape at this point. The Bunsen burner is next hghted, the flame applied to the flask, and rapid boiling is encouraged. 4. The water-tap is turned to such an extent that the water, after circulating in the outer tube of the condenser, flows in a small stream to the waste-sink. 5. A Nessler glass is placed so as to catch the distillate, and when sufficient of this is collected so as to reach up to the level of the 50-c.c. mark, a second glass is substituted, and then a third. 6. When three Nessler glasses are thus filled up to their 50-c.c. marks with distillate, a fourth is placed to catch more of the dis- tillate, while 2 c.c. of Nessler's reagent are added to each of the three glasses. If these glasses be placed upon a white porcelain slab from left to right in the order in which they received the distillate, the yellow colour furnished in each of them by the reagent will show a decrease in amount from left to right, since the first 50 c.c. collected will contain tlie most " free and saline ammonia," and the third the least. 7. The gas may be turned out and the distillation stopped if wanklyn's process ^^ there is no colour in the third Nesslcr glass, or if it be extremely faint, since all the " free and saUne ammonia " will then have come over. If, however, the colour is distinct in the third Nessler glass of distillate, a fourth must be collected andtested with 2 c.c. of the reagent, and even a fifth may be occasionally necessary. It is, of course, imperative that all the ree and saline ammonia " in the original 500 c.c. of water shall be re- moved, and it is seldom in a drinking-water that 150 c.c. ot distillate does not contain the whole of this. 8 The amount of ammonia must be determined by matching the'colour in each glass. To make this match it is necessary to take another Nessler glass, and to deliver into it by a burette the amount of ammonium chloride standard solution that is iudged necessary to effect the match; the cylinder is then filled up to the 50 c.c. mark with distilled ammonia-free water, and then 2 c.c. of Nessler's reagent are added. If the match is not correct, then a fresh comparison must be made with more or less of the standard solution, as the case may be. A very httle experience will enable the operator to effect this matching with great rapidity, . . Notes -It is pointed out by Wanklyn that it is not necessary to match each glass separately, since three-quarters of the total amount of " free and saline ammonia " is contained m the hrst glass of distillate. OccasionaUy, however, there is a famt dis- crepancy between the amount thus calculated and that obtained by matching each glass; this may be due to differences m the degree of alkaUnity of waters and in the rate of boihng. On this account, and for the reason that the statement does not hold true with very foul waters, it is preferable m every case to estimate the amount of ammonia in each glass. In every case when, after stirring with a clean glass rod, the colour in the comparison cyhnder is found to approach that m the distillate, the operator should cover up the cyhnder and wait about three minutes before adding more standard solution, since the colour deepens a Httle upon standing. The presence and degree of coloration must always be judged by looking down through the depth of the water on to a white slab and care must be taken that the bottoms of the glasses and the upper surface of the slab are perfectly dry, as a thm layer of intervening water diminishes materially the depth of colour, and leads to error in matching. 6 82 LABORATORY WORK 9. Having thus effected a colour match by placing the two glasses side by side upon the white slab under exactly the same conditions of light access, the amount of ammonium chloride solution which has been used to effect this is noted, and the ammonia which this is equivalent to will be the amount of the " free and saline ammonia " in the glass of distillate. Example. — One hundred and fifty c.c. of distillate were col- lected, and the last 50 c.c. are found to contain no trace of ammonia. The whole of the " free and saline ammonia " in the 500 c.c. of water was therefore collected in two Nessler glasses. It was necessary to add 3 c.c. of the standard solution of ammonium chloride to the comparison test-glass in order to match the colour in the glass containing the first 50 c.c. of distillate, and i c.c. of the standard solution was required to match the colour in the second 50 c.c. of distillate. The total amount, then, of " free and saline ammonia " in the 500 c.c. of water corresponds to the ammonia present in 4 c.c. of the standard solution of ammonium chloride. But I c.c. of this standard solution contains o-oi milhgramme of ammonia. .'.4 c.c. contains 0-04 milligramme of ammonia. .•. there is 0-04 milligramme of "free and saline ammonia" in the 500 c.c. of water (or 500,000 miUigrammes), .'. there is o-oo8 milligramme of " free and saline ammonia " in 100 c.c. (100,000 milligrammes) of water, or o-oo8 part per 100,000. 10. The next step in the process is to continue the distillation more slowly after adding 50 c.c. of the recently boiled alkaline solution of permanganate of potassium to the boiling- flask; to collect the distillate in three Nessler glasses; and to repeat the process of " Nesslerizing " precisely as before. The ammonia now obtained is called " albuminoid ammonia," since it is derived from the breaking up of albuminoid and other nitrogenous organic matter by means of the alkaline permanganate. It is important to remember that this albuminoid ammonia comes over more slowly and much less evenly (the second Nessler glass sometimes containing almost as much as the first), so that the first 50 c.c. of distillate must never be taken to contain three-quarters of the total " albuminoid ammonia." Example. — It was necessary to distil over 200 c.c. in four Nessler glasses before all the ammonia had come over. The wanklyn's process ' S^ fourth glass of distillate had colour equal to that furnished by 0-2 c.c. of the standard solution, the third to o-8 c.c, the second to 3 c.c, and the first to 2-5 c.c. .-. (o-2 + o-8 + 2 + 2-5) = 5-5 c.c. of the standard solution were required to match the colour furnished by the " albuminoid ammonia " in 500 c.c. of water. But each c.c. of the standard solution = o-oi mihigramme of NH3. .•. 5-5 c.c. = 0-055 milligramme of NH3. ,'. there is 0-055 milligramme of HN3 (" albuminoid ") in 500,000 milligrammes of water, or o-oii milligramme in 100,000 of water. Conclusions to he Drawn from the Amount Estimated. — In the case of contamination with animal matter the " free " ammonia exceeds the "albuminoid"; while vegetable matter furnishes " albuminoid " ammonia and practically no " free." Therefore much " albuminoid " along with a very small amount of " free " ammonia indicates vegetable contamination, and this indication gains further support if there is no excess of chlorides and of nitrates. Relatively high " free " ammonia along with " albuminoid " ammonia above 0-005, and excess of chlorine (and often of oxidized nitrogen) will denote recent animal pollution. If the " albuminoid " ammonia exceeds 0-005 part per 100,000, the " free " should be below this amount; but if the albuminoid ammonia is much below this, then a high figure of "free" ammonia is probably due to a reduction of nitrates, and not to recent animal contamination. Conversely, if there is practi- cally no " free " ammonia — i.e., 0-002 or less — then the " albu- minoid " ammonia may be allowed to exceed o-oi, as it is evident that the organic matter present is purely vegetable. It may be said that if the free ammonia in upland surface waters exceeds 0-002, a suspicion of animal contamination is warranted. "Free" ammonia, accompanied by practically no "albu- minoid," is found in the following circumstances: {a) The water has been in contact with a stratum contain- ing a reducing agent (greensand contains a reducing salt of iron) which has decomposed the oxidized nitrogen originally present in the water; or metal pipes, cement, etc., with which a well-water has come in contact may effect this reduction to a less extent. 84 LABORATORY WORK {b) Other waters containing iron frequently possess a marked amount of ammonia derived from the reduc- tion of nitrates, (f) The water has percolated a deposit in which some ammonia salt is present. (d) The sample is rain-water, collected in town districts, in which ammonia may exist in considerable quantities. Other steps of the analysis will serve to indicate the source of "free ammonia"; and where it is not derived from organic pollution the " albuminoid ammonia " \\dll always be very low indeed. If, in spite of the previous dilution, the ammonia in the first Nessler glass of distillate still furnishes too deep a colour to admit of a satisfactory match, the whole of the distillates con- taining free ammonia may be mixed together, and the ammonia in 50 c.c. of the light-coloured mixture Nesslerized and estimated; and from the amount found in this measured part the amount in the whole distillate may be calculated. Sometimes while extracting the " albuminoid ammonia " the contents of the boihng-flask boil too violently, and " bumping " ensues; to obviate this a gentle shaking of the flask will often suffice, but in default a few fragments of freshly ignited pumice- stone afford an excellent remedy. The foulest waters and those containing much saline matter are most apt to bump, and it is highly important to prevent this, since uncondensed vapour thereby escapes at the distal end of the tube, and sometimes some of the water is shot over from the boiling-flask, both of which occurrences obviously vitiate the results. When some of the water to which the alkaline permanganate has been added thus spurts over into the Nessler glass placed to collect the distillate, it is of course impossible to " Nesslerize," since the distillate has a pink colour. There is no alternative then but to pour back the distillate into the flask and renew the distillation. When the " albuminoid ammonia" comes over so slowly (as in some peaty waters) that almost all the water in the retort threatens to be used up, 200 c.c. of ammonia-free water may be added to the flask and the distillation continued. In those rare cases where " the free ammonia " continues to come over in small quantities, it is a good plan to adopt the measure (Rich) of starting the process afresh, " Nesslerizing " the first 50 c.c, wanklyn's process 85 and then returning the rest of the distillate to the flask, and redistilling it before " Nesslerizing." Strange to say, though urea is decomposed by the boiling with the alkahne permanganate, its decomposition does not yield any ammonia, and this at first sight would seem a grave defect in the process. When, however, it is considered that this is probably the only nitrogenous contamination of animal origm with which a water is hable to be polluted which does not, in the circumstances, yield ammonia, and that urea in urine naturally becomes very rapidly changed into ammonium car- bonate and as such is detected in the sahne ammonia, the matter is not one of importance. By Wanklyn's process only about one-half of the nitrogen in organic combination is liberated as " albuminoid ammonia " ; but it is not necessary that in the process the total nitrogen contained in organic matter should be evolved as ammonia, so long as that which is evolved gives an index which bears a fairly fixed and constant ratio to the total amount; so that from this index an empirical standard of purity can be formed. The process efficiently meets this requirement. When but httle water remains in the boiling-flask, the flame must be lowered, as the naked flame must not be allowed to play upon the glass above the water-level. The presence of considerable sulphuretted hydrogen in water interferes with Nesslerization ; this must therefore first be remedied before the free and sahne ammonia are distilled over and estimated, in the following manner: The ammonia should be fixed with 10 c.c. of normal sulphuric acid; then if 100 c.c. of the water are distilled over, this amount of distillate will contain all the sulphuretted hydrogen. The water remaining in the boihng-flask is then neutralized with 10 c.c. of normal sodic-hydrate, when Wanklyn's process can be performed. When it is found necessary so to deal with sulphuretted hydrogen, a blank experiment should be performed, by which any ammonia found in 500 c.c. of ammonia-free distilled water containing 10 c.c. of normal sulphuric acid and 10 c.c. of normal sodic-hydrate is distilled over and estimated, and this is deducted in arriving at the figure of the free and sahne ammonia in the sample. CHAPTER XI THE OXIDIZABLE ORGANIC MATTER— E. FRANKLAND'S PROCESS In the presence of organic matter the permanganate of potassium, under favourable conditions, will part with oxygen until all the permanganate has become reduced to hydrated manganese dioxide, as indicated by the loss of the original pink colour. While a certain proportion of the organic matter present in water is always oxidizable by the permanganate of potassium, this varies with the nature of the organic pollution, and it there- fore bears no constant ratio to the total quantity of such pollution present. Some forms of animal matter reduce less perman- ganate than others, and comparatively harmless peaty waters may absorb much more ox^^gen than waters dangerously polluted with animal matter. Despite these drawbacks, and the fact that in Wanklyn's process we possess the means of making a far closer estimation of organic matter, the test under consideration frequently furnishes corroborative e\ddence of value, but it is most service- able as a means of gauging the comparative purity of a series of waters, or of the same water from time to time. A two-hours' exposure of the water to the permanganate is quite short enough for the test to be of much value, since it is chiefly the putrescent, or very easily reducible organic matter, which is oxidized in the first half-hour. There is little or no advantage, however, in making the test extend to four hours. It must be clearly understood that even at the end of four hours the oxidation of the more stable organic matter by acid per- manganate would be incomplete; and so the two-hours' test is generally adopted for the purpose of obtaining a standard or figure for comparison. The conduction of the process at a precise temperature has 86 THE OXIDIZABLE ORGANIC MATTER ^7 been proved by experiment to be an important factor for the amount of oxygen'taken from the permanganate vanes con- siderablv at different temperatures. ,-, f If Jwater is bottled long before analysis, the quanWy o oxygen absorbed frequently increases; this is due to the fact that r organic matter is slowly passing into less stable forms and is therefore less resistant to the permanganate), and rarely . may also result from a reduction of nitrates by organic matter (bacteria, etc.) to nitrites. Tidy's Modification of the Forchammer Process. Special Reagents required : I. A standard solution of the permanganate of potassium lo c.c of which contain i milligramme of available oxygen; made by dissolving which contain ^ " ^ ^.^ ^^ distilled water which has SfSy" n°/.dtir;:iman "nate solution to oxidise any impurities been larntiy 5 ^^^j^ble, and must be frequently renewed. ^ri-Jhly prepared solution of potassium iodide, made by dissolvmg nnp Dart of the pure salt in ten of distilled water. °C«te sulphuric acid (r in 3); a solution of tire P«ga„ate o potassium is dropped in until a faint pinlr tint remains after four hours ^*riTo!utrof':Sium thiosulphate, made by dissolving t gramme of the crystallized salt in a litre of distilled water. r A freshly prepared solution of starch, made by adding 0-5 gramme of wen-washed ItLS to .00 c.c. of cold distilled water, and ^^^Y^^^^^ for five minutes; then let settle and decant the almost clear supernatant liquid. Special Apparatus required : Two thin glass-stoppered bottles or flasks of only a little more than 100 c.c. capacity. ^ . j 1 Two thermometers graduated on the Centigrade scale Graduated burettes; glass stirring rods; white porcelam slabs. The Process. I To 100 c c of the water in one of the thin glass flasks add of the dilute acid; then add 10 c.c. of the standard solu- 10 c.c tion of permanganate, and insert the stopper. 2 The solution of thiosulphate is unstable, and it is therefore advisable to always include a control test as follows: 100 c c. of cold recmtly boiled distilled water are treated m precisely the same manner. 88 LABORATORY WORK 3. Place both flasks in a hot-water o\'en kept at a constant temperature of about 27° C, this being a temperature which facilitates the parting of the oxygen from the permanganate of potassium. In the presence of organic matter live-eighths of the oxygen is liberated from the permanganate in the following manner: KaMugOg -1- 3H2SO4 + oxidizable matter = 2MnS04 + K2S04-i-3H20-f 5O (combined with oxidizable matter). The amount of sulphuric acid added in the process is not sufficient to make the permanganate part with its oxygen, but merely to assist it in doing so in the presence of oxidizable organic matter. 4. After two hours remove the flasks from the oven, and proceed to estimate the amount of iindecomposed permanganate. Add a drop or two of the solution of iodide of potassium, stirring well with a clean glass rod; when the pink colour is entirely replaced by a yellow one (due to free iodine). The undecom- posed permanganate immediately reacts upon the iodide, with the result that an amount of free iodine is liberated proportionate to the amount of undecomposed permanganate, according to the following equation: KaMngOg + loKI + 8H2S04= 2MnS04 + 6K0SO4 + 8H20+5I2- 5. The next step is to ascertain the value of this free iodine in terms of sodium thiosulphate. Add by a graduated burette the standard solution of sodium thiosulphate until the yellow colour has almost completely disappeared — i.e., very little free iodine remains ; and so as to estimate the remaining trace with greater precision, create the blue colour of the iodide of starch by adding a drop or two of starch solution, then resume the addition of the standard solution of sodium thiosulphate until this blue colour has just disappeared. The reaction of the thio- sulphate solution with the free iodine is according to the following equation: 2Na2S203 + 12 = 2NaI -1- Na2S406. If this process of decoloration has been properly performed, and if the necessary amount of thiosulphate solution has not been exceeded, a drop of the permanganate solution will suffice to restore the blue colour to the water. THE OXIDIZABLE ORGANIC MATTER 89 6. When we come to similarly estimate the free iodine of the control test in terms of the thiosulphate solution, the quantity of the latter used will be the amount which is equivalent to 10 c.c. of the standard solution of permanganate (containing i milli- gramme of available oxygen). The difference, therefore, between the amount of thiosulphate solution here required and that required to titrate the amount of free iodine liberated by the permanganate in the sample of water, will represent the amount of permanganate decomposed. Example. — ^The distilled water +10 c.c. of permanganate used up 26-5 c.c. of the thiosulphate solution. .-. 26-5 c.c. of the thiosulphate solution may be considered as equivalent to 10 c.c. of permanganate, or the i milligramme of oxygen which this will part with. The sample water + 10 c.c. of permanganate required only 24-5 c.c. of the thiosulphate solution, and therefore an amount of oxygen equivalent to 26-5 -24-5 = 2-0 c.c. of thiosulphate solution has been taken up by the organic matter. But if 26-5 c.c. of thiosulphate solution is equivalent to i milligramme of oxygen, then 2-0 c.c. = 0-075 milligramme of oxygen. .•. 0-075 milligramme of oxygen is taken up by 100 c.c. of water (100,000 milligrammes); or the oxidizable organic matter in a hundred thousand parts of water required 0-075 part of oxygen to oxidize it in two hours at 27° C. Notes. — It is essential to bear in mind the important fact that there are other substances which water is liable to contain which will reduce the permanganate besides organic matter — i.e., nitrites, ferrous salts, and sulphur compounds other than sul- phates — so that it is necessary to dispose of or account for these before attributing the reduction in the permanganate solely to oxidizable organic pollution. Since as little as | grain to the gallon of iron can be detected by the chalybeate taste which it imparts to the water, in the absence of any such taste the pres- ence of iron may be disregarded. If, however, iron is markedly present as a ferrous salt, one may deduct from the total ox3^gen consumed the amount necessary to convert the iron into the ferric condition (112 parts of iron in a ferrous form will require to this end 16 parts of oxygen). To get rid of the nitrous acid and sulphur compounds other than sulphates, it is necessarj' to boil the water after acidulation with the sidphuric acid for about twenty minutes. Then it is made up to the original bulk with 90 LABORATORY WORK distilled water, allowed to cool to 27° C, and the test is then proceeded ^^dth as above. The amount of permanganate added must always be sufficient to leave a distinct pink colour at the end of the heating. There- fore, in some foul waters it is necessary to make further addi- tions of the permanganate solution, carefully noting the total amount which has been employed. The calculation is facilitated if in these cases the same amount is added to the control flask. At the end of the process — i.e., after titration — the blue colour returns when the fluid has been exposed a few minutes to the air. Conclusions to be Drawn from the Amount Estimated. — In very pure waters the oxygen thus absorbed in two hours is below 0-05 part per 100,000; but a figure not exceeding o-i is not un- favourable to the water's purity. Even when the latter figure is exceeded no definite conclusion can be come to unless the nature of the organic pollution is known, since a peaty water would not necessarily be judged as harmful even if it required three or four times this amount of oxygen to oxidize its vegetable organic matter. The following table of approximate standards for this process was drawn up by Frankland and Tidy : Amounts of Oxygen Absorbed by 100,000 Parts of Water. Derived from Upland Surfaces. Derived from Sources other than Upland Surfaces. Water of great organic purity Water of medium purity . . Water of doubtful purity Polluted water Not more than o-i 0-3 ^love than 0-4 Not more than 0-05 0-15 More than 0-2 1 E. FRANKLAND 'S PROCESS A short reference only to this ingenious process is here given, since it is too difficult and complex for any but trained chemists to perform, and it is generally held that Wanklyn's method attains to as true an estimate of the organic matter at the cost of far less trouble ; and as regards the opinion which the results enable one to form upon the water, it closely coincides with that formed when the same water is analyzed by Wanklyn's method. The process has been employed in certain official analyses, and E. frankland's process 91 the public health student should understand the significance of the terms used to express results. As pointed out on a previous page, the rationale of the method is as follows: When water is evaporated to dryness, and the residue is burnt with the oxide of copper, the nitrogen and the carbon which result from the combustion of the organic matter can be collected and estimated as " organic nitrogen " and " organic carbon." Steps are taken, of course, to ehminate the nitrogen and carbon which are not present in the form of organic matter. The chief objections raised against the process are that it is tedious, costly, and difficult of performance. Inferences are drawn from the ratio which " organic nitrogen " bears to " organic carbon," while the amount of the former is less reliably estimated than that of the latter. Dupre points out that sea-water shows a ratio between the two worse even than is found in pure sewage. By this process the purity of water is judged from a considera- tion of the actual amounts of organic carbon and organic nitrogen present and their relative proportions to each other; and both a low quantity of each and a small relative amount of organic nitrogen to carbon is favourable to the water. Much C and little N is indicative of vegetable pollution; whereas, if the relative proportion of N to C is high, the inference is that the pollution is largely of animal origin. The Rivers Pollution Commissioners held that " a good drink- ing-water should not yield more than 0-2 part of organic carbon, or 0-02 of organic nitrogen in 100,000 parts." They found that in peaty waters the ratio of nitrogen to carbon was i : ii-g, while in similar waters that had been stored in lakes the nitrogen to carbon=i : 5-9. In sewage the average of a large number of samples gave nitrogen to carbon = i : 2-1. Highly polluted well- waters gave nitrogen to carbon = i : 3-1. CHAPTER XII OXIDIZED NITROGEN (NITRATES AND NITRITES) Nitrates and nitrites in water represent the oxidized nitrogen derived, in the main, from the decomposition of nitrogenous organic matter. When organic matter undergoes decomposition, much of the N passes off in the free state, the remainder com- bining with hydrogen to form ammonia; hence when " free and saline ammonia " is found in large quantities in a water it almost invariably affords evidence of the presence of very recent organic pollution, such as raw sewage. As the water continues on its course, the N, mainly through the action of so-called " nitrifying organisms " in the soil, becomes partially oxidized to nitrous acid (HNO2), which, combining with bases (commonly of lime and less often of soda and potash), forms nitrites ; therefore the presence of these salts generally indicates recent organic pollu- tion. The same purifjdng agencies continuing to act, the nitrous acid combines with more oxygen, and becomes nitric acid (HNO3), which forms nitrates of the above-mentioned bases, until ultimately none of the original N may have escaped this complete oxidation. When both forms of ammonia by Wanklyn's process are very low, then practically the whole of the organic matter may be considered as thus purified; when this is not the case, however, purification has only been partially effected. When, as in some rare cases, the water in its subsequent flow meets with reducing agents (either inorganic or organic), the nitrates which have been built up may become gradually deoxidized, and reduced through nitrites to ammonia again; but in these cases one finds practically no albuminoid ammonia, oxidizable organic matter, etc., so that the large amount of free ammonia would not be taken as due to recent animal pollution. It is necessary, however, in all cases where nitrates exist, 92 OXIDIZED NITROGEN (NITRATES AND NITRITES) 93 before ascribing their presence to relatively recent organic pollution (which will be almost entirely of animal origin), to preclude the possibility of their origin from soluble nitrates derived from remote organic matter in the strata permeated, since waters of great organic purity from the chalk, the oolite, the red sandstone and the lias may contain marked traces. It has been suggested that these nitrates may sometimes be derived from fossil remains. An appreciation of these facts enables a true estimate of the importance of the presence and amount of nitrates to be made. Thus nitrites either indicate the incomplete nitrification of ammonia, or the reduction of nitrates by mineral reducing agents or microbes; thus, when they occur in shallow wells or rivers their presence should suffice to condemn the water for drinking purposes, since they would point to the probability that animal pollution is present or very recent ; but when they occur in deep- well water they may not denote present danger, for they may result from the reduction of nitrates, by iron in natural deposits or even by iron pipes. Generally speaking, the importance to be attached to their presence will depend upon the results ob- tained from Wanklyn's process. Nitrites have, of course, a tendency to become nitrates, so that whereas a water often con- tains the latter without any evidence of the former, nitrates will always be found accompanying nitrites; and, owing to their instability, it is exceptional to find nitrites in polluted samples of water. Nitrates and nitrites exist only in traces in waters vitiated by vegetable matter alone, and plant life tends to remove nitrates and nitrites from a water; thus a polluted water, subsequently exposed to plant life, may furnish in its oxidized nitrogen but slight evidence of its previous pollution. Even when the whole of the N of sewage matter is fully oxidized to nitrates, the water must be regarded as dangerous for drinking purposes, for at any time the agencies responsible for the purification may be overtaxed, and dangerous pollution may pass unchanged into the water. Traces of nitrates are present in almost all waters, including rain-water. 94 LABORATORY WORK Qualitative Tests for Nitrates. The old brucine test in careful hands will detect extremely faint traces. A few drops of a saturated solution of brucine are well mixed with half a test-tubeful of the suspected water; then, with the test-tube held well on the slant against a white background, pure sulphuric acid is poured gently down the side until the acid forms a distinct layer at the bottom of the test-tube. When the test-tube is brought to the vertical, a pink zone is seen to occupy the line of junction between the mixture of brucine and water and the sulphuric acid; the pink is transitory, however, and soon changes to a brownish-yellow. Especially does the colour change quickly when the nitrates are liigh in amount. If no coloured zone appears, the test-tube should be gently swayed to and fro, so as, without mixing them, to bring more of the water and brucine in contact with the sulphuric acid; if the results are still negative, nothing but an insignificant trace of nitrates can be present. A control test should be made with nitrate-free water, in order to test the purity of the sulphuric acid. A still more delicate mode of applying the same test is to place 5 c.c. of the water in a perfectly clean platinum dish and evaporate to dryness. Then a drop of pure sulphuric acid is allowed to fall into the dish, and a minute crj^stal of brucine is added. A pink colour will appear with an extremely faint trace. The Diphenylamine Test. — The sensitiveness of this test depends greatly on the mode of performing it. The reagent to be em- ployed is a solution of diphenylamine in sulphuric acid and 5 per cent, hydrochloric acid; three or four drops of this reagent are added to i c.c. of the liquid to be tested, then 2 c.c. of concen- trated sulphuric acid, and the whole shaken. In the presence of nitric acid or nitrous acid the mixture acquires a blue colour. When nitrites are present, the blue colour appears at once, whereas it forms slowly when due to nitrates. No other con- stituent of natural water gives a similar reaction. Most tests for nitrates are responded to equally by nitrites. The brucine and sulphuric acid test responds also to nitrites, but not to nitrites in the absence of nitrates if the acid is diluted with an equal amount of distilled water. OXIDIZED NITROGEN (NITRATES AND NITRITES) 95 Qualitative Tests for Nitrites. The old starch test for nitrites is sufficiently reliable and deli- cate, when carefully performed, for most purposes; but there must be no sulphuretted hydrogen in the water. It consists in the addition of a little clear starch solution, and a drop of a solution of potassium iodide to some of the water in a test-tube. Dilute sulphuric acid is then added, when in the presence of nitrites a dark blue tint appears immediately ; nitrous acid being liberated by the sulphuric acid; it then oxidizes the potassium iodide, leaving the iodine free to combine with the starch as the hhie iodide. The test should be performed at the lowest possible temperature, and an instant reaction must take place, for nitrates give similar results after standing awhile. Ilosvay's Naphthylamine test is more delicate. The following solutions are required: {a) Solution of sulphanilic acid, 0-5 gramme in 150 c.c. of dilute acetic acid (specific gravity, 1-04). (6) Solution of naphthylamine, made by dissolving o-i gramme in 20 c.c. of distilled water, filtering, and adding 150 c.c. of dilute acetic acid. If I c.c. of each of the above solutions be added to 50 c.c. of the suspected water in a Nessler glass, placed upon a white porcelain slab, a pink colour develops if nitrites are present. If no colour appears within fifteen minutes, nitrites may be con- sidered as absent. This method is sufficiently sensitive for ordinary purposes ; but the most delicate appreciation of nitrites is made by acidify- ing a large bulk of the water with acetic acid, and then testing a little of the first part of the distillate from the water. If the water contains sulphuretted hydrogen this must first be separated by means of a little well-washed carbonate of lead and subse- quent filtration. The Quantitative Estimation of Nitrites. The estimation may be based upon Ilosvay's reaction, the degree of colour thereby furnished being matched by means of a standard solution of potassium nitrite, in the manner of the colorimetric estimation of lead, as previously described. 96 LABORATORY WORK The standaxd solution of potassium nitrite is made of the required strength by dissolving i-i grammes of pure silver nitrite in hot distilled water, and then adding a slight excess of potassium chloride. This solution is allowed to cool, and is then made up to i litre; the silver chloride is allowed to settle, and each loo c.c. of the clear supernatant liquid is diluted to i litre; i c.c. of this liquid contains o-oi milligramme of N as nitrite. The solution should be kept in the dark in a number of small bottles filled to the level of the stopper. The quantity of nitrite present is estimated by taking several cylinders containing known amounts of standard nitrite solution, varying, say, from o-02 to 0"i milligramme of N as nitrite in lOO c.c. of distilled water; and i c.c. of each of the two solutions employed in Ilosvay's test must be added to the sample and comparison waters at the same time, since the colour gradually deepens upon standing. As the colour takes nearly a quarter of an hour to fully develop, the cylinders should be covered and set aside for this period before they are compared. Example. — The comparison cylinder containing 8 c.c. of the standard nitrite solution is found to haye the same tint of colour as that produced by the nitrite in the sample, and therefore the amount of N as nitrite in the sample of water is equivalent to that contained in 8 c.c. of the standard solution. But I c.c. of this= o-oi milligramme of N as nitrite. .•. 8 c.c. = o-o8 milligramme of N as nitrite. .•. there is o-o8 milligramme of N as nitrite in lOO c.c. (or 100,000 milligrammes) of water, or O'oS part of N as nitrite in 100,000 parts of water. The starch, iodide, and zinc reaction may also be taken advantage of as a means of making a quantitative estimation on colorimetric principles. If the sample is coloured it must be decolorized as much as possible by adding to 200 c.c. of the water, 3 c.c. of a solution of sodic- carbonate (1:3) and i c.c. of soda-lj'e (1:4); when in most waters the precipitated carbonates of the alkaline earths carry down with them much of the colouring matter. If the water is soft, a few drops of a solution of alum should first be added. As distilled water wiU often give a reaction for nitrite, care must be taken to see that the distilled water employed in the standard solution and in the comparison cylinders does not so react. oxidized nitrogen (nitrates and nitrites) 97 The Quantitative Estimation of Nitrates and Nitrites. A reliable method is that known as the copper-zinc couple process, by which all the oxidized nitrogen in nitrates and nitrites is reduced to ammonia by a wet copper-zinc couple. The ammonia thus obtained can be distilled over and estimated in the manner described in Wanklyn's process. The process is not suited to the estimation of exceptionally large quantities of nitrates, but it is very accurate up to i part of N as nitrates and nitrites per 100,000, and is therefore applicable to the very large majority of waters which are examined as to their fitness for drinking purposes. The Process. 1. A wet copper-zinc couple is prepared by taking a clean and bright piece of thin, well-crumpled zinc foil, and well cleansing this with dilute sulphuric acid. Then the zinc foil, which should measure about g square inches, is covered with a saturated solu- tion of copper sulphate, and very quickly the surface of the zinc loses its bright appearance and becomes covered with a black adherent coating of metallic copper. As soon as this coat has thoroughly formed — and generally about three minutes will suffice — the zinc is removed, or the coating becomes pulverulent and falls away. It is then well washed with distilled ammonia- free water. The wet copper-zinc couple is placed in a thoroughly clean 8-ounce glass-stoppered bottle, with a wide mouth in order that it may take the " couple," and no c.c. of the water are poured in so as to cover the " couple," when the bottle is tightly stoppered and left all night in a dark, warm place (about 20" C ) With very soft water a trace of sodium chloride should be added (about o-i gramme); and with very hard waters a small quantity of pure oxalic acid, to precipitate lime. 2. On the following morning 10 c.c. of the water should be removed and tested for nitrous acid by Ilosvay's test ; the absence of this acid proves the completion of the reducing process, and its presence demands that the reaction should be given more time in which to complete itself. 3. In the absence of nitrites the remainder of the water (100 c.c.) is decanted into a boiling-flask, the bottle is well washed out with ammonia-free distilled water, the washings 7 98 LABORATORY WORK being also added to the flask, and then about 400 c.c. of ammonia- free water are added. 4. The water is next distilled until all the ammonia present has come over. This is then Nesslerized as in Wanklyn's method, and the nitrogen present is calculated from the ammonia thus : The molecular weight of ammonia being 17, and the atomic weight of N 14; N = yi of the ammonia estimated. Of course, the amount of free ammonia originall}- present in 100 c.c. of the water (and which has already been estimated by \Vankl3-n's method) must be deducted from the total ammonia produced by this process. Example. — The water furnishes 0-25 milligramme of ammonia, and since all of this must have been yielded by 100 c.c. of sample, 0'25 part per 100,000 is present. But by Wanklyn's method the water showed o-oo8 part per 100,000 of free and saline ammonia as originally present. After deducting this amount, there is (o'25- 0'008= ) 0*242 part of ammonia due to nitrates and nitrites in 100,000 parts of the water sample. The results are expressed in terms of " nitrogen as nitrates," or as " nitrogen as nitrates and nitrites " in those cases where nitrites are also present. Nitrogen has been seen to form \i of ammonia, therefore there are |i of 0*242 = O'lgg of " nitrogen as nitrates," or of " nitrogen as nitrates and nitrites," as the case may be, in 100,000 parts of water. If in those cases where nitrates co-exist with nitrites it should be desired to express the nitrogen of the nitrates alone, the nitrogen yielded by nitrites may be deducted. Assuming that the water has been found by the Ilosvay colorimetric method to contain nitrogen as nitrous acid to the extent of 0-029 P^-rt per 100,000, then 0'i99-0"029 = 0"i7 is the amount furnished by nitrates alone in 100,000 parts. The ammonia thus furnished is generally in considerable quantities, and the colour in the first one or two Nessler glasses of distillate cannot on this account be directly matched by the chloride of ammonium solution; it can best be estimated by mixing the distillate collected in five Nessler glasses, adding 10 c.c. of Nessler reagent, and then matching the colour in an aliquot part, as previously recommended (see Wanklyn's Method), and then calculating. OXIDIZED NITROGEN (NITRATES AND NITRITES) 99 The phenol-sulphonic acid colorimetrie method of estimating nitrates is not quite so exact as the copper-zinc couple process, but the results can be very much more rapidly arrived at. The lesser delicacy of the process results only in a very slight error of under-estimation, which docs not affect one's judgment upon the water. The Special Reagents required are : 1. Phenol-sulphonic acid, made by mixing 6 grammes of pure phenol, 3 c.c. of distilled water, and 37 c.c. of pure sulphuric acid. Digest for several hours at 82° C. Preserve in a tightly stoppered bottle. 2. A standard solution of potassium nitrate (o-72i gramme to the litre), each c.c. of which contains o-i milligramme of nitrogen. Dilute tenfold, so that each c.c — o-oi. The Process. 1. Ten c.c. of the water sample and 10 c.c. of the standard nitrate solution are each placed in clean platinum dishes and almost evaporated to dryness. 2. Three c.c. of the phenol-sulphonic acid are then run into the dishes, which are subsequently placed on the water-bath for about five minutes. 3. The contents of the two dishes are poured into two separate Nessler glasses, and the dishes are carefully washed out with 25 per cent, ammonia solution. The washings of each dish are added to the Nessler glass which originally received its contents, and then more 25 per cent, ammonia solution is cautiously added to each glass until a yellow colour remains. 4. The contents of the glasses are then filtered (if necessary), and made up to 50 c.c. with distilled water. The Nessler glass containing the standard solution assumes a distinct yellow colour (due to the formation of potassium nitrophenol-sulphonate) , and the contents of the other Nessler glass are also coloured, more or less, in proportion to the amount of nitrate in the 10 c.c. of water sample. By transferring measured quantities from the deeper coloured liquid (which will almost always be that containing the potassium nitrate standard solution) into other Nessler glasses, which are again filled up with distilled water to their marks, a match is obtained; thus it is learnt how much of the deeper coloured liquid is required, when diluted to the 50-c.c. mark with distilled water to match the tint in the cylinder with the less colour; or the darker solution may be poured into a measuring glass and sue- 100 LABORATORY WORK cessive additions of water made until 50 c.c. poured into a Nessler glass is found to effect the match. Suppose that 5 c.c. of the 50 c.c. of the darker coloured (standard) liquid effect a match. Then, since the colour in the whole of the standard liquid represents 10 c.c. of the standard solution of nitrate of potassium, the colour created by nitrates in the 5 c.c. = (J\j or y\j of the 10 c.c.) i c.c. of the standard solution. But I c.c. of the standard solution contains o-oi milligramme of nitrogen as nitrate; therefore there is o-oi milligramme of such nitrogen in 10 c.c. (10,000 milligrammes) of water, or o"i milli- gramme in 100,000 milligrammes of water, or 0"i part per 100,000. If the sample is darker than the standard, then measured quantities of the sample must be removed and made up to 50 c.c. with distilled water until a match is obtained — e.g., sup- posing 25 c.c. suffice for the match, then the sample cylinder contains twice as much oxidized N as the standard cylinder ; and therefore the 10 c.c. of original water contained 0"2 milligramme of N as nitrates; or 2 parts per 100,000. Nitrites slightly add to the colour formed in this process, and chlorides interfere with the delicacy of the estimation by furnish- ing lower results. If chlorides exceed 5 parts chlorine per 100,000 in the original water, a trace of pure sodium chloride should be added to the standard solution of nitrate of potassium. Conclusions to he Drawn from the Amount Estimated. — The significance of the presence of nitrites and nitrates has already been discussed; and it has been seen that high nitrate indicates previous pollution, either distant and old, or near and recent. When the " nitrogen in nitrates " exceeds 0"i part per 100,000, suspicion is certainly justified in those cases where the strata may be excluded as the source from which the water may have derived such nitrogen; but where such a source cannot be ex- cluded, an amount exceeding 0*5 would be regarded as suspicious, (or it is exceptional that more is derived from entirely harmless sources. More than O'l part per 100,000 in rain or upland surface water is therefore significant of animal contamination. No hard limits, however, can be accepted for all waters, and the amount of oxidized nitrogen must be considered in conjunction with the results of the other processes that help to furnish evi- dence of contamination. CHAPTER XIII THE GASES IN WATER It is the aeration of water which furnishes its pleasant taste and sparkhng appearance. The degree of aeration — as has been already pointed out— affords no evidence of the water's purity or impurity, since the foul water of a shallow polluted well is frequently markedly aerated, whereas the pure water collected from great depths is sometimes poorly so. Rain water, when thoroughly aerated, contains about 2073 c.c. of gases per litre — i.e., nitrogen i3"o8, oxygen 6-37, and carbonic acid I "28 c.c. In addition to the innocuous gases upon which the aeration of a pure water depends — i.e., nitrogen, oxygen, and carbonic acid— it is obvious that water may take up noxious gases, or those which, as they are generally the products of organic decomposi- tion, may indicate danger (such as sulphuretted hydrogen, ammonia, marsh gas (CH4), etc.). To ascertain whether free carbonic acid exists in the presence of bicarbonates, a solution may be used of i part rosolic acid in 500 parts of 80 per cent, alcohol (to which baryta water has been added until it begins to acquire a red tint); when | c.c. of this is added to 50 c.c. of water, no change takes place if free CO2 is present, but a distinct reddening occurs in the absence of free COg. The Lunge-Triliich method of estimating the free carbonic acid is very easy and accurate. When NajCOg is added to water containing free COo, sodium bicarbonate is formed (Na2C03 + C02 + H20= 2NaHC03), and when all the free CO2 is combined the water reacts alkaline to phenolphthalein. The amount of Na^COg used may therefore be made to indicate the amount of CO2 present. One hundred c.c. of the sample are mixed with a few drops of 102 LABORATORY WORK a neutral alcoholic solution of phenolphthalein and titrated in a narrow glass cylinder with a ^^f solution of sodium carbonate, until a faint, but permanent, red tint appears. This gives the amount of free CO2. Example. — 2"2 c.c. of ^ NagCOg were required to neutralize the free CO2 in 100 c.c. of water. One c.c. of ~y Na^COg contains ( i;'^"" ) o"00265 gramme of Na-^COg. But 106 parts of Xa^CO^ neutralize 44 parts of CO,. .•. One c.c. of -.^^ NaXOg^/j/.v of 0"00265 = 0001 gramme of CO.. .•. 2'2 C.C. of ttV Na2C0j= 2'2 X o'OOi gramme CO^ o"0022 gramme of 002- .•. 0022 gramme COo in 100 c.c. water, or 2'2 parts per 100,000. CO. as Carbonate and Bicarbonate (Thorpe's Method) : 1. Take 100 c.c. of water in a flask and add a drop or two of phenolphthalein. 2. Add standard oxalic acid solution (2"863 grammes of pure recr3'stallized oxalic acid to the litre of distilled water, i c.c. of which equals i milligramme of COg) until the phenolphthalein is decolorized, carefulh^ noting the amount of solution used. This will indicate the CO2 present as carbonate. 3. Next boil the water for about ten minutes, and then note the amount of standard acid required to decolorize. When the acid is first added to the water the carbonates are converted into bicarbonates, and when this conversion is com plete the phenolphthalein is decolorized. When the water is boiled CO2 is driven off, and the converted bicarbonates and the original bicarbonates are reduced to car- bonates. The second titration will furnish the amount of CO2 remaining after boiling. Twice this quantit}/ will represent the total CO2 as bicarbonates prior to boiling, and this amount less the original CO2 in carbonates will furnish the amount of CO2 as bicarbonates originally in the water. Example. — One hundred c.c. of water required 2 "5 c.c. of oxalic acid= 2*5 milligrammes of CO2 as carbonates, or 2*5 parts, per 100,000. After boiling, 3-5 c.c. of oxalic acid were required. The total CO2 as bicarbonates is therefore 7 milligrammes. Deduct the THE GASES IN WATER IO3 2-5, and the amount in the original water was 4-5 i)arts per 100,000. Sulphuretted hydrogen frequently gains access to water through organic substances undergoing putrefaction, or in- directly from industrial waste matters. It may be derived from mineral sulphides, or from reduction of mineral sulphates in soil, etc. (which reduction is frequently effected by organic matter and hving organisms, such as Beggiatoa alha). Good examples of waters charged with sulphuretted hydrogen from harmless sources are to be found notably at Harrogate and Aix-la-Chapelle. The waters from some clays have a distinct amount of sulphuretted hydrogen, derived from the tiny particles of iron pyrites which enter into the composition of the clay. When sulphuretted hydrogen, ammonium sulphide, or the constituents of coal gas are present, the water will generally be condemned, as either unsuitable for a domestic supply or as polluted. If HgS is in considerable amount, the addition of a solution of acetate of lead produces a brownish coloration (Pb(C2H30,)2 + HgS^ PbS + 2C2H4O2). The gas may be estimated while in the water in the following manner : Take a large flask and add 10 c.c. of centinormal iodine solu- tion; then run in a measured quantity of the water until the yellow colour of the free iodine disappears (I2+ H2S=2HI + S); then add 5 c.c. of starch solution, and run in more of the iodine solution cautiously until a blue colour just begins to show itself. Of the iodine solution used each c.c. will have decomposed 0-17 milligramme of HgS, and therefore the total will represent 17 milligrammes of HgS. The slight excess of iodine required to produce the blue colour is trivial, but it may be estimated and deducted by titrating back with sodium thio- sulphate. Centinormal iodine (1-26 grammes iodine per litre) is pre- pared as follows : Iodine is never quite pure, and is very volatile. Dissolve therefore about I -3 grammes in a solution of 2 grammes of potassium iodide in 50 c.c. of water, and dilute to a litre; then further dilute until 10 c.c. of the solution, coloured blue with a few drops of starch solution, are decolorized by exactly 10 c.c. of centinormal sodium thiosulphate (2-464 grammes to a litre) . The solution should be kept in the dark. 104 LABORATORY WORK Evidence of odorous gases may be obtained by heating the water to 60° C, in the manner already described in reference to "odour." A few drops of a solution of the nitro-prusside of sodium \vill distinguish between sulphuretted hydrogen and ammonium sulphide, for that solution furnishes a violet-purple colour with ammonium sulphide, but no change results if sul- phuretted hydrogen alone be present. Some waters issuing as springs in the vicinity of volcanoes are charged with sulphurous acid. The Estimation of the Oxygen Dissolved in Water. The estimation of oxygen is of some importance, for it gives a clue to the self-purifying power of the water, and it is rapidly diminished in oxidizing the C and N of any putrescent organic matter present. Dr. Thresh has devised a satisfactory process for this estima- tion, one that is quickly performed and easy of execution. When dilute sulphuric acid and the iodide of potassium are added to water containing nitrite, iodine is liberated, the amount varying with the length of time during which the mixture is exposed to the air. If air be carefully excluded there is no in- crease in the amount of iodine set free after the first few minutes. We have only, therefore, to add to a known quantity of the water a definite amount of sodium nitrite, together with excess of potassium iodide and acid (avoiding access of air), and then to determine volumetrically the amount of iodine liberated. After certain deductions the remainder represents the dissolved oxygen present in the water. The following are the reagents required : 1. Solution of sodium nitrite and potassium iodide: Sodium nitrite . . 0-5 gramme. Potassium iodide . . 20-0 grammes. Distilled water . . 100 c.c. 2. Dilute sulphuric acid, i in 4. 3. A clear or fresh solution of starch. 4. A volumetric solution of sodium thiosulphate: Pure crystals of thiosulphate, 7-75 grammes; distilled water, to i litre. One c.c. corre- sponds to 0-25 milligramme of oxygen. Keep in small bottles in the dark. The apparatus is used in the following manner : The bottle A being clean and dry, the perforated bung is inserted, the burette charged, and the tube B fixed in its place. THE GASES IN WATER IO5 E is connected with the gas supply. The tube D is filled to the level of the stopper with the water to be examined, then i c.c. of the solution of sodium nitrite and potassium iodide and i c.c. of the dilute acid are added, and the stopper instantly fixed in its place, displacing a httle of the water and including no air. If the pipette be held in a vertical position, both the saline solution and the acid, being much denser than the water, flow FIG. 15. APPARATUS FOR THRESH S PROCESS. in a sharply defined column to the lower part of the tube, so that an infinitesimally small quantity (if any) is lost in the water which overflows when the stopper is inserted. The tube is next turned upside down for a few seconds for uniform ad- mixture to take place, and then the nozzle is pushed through the bung of the bottle, and the whole allowed to remain at rest for fifteen minutes to enable the reaction to become complete. A rapid current of coal gas is now passed through the bottle A I06 LABORATORY WORK until all the air is displaced, and the gas lighted at G burns with a full luminous flame. The flame is then extinguished, the stopper of D removed, and the cork G rapidly inserted in its stead. On turning the stopcock the water flows from D into the bottle A, which now contains no oxygen. The stopcock is turned off, the cork G removed, and the supply of gas regulated so that a small flame only is produced when this gas is ignited at G. The thiosulphate is now run in slowly from the burette C until the colour of the iodine is nearly discharged. A little solution of starch is then poured into D, and about i c.c. allowed to flow into the bottle by turning the stopcock. The titration with thiosulphate is then completed. After the discharge of the blue colour, the latter returns faintly in the course of a few seconds, due to the oxvgen dissolved in the volumetric solution; therefore, after about two minutes o-i c.c. of thiosulphate must be added to effect the final discharge. The total amount of thiosulphate solution used must now be noted. This will repre- sent the iodine liberated on account of — (a) The oxygen dissolved in the sample used. (b) The chemical interaction between the KI.NaNOg and H2SO4. (r) The oxygen dissolved in the reagents added. (d) Any nitrous acid present as nitrite in the sample would liberate a little more iodine. To find the value of (a), it is obvious that the value of (&), (c) and (d) must be ascertained. The value of {b) and (c) can be most readily found by making a blank experiment, by adding to the apparatus 2 c.c. of the nitrite iodide solution, 3 c.c. of acid, 3 c.c. of starch, and distilled water equal to twice the volume of thiosulphate used in the analysis, and titrating with thiosulphate as in the actual experiment. Half the thiosulphate solution used represents the amount which must be deducted on account of (b) and (r). To find the value of {d), a water containing nitrites will require the amount of the nitrous acid to be determined if the utmost accuracy is required (A water containing i part HNOg in 1,000,000 will effect the results +0-17 milligramme of oxygen per htre; 94 parts of the acid corresponding to 16 of oxygen). Where nitrites are present in sufficient quantity to interfere, the amount may be determined by any of the ordinary processes. THE GASES IN WATER 107 Notes.— The test depends on the liberation of iodine anrl nitric oxide (NO), by the interaction of sodium nitrite, potassium iodide, and sulphuric acid in the water, and the NO combines with what dissolved oxygen is present to form nitric trioxide (N2O3). This N2O3 then liberates a further quantity of free iodine, which is therefore proportionate to the amount of dis- solved oxygen in the water. The total iodine is then determined by sodium thiosulphate solution, and the amount furnished by the dissolved oxygen can be calculated. The various chemical reactions are as follows: (i) 2KI + 2NaN02 (2) 2HI + 2HNO2 - h 2H2SO4 == K2SO4 + Na2S04 + 2HNO2 + 2HI. I2 + 2NO + 2H2O. (3) 2NO + dissolved in water =N203; and the N2O3 in contact with hydriodic acid hberates a further quantity of iodine, which is therefore equivalent to the oxygen present : (4) 2HI + N2O3 = 2NO + 12 + H2O. The sodium thiosulphate reacts with the liberated iodine as follows : (5) 2Na2S203 + l2 = Na2S406 + 2NaI. Samples should always be drawn from the water in a quiescent condition, and every precaution must be taken to avoid splashing or further aeration. A few of the results got by Thresh are as follows : Source of Water. i Amount of Water employed. 1 Total Thio- sulphate required. Thio- sulphate required by the Dis- solved in the Water. 1 1 MiUi- i grammes O per Litre. 1. Spring water 2. Rain water . . 3. Shallow-well water . . 4. Rain water . . 5. Distilled water shaken with air 322-0 322-0 322-0 322-0 322-0 12-35 13-05 11-35 12-95 16-00 9-87 10-55 8-90 10-45 13-40 7-66 8-19 6-91 8-II i 10-40 Nitrites, ferrous iron, sulphides, and a large quantity of organic matter will vitiate the estimation, but Rideal has pointed out that it is practicable to get rid of the interference due to nitrites, organic matter, etc., by oxidizing with permanganate in acid solution out of contact with air. I08 LABORATORY WORK The process gives results closety approximating to those obtained by a gasometric process. A simpler method for the estimation of dissolved O in water is that of Winkler : Winkler's Method for the Estimation of Dissolved Oxygen in Water. In collecting the sample of water, care must be taken to avoid agitating it and exposing it for any length of time to the air. 1. A portion of the sample is transferred with the above- mentioned precautions to a glass-stoppered bottle of known capacity. A suitable capacit}- is about 300 c.c, and the bottle must be completely filled. 2. One c.c. of strong manganous chloride solution (40 grammes of MnCUHgO to 100 c.c. of distilled water) is added, followed by 2 c.c. of a solution containing 33 per cent, caustic potash and 10 per cent, potassium iodide. 3. The bottle is stoppered without including any air-bubble, and the liquids are mixed by several times inverting the bottle. The manganous hydroxide precipitate which forms will be more or less discoloured by higher hydroxide, according to the propor- tion of oxygen which was dissolved in the water sample. 4. As the oxidation of the manganous hydroxide is not immediate, and the result is influenced by light, the bottle is put aside in a dark cupboard for fifteen minutes; 2 to 3 c.c. of pure strong hydrochloric acid are then added by means of a pipette inserted into the bottle, so that the acid will fall upon the precipitate, when the precipitate disappears and leaves the liquid coloured with dissolved iodine, which is proportionate in amount to the higher hydroxide formed, and therefore to the dissolved oxygen in the water. 5. Pour the contents of the bottle into a clean beaker, washing out the bottle with distilled water and also adding the washings to the beaker, and then titrate the iodine with decinormal sodium thiosulphate, of which i c.c. is equivalent to o-ooo8 gramme of oxygen. Starch should be used for the end reaction, as recom- mended in Tidy's process. Example. — Capacity of bottle (283 c.c.) less the 3 c.c. of solu- tion added =280 c.c. The decinormal sodium thiosulphate required in the test was 2-8 c.c. = 0"00224 gramme of oxygen. THE GASES IN WATER lOQ Therefore there is 0-00234 gramme of dissolved oxygen in 280 c.c. of water=0"0008 gramme of oxygen in 100 c.c. of water = o-oo8 gramme of oxygen in 1,000 c.c. of water, or 8-0 milHgrammes per Htre. Notes on the Process.— Sometimes the amount of thiosulphate required by the same volume of fully aerated distilled water is determined, and the percentage of dissolved oxygen in the sample is compared with the amount of thiosulphate which equal volumes of these two waters require. The manganous chloride must be free from iron, and all the reagents must be free from nitrites. The process must be done rapidly. Nitrites liberate iodine and so vitiate the result by increasing it. Much organic matter interferes with the method, for it absorbs liberated iodine, and thus diminishes the result. The chemistry of the process is explained by the following equations : (i) 2Mna2 + 4NaOH = 4NaCl + 2Mn(OH)2. (2) 2Mn(OH)2 + + H20=2Mn(OH)3. (3) 2Mn(OH)3 + 6HCl=2MnCl3 + 6H20. (4) 2Mna3 + 2KI=2MnCl2 + 2KCl + l2. The amount of dissolved oxygen in a water is influenced by temperature, being less in summer and more in winter. Ordinary tap water in this country contains between 6 and 7 c.c. per litre, or about i part by weight in 100,000. Water is saturated at 5° C, 10° C, 15° C, and 20° C. respectively, by 8-68 c.c, 777 c.c. 6-96 c.c, and 6-28 c.c per litre. When nitrites excaed faint traces the results are too high o^Adng to the reaction between the nitrous acid and hydriodic acid; the reaction is catalytic, the nitric oxide formed absorbing oxygen from the air and yielding nitrous acid, which in turn decom- poses a further quantity of hydriodic acid. This effect may be prevented by carrying out the method in the usual way, and introducing 2 c.c. of potassium acetate solution (1,000 grammes per htre) when the precipitate has dissolved in the added hydro- chloric acid. The acetate solution should be added by means of a pipette reaching to the bottom of the bottle (Hale and Meha). CHAPTER XIV COMPOSITION OF WATER FROM VARIOUS SOURCES— THE OPINION ON WATER SAMPLES Waters from the subsoil, from cultivated surfaces and from rivers are especial!}^ liable to be organically polluted; and the character of the deposit from which the water is collected in- fluences its composition to an extent which, though variable, may be approximately defined. I. Surface Waters. — Those waters collected from the hard sur- faces of the practically impervious rocks, which support httle animal or vegetable life, are very pure. They commonly contain less than lo parts of total solids, 5 of total hardness, i of chlorine, and o-i of nitrogen as nitrates, in 100,000 parts of water. The mineral solids consist mainly of sodium carbonate and chloride, and a trace only of lime or magnesia. The organic matter, which is often exclusively of vegetable origin (peat), yields practically no free ammonia; but the organic ammonia figure and that of the oxygen absorbed by organic matter may be high, in which case the water is often highly coloured and acid in reaction. Such characters are presented by the waters collected from the sur- faces of the igneous, metamorphic (quartz, mica, granite, etc.), Cambrian, Silurian and Devonian, rocks. Waters from the surface of the non-calcareous carboniferous rocks (Yoredale rocks, millstone grits and coal-measures) are very similar; but those which have flowed over the surfaces of the calcareous carboniferous rocks — the mountain limestone and limestone shales — differ from the former in possessing a moderate degree of hardness, higher total solids and a neutral or faintly alkaline reaction, the mineral solids consisting chiefly of sulphate and carbonate of calcium and magnesium. Surface waters from the lias, new red sandstone, magnesian limestone and oolite vary considerably' in their composition. COMPOSITION OF WATER FROM VARIOUS SOURCES III The total solids are generally between lo and 20 parts per 100,000, the total hardness between 10 and 15; the chlorine is below 2. and the nitrogen as nitrates below 0-2 of a part per 100,000. Clay surface waters are, as a rule, opaque from a variable quantity of suspended matter, but generally there are few dis- solved solids, and the water is fairly soft. They vary greatly, however, in their composition. Waters collected from cultivated land present great varia- tions in composition; the total hardness may range from 5 to 25 parts per 100,000, according as to whether the soil is non- calcareous or calcareous. Alluvium is generally a mixture of sand, clay and organic matter; and waters from such a source mostly contain high mineral sohds (50 to 100 parts), consisting of calcium and magnesium salts, sodium chloride, iron, and silica. 2. Waters from a Depth. — Those collected from the chalk are generally clear, bright, and well charged with carbonic acid. The total sohds are generally from 25 to 50 parts per 100,000, and the total hardness from 15 to 30 parts; the hardness is mostly temporary, and calcium carbonate may vary from 10 to 30 parts. The chlorine is commonly from 2 to 4, but it may reach a higher figure in some pure chalk waters. The nitrogen as nitrates is generally below 0-5 part per 100,000, and is com- monly about 0-2. Sulphates are present in small quantity, and there is often a trace of phosphates and of iron. Although the car- bonic acid present may be sufficient to turn blue litmus red, when this gas is driven off by heat an alkaline action is invariably obtained. Some waters from the chalk are very soft, and contain sodium carbonate. They are only found where the chalk lies buried beneath a thick mass of London clay (Thresh). Waters from the oolite present characters very similar to those from the chalk. Those derived from limestone and magnesium limestone for- mations only differ from the chalk waters in generally containing more total solids, far more calcium or magnesium sulphate (which may reach nearly 20 parts per 100,000), and less calcium or magnesium carbonate; and by consequence the hardness is generally higher and in a greater proportion " permanent." In dolomite districts the mineral solids contain much mag- nesium carbonate and sulphate, and a large proportion of the 112 LABORATORY WORK total hardness is" permanent," dolomite being a double carbonate of lime and magnesia. The greensands are porous strata containing a reducing salt of iron, which, by reducing oxidized nitrogen to ammonia, often furnishes to the water a very high figure of free and saline ammonia. The total soHds vary considerably, but they some- times approach loo parts per 100,000 where the water is col- lected at great depths in greensand underlying the chalk; the chlorine may reach a figure of from 4 to 15; the total hard- ness (much of which is " permanent ") is very variable — from a low to a high figure; and the nitrogen as nitrates is generally from about 0-3 to o-6 part per 100,000. Where the lower greensand is exposed it is very porous, and many of the waters yielded from it contain but little lime and hardness. The total solids are often small in amount, and the chlorine and sahne constituents (including ammonia) may be low. A marked amount of nitrate is often present, and not infrequently a considerable quantity of iron is found in solution. Such waters differ materially from those obtained from the covered beds; the latter often containing large amounts of saline matter (chiefly alkaline chlorides, sulphates, and carbonates). This difference in the composition of the waters from exposed greensand and those from covered beds is probably accounted for by the ex- planation that the soluble constituents in the very porous exposed part of the bed have been largely washed out by the rapidly percolating waters. Waters from red sandstone strata vary considerably in their composition, according as the deposit is pure or impure, soft or hard. The total solids and total hardness are both sometimes high, and the former may reach 100 parts per 100,000; the latter is mainly of a permanent nature, but the water may sometimes be soft and possess a total hardness figure not exceeding 10 parts per 100,000. The chlorine may vary from 3 to 6; and traces of phosphates are always to be detected in the mineral solids, which consist in the main of sodium chloride, carbonate and sulphate, calcium and magnesium carbonates and sulphates, and a trace of iron. Waters from selenitic deposits are sometimes objectionable, on account of the large proportion of calcium sulphate (10 to 30, or more, parts per 100,000) which is taken up from this deposit, which itself consists of calcium sulphate in clear crystals. THE OPINION ON WATER SAMPLES 113 Waters collected from loose sands are of variable composition. Some are soft, with total solids of from only 6 to 12 parts per 100,000, and others are rather hard (permanent) with mineral solids amounting to even 100 parts. The chlorine figure is generally rather high, and may reach to a very high figure in some cases. The mineral solids consist of sodium chloride, carbonate and sulphate, calcium and magnesium salts, and traces of iron and silica. Those from gravel are generally soft, but some are liard, with rather high total solids. Waters of the latter class coming from a depth have often very high mineral sohds (often consisting largely of calcium and magnesium sul- phate). There is, as a rule, some opacity, and the physical characters generally are not favourable to the water. The hard- ness, which is almost entirely " permanent," is often over 20, and the mineral solids may in some cases reach a high figure. Deep wells, when protected from surface drainage and ground water in their upper parts, are but rarely polluted, even when situated in the centres of towns. But it does occasionally happen that Hquid soakage from sewers or cesspools finds its way into fissures in chalk or sandstone, which conduct it to the water of the well, maybe from a considerable distance. The Opinion on Water Samples. If the analysis does not justify suspicion, and local circum- stances do not favour any form of dangerous contamination, then, and then only, may the water be judged safe for drinking purposes. Either a chemical analysis or a bacterioscopic examination may alone suffice to demonstrate the fitness of a water for drink- ing purposes, but the two processes are complementary to each other, and it is essential in many cases that they should both be performed. Neither is infallible; the value of one is often greatly enhanced by the other; and the value of both depends upon the correct interpretation of results. Between an undoubtedly bad water and an undoubtedly good water there are waters regarding which no opinion ought to be advanced unless advantage is taken of a careful local inspec- tion to detect any possible source of pollution and both a bacterio- logical examination and a chemical analysis are performed. It is true that so small an amount of organic matter as would not call for condemnation of the water may yet contain the 114 LABORATORY WORK specific germs of disease; but chemical analj'Sis will generally reveal impurity and risk. As specific germs always gain access to the water in the media of dirt and animal matter, it is rare that. the chemical examination fails to indicate the danger; but whenever there is any reason to suppose that a water-supply is infected with typhoid or cholera organisms, a bacterioscopic examination becomes imperative. Indeed, the circumstances which call for the analysis of a particular sample should always weigh with one in fixing the standard of purity which will justify the opinion that the water may be consumed with safet3^ The detection of a very fine amount of organic contamination in shallow wells may often be made by collecting samples from several wells in the near neighbourhood of each other, and taking the purest of these waters as a standard. In judging of the purity of water from a river or small stream, it will sometimes be advantageous to make an analysis of tribu- tary streams emptying into it. Water from different sources (peaty and non-peaty surface waters, waters from shallow and deep sources) have their own characteristics, and the various quantitative estimations must be interpreted accordingly. It is impracticable to lay down hard- and-fast standards of purity (either chemical or bacteriological) to which all waters must conform, irrespective of their source. In expressing an opinion upon the analytical results, it is very desirable to give expression to the necessary limitations of the examination. Thus, it will be better to say, " The chemical analysis of the sample shows no evidence of harmful organic pollution," than to employ such a statement as, " There is no harmful organic pollution in this water." In order to commit oneself to a statement which would imply that the water-supply is a constantly pure one, it will often be necessary to examine several samples at different seasons of the year and under dif- ferent conditions as to rainfall. Water Standards. — It would often be of assistance to analysts if they carefully constructed " water standards " in their dis- tricts. Such would be prepared from the purest water in each locality, and they would form a reliable means of detecting the smallest amount of impurity which gains access to any particular supply. For instance, a water sample may well contain 0"004 part per 100,000 of free and saline ammonia and 2 parts per 100,000 of chlorine, without any suspicion of danger being war- ranted; but if the average of the pure water of the particular THE OPINION ON WATER SAMPLES 115 locality is 0'002 part of free and saline ammonia and i'5 parts of chlorine, then the excess found would furnish important evidence of animal pollution. In some of the American States, Massachusetts notably, the normal distribution of chlorine has been mapped out. This has been done by estimating the amount of chlorine in the unpol- luted waters of the district at a large number of points, and thus ascertaining the normal distribution of chlorine in every par- ticular locality. When lines are drawn upon the district map which join together the localities where the chlorine figures corre- spond, these lines are known as " isochlors." The plans of the " isochlors" are valuable, as when in a given spot an amount of chlorine is found in excess of the figure of the particular " isochlor " of the locality, it furnishes material evidence of sewage pollution. Chlorine standards are of the most value when they relate to surface water, in which the small clilorine figure remains very uniform indeed in the absence of animal pollution; but even then they often fail to indicate harmful pollution. The superior value of the " free and saline " ammonia figure is very apparent in the following results: ' 1 A. Water. B. Sewage. 0. Water+i^n per cent, of Sewage. Parts per 100,000. Parts per ioo,coo. Parts per 100,000. Free and saline ammonia 0-0005 6-2IOO 0-0067 Albuminoid ammonia 0-0035 1-3200 0-0048 Oxygen absorbed in 2 hours at27°C. 0-0300 4-1430 0-0341 Total solids 30-1 150-0 30-25 Chlorine 1-9 10-2 I-QI N as nitrates o-ig O-OO 0-18 Thus the water polluted with —^ per cent, of the crude sewage taken in the experiment (or i gallon of sewage to i,ooo gallons of water) shows the free and saline ammonia increased about 13 times, the albuminoid ammonia increased by not much over I of its original quantity, the total solids remain practically the same, and the chlorine shows no appreciable increase. When the health officer has to make periodical examinations of the same water, it will not be necessary, in order to detect contamination, always to perform a complete analysis. No object will be gained, for instance, by the estimation of the total solid matter and the hardness; but one must never neglect Il6 LABORATORY WORK the free and saline ammonia ligure, which is of paramount im- portance. It ma}- be said with certainty that in such cases the shghtest increase of ammonia slioiild be regarded with grave suspicion. In the hght of a chemical anal3'sis the fitness of the water for drinking purposes will be determined by the following considera- tions: Is any evidence forthcoming of animal contamination present or past ? Is there evidence of excessive ^•egetable pollution ? Is the nature and amount of the saline constituents likely to pro\'e harmful ? Are poisonous metals present ? The suitability of the water for other uses, such as for washing and trade purposes, will depend upon its ligure of hardness, the absence of a marked colour and freedom from suspended or deposited matter. The evidence of present or recent animal pollution is more especially indicated by high chlorine and oxidized nitrogen, in association with marked free and albuminoid ammonia; and that of past or remote animal pollution by high clilorine and oxidized nitrogen (not accounted for by the strata permeated) with little free and albuminoid ammonia. If the pollution is solely of a vegetable origin it is indicated by high figures of albuminoid ammonia and of oxygen absorbed by Tidy's method, in association with very low figures of clilorine and oxidized nitrogen in the case of surface waters, and with practically no increase in these latter two figures if the water is collected from below the surface. Waters containing excessive vegetable pollution would generally be coloured, and the solid residue would char considerably upon ignition. As to excessive or harmful mineral matter in a drinking-water, a limit of lOO parts per 100,000 ought not to be exceeded; but it is generally the nature of the mineral matter rather than its amount which will affect the opinion. Sulphates should not furnish more than 10 parts of SO3 per 100,000; iron is only permissible when in traces, and other poisonous metals should be absent. A water containing over 30 parts of hardness may be regarded as unfit for trade, washing, and cooking purposes. It is rare with these waters that the total hardness cannot readily be reduced considerably by a water-softening process. A few examples of waters from difierent sources, together with the opinion upon them which the chemical analysis (expressed as parts per 100,000) appears to warrant, may now be given. THE OPINION ON WATER SAMPLES 117 Physical characters Reaction Free and saline ammonia. . Albuminoid ammonia O absorbed from permanganate (in two hours at 27° C.) Total solid matters (a) Volatile . . {b) Fixed (c) Appearance on ignition Total hardness {a) Temporary (6) Permanent Chlorine N as nitrates Note. — ^Sample 2 is evidently a chalk water contaminated with animal matter, as evidenced by the high ammonias (the " free" being higher than the "albuminoid"), and the high figures of oxidized nitrogen and chlorine (for a chalk water) . The very high figure of free and saline ammonia points to recent and therefore specially dangerous contamination. 1. 2. A wry I'nre A Foul anfl WhId: Dangerous Watur. Excellent Excellent Faintly Markedly alkaline alkaline o-ooi 0'022 0'002 0-014 0'0I2 0-114 l8-o 38-4 4-6 18-1 13-4 20-3 Nil Marked charring 9-0 24-0 4-0 16-0 5-0 8-0 I'O 6-2 O'OI 0-8 Physical characters Reaction Free and saline ammonia Albuminoid ammonia O absorbed from permanganate (in two hours at 27° C.) . . Total solid matters {a) Volatile {b) Fixed {c) Appearance on ignition Total hardness {a) Temporary (&) Permanent Chlorine N as nitrates Notes. — Sample 3. In industrial towns the reaction may be slightl}^ acid, from the sulphuric acid in the atmosphere; and the water is a little different in other respects owing to further impurities taken up, such as soot, sulphur compounds, and increased ammonia. Thus the rain falling in INIanchester has been found to contain 0-7 part per 100,000 of free ammonia, 0-03 of albuminoid ammonia, 4-7 parts of sulphuric acid, and 0-5S of hydrochloric 3. 4. Rain Water (Country). Good Subsoil Water (Gravel over Chalk) Good Faintly alkaline Alkaline 0-050 0-002 0-005 0-006 0-005 o-o6i 3-0 29-2 1-5 12-2 1-5 Nil 17-0 Slight charring 0-5 20-0 o-o 5"5 0-5 14-5 0-25 2-1 0-02 0-2 ii8 L\BOKATORY WORK acid. Rain collected on the sea coast has been found to contain as much as 5*4 parts per 100,000 of chlorine (chlorides). Rain water which is collected in countr}' districts after long periods of continuous rainfall provides the purest possible nahtral water. Its com- position varies throughout the year somewhat. Sample 4. The analysis does not furnish evidence of harmful contamina- tion; but the ammonias suggest the presence of a little vegetable matter. Physical characters Reaction Free and saline ammonia. . Albuminoid ammonia O absorbed from permanganate hours at 27° C.) . . Total solid matters («) Volatile . . (6) Fixed (c) Appearance on ignition Total hardness (a) Temporary {b) Permanent Chlorine N as nitrates (in tw A Peaty Surface Water. Brownish- yellow Acid o-ooi 0'022 0'i6o I2"0 9-5 2-5 Marked charring 3.0 o-o 3-0 0-7 O'OI 6. A N on -peaty Surface Water on Millstone Grit. Nearly colourless Neutral 0'002 0*004 0-040 5-5 2-0 3-5 Faint dis- coloration 2-5 O'O 2-5 0-8 0-05 Note. — In many "peaty waters" the "organic ammonia" and the " oxygen absorbed " will be found to much exceed the amounts given above. Neither of the above analyses furnishes evidence of harmful con- tamination. Physical characters Reaction Free and saline ammonia. . Albuminoid ammonia O absorbed from permanganate hours at 27° C.) . . Total solid matters {a) Volatile . . {b) Fixed (c) Appearance on ignition Total hardness (a) Temporary (6) Permanent Chlorine N as nitrates in tw 7. Deep-well Water (from Chalk). Excellent Alkaline 0'Oo6 O'OII 0'oS4 40-0 15-5 24-5 Marked charring 26*0 15-5 TO-5 4-5 0-6 8. Deep-well Water (from New Red Sandstone). Good Alkaline o-ooi 0'002 0'0I2 30-2 8-6 21'6 Nil 19-5 8-0 "•5 2-2 0.3 THE OPINION ON WATER SAMPLES 119 Notes. — A deep-well water from the chalk may ( ontain total solids up to 200 parts per 100,000, but such an amount is rare. Sample 7 is a water the analysis of which warrants suspicion of organic contamination. Slight animal contamination is probable in such a water, having regard to the figures of the saline and albuminoid ammonia, the chlorine and the oxidized nitrogen. A careful local inspection might detect the source of some pollution. There is no reason to question the purity of Sample 8. Physical characters Reaction Free and saline ammonia. . Albuminoid ammonia O absorbed from permanganate hours at 27° C.) . . Total solid matters {a) Volatile . . {b) Fixed (c) Appearance on ignition Total hardness . . (a) Temporary {b) Permanent Chlorine N as nitrates 9. 10. River Water. Good New River Water (as supplied in London). Excellent Faintly alkaline Faintly alkaline O'oog O'OOI 0-017 0'002 o-ogg 0-014 32-5 31-5 i4'0 9-0 .. i8-5 Marked 22-5 Nil charring 20-5 22-0 9-0 8-5 II-5 2-4 13-5 1-8 0-4 0-2 Notes. — Sample 9. The composition of river water will alwaj^s, of course, vary with the following circumstances : 1. The nature of the country through which the river courses, and which it therefore drains — i.e., whether this be cultivated and manured or wild, whether there be much or little vegetation, and whether it be thickly or sparsely populated. 2. The amount of pollution by sewage, waste products of manufactories, etc., which gain access to the water. 3. The nature of the bed of the river, and of the strata through which any springs (which feed the river) rise. 4. The rapidity and smoothness of flow — i.e., the more rapid and inter- rupted this is, the greater the powers of the river in the direction of self- purification. No. 9 furnishes evidence of contamination. No. 10 is a water of great purity. 120 LABORATORY WORK 11. 12. Jjeep spring w.iter (from Green'vand below Chalk). Spring Water (from Chalk). Physical characters . . Excellent Excellent Reaction Alkaline Alkaline Free and saline ammonia. . 0-030 O'OOI Albuminoid ammonia O'OOI 0-003 O absorbed from permanganate (in two hours at 27° C.) 0'020 O'Oig Total solid matters III'2 32-5 (a) Volatile . . 21'0 8-5 {b) Fixed 90'2 24-0 (c) Appearance on ignition . ..Nil Nil Total hardness .. i8-5 23-0 (a) Temporary' .. 7-0 i8-o {b) Permanent II-5 5-0 Chlorine 12-2 3-0 N as nitrates ..0.4 0'2 Notes. — In Sample 11 the high amounts of saline ammonia, chlorine and mineral matter so frequently present in pure waters from the lower green- sand, are shown. Sometimes the nitrates are higher than in this sample. The absence of recent animal pollution in this case is shown by the very low figure of albuminoid ammonia. Sample 12 is a very pure water. The ammonias are quite low. 13. 14. Well Water from " Carboniferous Well Water. Limestone. Physical characters E.xcellent Excellent Reaction Alkaline Alkaline Free and saline ammonia. . 0-005 O-OOI Albuminoid ammonia 0-006 O-OOI absorbed from permanganate (in two hours at 27° C.) 0-066 0-008 Total solid matters 31-9 48-5 {a) Volatile . . 9-8 17-3 (b) Fixed 22-1 31-2 (c) Appearance on ignition . . . Slight charring Nil Total hardness 24-2 31-5 (a) Temporary 17-9 9-5 {b) Permanent 6-3 22-0 Chlorine 1-9 6-2 N as nitrates 0-3 1-8 Notes. — Sample 13. The ammonia figures indicate slight animal con- tamination. Sample 14 furnishes evidence of previous (remote) sewage contamination in the high figures of N as nitrates and chlorine. The extremely low albuminoid ammonia figure shows that the organic matter has been almost completely mineralized. Nitrites were absent, but phosphates were markedly present. THE OPINION ON WATER SAMPLES 121 15. 16. Chalk Water. Peaty Water. Excellent Light brown tint; clear Alkaline Acid O'OiS 0-005 O-OIO 0-022 0-082 0-122 62-7 15-6 23-2 12-0 39-5 3-6 Marked Marked charring charring 33-5 3-0 22*5 o-o II-O 3-0 6-8 1-5 0-9 0'2 Physical characters Reaction Free and saline ammonia. Albuminoid ammonia O absorbed from permanganate (in two hours at 27° C.) Total solid matters {a) Volatile . . lb) Fixed (c) Appearance on ignition Total hardness (a) Temporary (6) Permanent Chlorine N as nitrates Notes. — Sample 15 is a chalk water polluted with animal matter, as evidenced by the high saline ammonia (along with a considerable amount of organic ammonia) and the high figures of chlorine and oxidized nitrogen. The hardness is also excessive, but this may readily be reduced to 11 parts by a water-softening process. Sample 16 is a peaty water polluted with animal matter, as evidenced by the fact that the figures of the saline ammonia, the chlorine, and of the oxidized nitrogen, are excessive for a peaty water. It is a water possessing a considerable plumbo-solvent action. 17. Physical characters . . . . . . Brackish • taste ; blue- green tint Reaction Alkaline Free and saline ammonia. . .. .. o-ooi Albuminoid ammonia . . . . . . 0-006 O absorbed from permanganate (in two hours at 27° C.) . 18. Excellent Alkaline 0-004 0-005 Total solid matters (a) Volatile (&) Fixed (c) Appearance on ignition . Total hardness {a) Temporary (6) Permanent Chlorine N as nitrates 0-050 249-2 32-8 216-4 Slight dis- coloration Very high 109-5 0-9 0-042 34'0 ii-o 23-0 Faint dis- coloration 23-5 13-0 10-5 1-9 0.3 Notes. — Sample 17 is a deep-well water in the chalk contaminated by sea water. This is evidenced by the fact that the chlorine is enormously high and magnesium chloride is abundant. The well was near the coast 122 LABORATORY WORK and, prior to the contamination, the chlorine was 4 parts per 100,000 and the total hardness 24. The water is quite unfit for domestic uses on account . of its excessive hanlness, the brackish taste, the deposit it will give rise to in boilers and kettles, and the fact that it will impair the palatability of tea, coffee, etc.; and it is altogether unsuitable for washing and cooking purposes. Sample 18 is a polluted water. The above figures would barely warrant such an opinion, but a previous analysis of the water from the same source gave the saline and organic ammonias as 0-002 and 0-003 respectively, and the chlorine and oxidized nitrogen as i-S and 0-20 respectively. Some pollution has therefore recently gained access to the water, and the sample is included to illustrate the value of periodical analyses of a water-supply in detecting intermittent pollution. An examination of the bed of a waterway or pond may serve to furnish corroborative evidence of the sewage contamination of the water. Of such contamination there is httle or no reliable indication by chemical anatyses, nor does a low-power micro- scopic examination and a bacteriological examination supply valuable evidence; and unless the pollution is gross, it is not possible to conclude that a mud is contaminated with human excrement, either from chemical or bacteriological data. If any parts of the bed are covered with gravel or large stones which are not clean, and especially if the greyish flocculent growths characteristic of certain sewage fungi are found to be attached to them or to any other part of the bed, and if the mud is of a dark colour and emits gas bubbles and offensive odour on being disturbed, then there are good reasons for suspecting gross sewage contamination. It is common in these circum- stances to find some opalescent floating bubbles, which have but little tendency to burst, on parts of the surface of the overlying \\ater. Although the organic matter in pond and river mud will be found on an ordinary microscopical examination to be largely in the form of unrecognizable debris, with only a relatively small (but variable) quantity of the vegetable structure of plant life distinguishable, a textile fibre or animal hair, etc., \\'ould indicate dangerous contamination {vide p. 155). Water markedly contaminated with sewage or sewage effluent is unlit to be used by cows for drinking purposes; for, apart from the risk to the health of the animal, there is a danger of specific organisms of intestinal origin getting upon the teats and udders of the cows, and thereby into the milk in the process of milking. Gerber and Sheldon botli agree that dirty drinking-water THE OPINION ON WATER SAMPLES 123 may give rise to impure and tainted milk. We know that improper food, such as fermented potatoes or cabbages, affect the taste and keeping quahties of cow's milk; and there is no reason why what applies to food should not apply to drink. It is, moreover, only reasonable to suppose that the drinking of polluted water is injurious to the cow as well as to the milk; and that the purer the food and water given to cows tlie better both for the animal and for the milk she furnishes. The presence of a small amount of domestic sewage in a stream of fair volume and flow is apparently not injurious to fish. Oysters, mussels, and cockles are tolerant of considerable sewage pollution, although there is evidence that oysters become scarcer and smaller in the presence of gross pollution; but in the case of these shellfish, their capacity to retain specific disease-pro- ducing bacteria when bathed in polluted water makes the con- sumption of them, when collected from such waters, a grave danger, the reality of which has been abundantly demonstrated in this and other countries. Fresh-water fish generally are more affected by pollution from chemical wastes than by sewage; but they vary considerably in their susceptibihty to sewage contamination. In experimenting upon the effect of the sewage contamination of a stream upon fish life, allowance must be made for this fact. Trout appear to be very susceptible, and they require to be kept in running water. Gold-fish, gudgeon, and roach (of which the latter two are very sensitive to various forms of pollution in water, while the former is relatively resistant) are suitable fish to experi- ment with. These may be kept in the contaminated water, while at the same time control fishes are kept in pure water; and by observation of their active movements, their food con- sumption, the healthy appearance of their eyes, fins, tails, etc., the weight of the survivors at the end of the experiment, and the rate of mortality, it is not difficult to learn from the com- parative data collected whether the polluted water has proved inimical or not. Bacteriological Evidence. Like the chemical, the value of this evidence has certain limitations. As ordinarily performed, even the bacterial counts furnish evidence only of the discrete masses of organisms. There is no necessary relationship between these and the numbers of 124 LABOKATOKY WORK separate organisms originally present, because with efflux of time the organisms, like other suspended particles, tend to agglutinate. Again, at the present day it is often impossible to reco\ier or recognize the specific germ, even shortly after this has been experimentally added to water; and the organisms which denote sewage contamination are the same whether they are derived from the lower animals or from human beings. Yet, despite these facts, the results of a bacteriological examination of water samples, interpreted in the light of topographical cir- cumstances, is generally of great value. The Colleciion and Transmission of Samples. — Great care is required in the collection of the samples; even apparently trivial errors or omissions may entirely vitiate the result. Very precise and seemingly trifling directions must be given, unless the sample is collected by an expert. For the ordinary examination 2-ounce (57 c.c.) glass-stoppered bottles are sufficient. When larger amounts are required, a Winchester quart bottle may be used. The bottles should be sterilized, with their stoppers loosely inserted, at 160° C. for one hour, and allowed to cool slowly. If the specimen cannot be examined at once, and delay is unavoidable, the sample should be packed in ice, and then trans- mitted to the laboratorj^ Special apparatus have been designed for this purpose. That figured on p. 125 is a convenient form. Two-ounce glass-stoppered bottles are used. Each of these, after thorough washing, and drying in the hot-air apparatus, has its stopper inserted, and is then placed in a tin into which it just shps. The bottom of the tin has a layer of cotton-wool and then a piece of asbestos cardboard. Several thicknesses of asbestos cardboard are also fitted in the cover of the tin, so that when in place the bottle is firmly in contact with the asbestos above and below. The tins with their contained bottles are then sterilized in the hot-air apparatus- Labels are placed on the outside of the tins, and they are ready for use. The ice-boxes are made to just receive one, two, or four such tins. The tins are not opened after sterilization until immediately before the sample is taken. To take samples from various depths, a number of different forms of apparatus have been devised. The ordinary collecting- bottle may, however, be also used for this purpose. It is tied BACTERIOLOGICAL EVIDENCE 125 into a leaden cage, and lowered to the required depth by catgut or string attached to the cage. The loosened stopper is then removed by a jerk upon a second string previously tied to the stopper, and the sample collected. In collecting samples from a reservoir, lake, or river, plunge below the surface before removing the stopper, thus avoiding scum and surface contaminations. If from a tap, allow the water to first run to waste for five to ten minutes. If from wells with a pump, pump away a considerable quantity of water FIG. l6. COLLECTING- BOTTLE AND TIN. A, Asbestos cardboard in lid; B, asbestos cardboard below bottle D ; C, cotton-wool layer. FIG. 17. ICE-BOX. A and B, felt lining; C, metal ice re- ceptacle, with depression D, to hold two collecting-tins (with contained bottles) . before collecting the sample; while if a complete investigation is required, a second sample should be obtained after several hours' pumping. Owing to the extreme difficulty of detecting the actual specific organisms of disease, such as the organisms of typhoid fever and cholera, it is necessary to resort to other methods of investigation. Hygienists are unanimous in recognizing that sewage and the excreta of human beings, diseased or healthy, must be looked upon as potential vehicles for disease production. The presence 126 LABORATORY WORK of the excreta of animals must also be looked upon as prejudicial, since it may contain harmful bacteria and other parasites. A number of organisms have been advocated as fulfilling the requirements necessary for indicators of sewage contamination. Of these, B. coli and allied organisms, B. ententidis sporogenes, and certain streptococci, are the only ones which have been extensively advocated and merit detailed consideration. For these organisms it is not only necessary to ascertain their presence or absence, but, in addition, their numbers. Significance and Interpretation of Results. — The detection of the cholera spirillum or the typhoid bacillus in a water, in whatever amount, is sufficient to condemn the water. The other results obtained in the bacteriological examination of water-supplies are, however, only data from which an opinion upon the purity or contamination of the water can be deduced with more or less confidence according to the data available. Such deductions require much special experience, and for a detailed consideration of the matter the reader is referred to Dr. Savage's book on the subject,* limits of space only allowing here a bald summar}' and review. The number of organisms developing upon gelatine plates is largely an index of the amount of organic matter in the water, although there is no constant or exact relationship between the two. Still, the addition of organic matter almost invariably means an addition both of foreign bacteria and of material which enables the water, for a time at least, to become a better nutrient medium, and so causes an increased proliferation of bacteria. A low gelatine count is, therefore, a satisfactory feature; but, on the other hand, a high gelatine count cannot in itself be con- sidered a sufficient reason for condemning a water. For surface waters the contamination is frequently with harmless organic matter, and of comparative unimportance. Good deep-well and spring waters frequently contain less than 50 bacteria per c.c. developing on gelatine plates, while in surface waters, even when free from pollution, up to 500 or more per c.c. are not infrequently met with. The blood-heat count (agar plates at 37° C.) is an index of the addition of bacteria other than those natural to pure water, but * " The Bacteriological Examination of Water-Supplies " (H. K. Lewis, London, 1906). BACTERIOLOGICAL EVIDENCE I27 they need not be harmful. The addition of harmless soil bacteria will cause a great increase in the number of the ^Y' ^^- organisms. The number present in deep-water sources, when pure, is very low, frequently less than i per c.c, and 10 or more per c.c. is not satisfactory. In the case of surface waters and rivers, soil washings are common, and a more generous margin (50 to 100 per c.c.) is necessary. On the whole, a marked increase in the number of bacteria growing at 37° C. is of greater significance than a proportionate increase of the gelatine count. Of much greater importance is the interpretation of the B. coli estimation. The views of different workers show considerable variance. This bacillus is abundant in human and animal excreta and in sewage, and it serves as a meas^ure of excretal pollution. Deep-well and spring water should not be liable to any poUu- tion by material containing B. coli. Water from these sources, even if originally polluted, must have passed through a con- siderable depth of soil, and thus have become purified from all bacterial evidence of contamination. If such sources are properly protected at their outlets, there is no reason why they should contain any B. coli. It is, therefore, justifiable to maintain an attitude of great suspicion towards any water from such sources which contains B. coli in 100 c.c. or less. In the case of surface supplies and shallow wells the position is different. For example, considering upland surface waters, the opportunity for contamination by B. coli contained in animal {e.g., sheep) excreta may be considerable. The B. coli from sheep excreta are indistinguishable from those from sewage or human feeces, yet no one would contend that they are of equal significance, or that it is equally important to prevent their presence. As a matter of experience, on the other hand, it will generahy be found that B. coli rigidly defined is not found in shallow wehs, or in the majority of surface supplies, in 10 c.c. or less, unless that water is being polluted with excrementitious matters in undesirable amount. While, therefore, admitting that dogmatic standards are especially untrustworthy for these classes of waters, a working standpoint that the finding of excretal B. coli in 10 c.c. or less points to undesirable pohution is both justifiable and in accordance with actual experience. If no B. coli are present in 50 c.c. the 128 LABORATORY WORK water ma}- be safely passed as satisfactory. For rivers used as sources of drinking-water, without artificial purification, similar standards are applicable. Sometimes the organisms isolated are not typical B. coli, but differing in the absence of one or more of the characteristic properties of this organism. In the opinion of most bacteriolo- gists of experience the nearer these lactose-fermenting coli-like bacilli approach typical B. coli in their characters, the more nearly are our numerical standards for that organism applicable to them, while if the}' lack essential characters a proportion- ately greater number must be present to justify an adverse opinion. Determinations of the number of streptococci have been made much less frequently than in the case of B. coli. As a pro- visional guide, and without attaching an equal significance to the findings, a standard similar to that for B. coli may be em- ployed — i.e., their presence in lOO c.c. or less of deep-well or spring water, or in lo c.c. or less of surface and shallow-well waters, would justify an adverse opinion as to the purity of the water in question. On its negative side the streptococcus test is not of great value, and the absence of streptococci, even in a considerable bulk of water, cannot be taken as showing purity or freedom from danger. Opinion is not united as to the value of B. enterilidis sporo- genes as an indicator of pollution. It is fairly abundant in sewage and excreta, but it is a spore-bearing organism with prolonged powers of resistance, and therefore, even if it be admitted that its presence indicates pollution, such pollution may have taken place at some long antecedent period, a con- tamination so old as to be of no significance. But its absence in a large quantity of water is some evidence of purity. It will be of assistance in the difficult matter of giving an opinion upon samples submitted for bacteriological examination if a few examples are given of samples from different sources. They represent actual analyses, in which the topographical conditions were accurately investigated either at the time of examination or subsequently. I. An upland surface water collected in an open artificial reservoir: Number of organisms developing per c.c. at 37° C.= 16. 2I°C.= 224. " Excretal " B. coli present in 40 c.c, but not in 10 c.c. or smaller amounts. Streptococci absent in 50 c.c. BACtERlOLOGICAL EVIDENCE I29 (Standard+i media; incubation lorty-two liours at 37' C, three days at 21° C.) The bacteriological opinion from this sample would be favourable. Careful topographical investigation showed no evidence of any human sources of infection on the gathering ground, but the water was liable to some pollution from sheep's droppings, etc. 2. A mountain stream feeding a large upland surface reservoir: Number of organisms developing per c.c. at 37° C. = 340. 21° C. = 1,640. " Excretal " B. coli isolated from 10 c.c, but not found in smaller quantities of the sample. Atypical B. coli isolated from i c.c. Streptococci present in 30, 10, and i c.c, but not in o-r c.c. The bacteriological evidence is here sufficient to condemn the water, and topographical investigation showed considerable opportunities for pollution from both inhabited houses and manured lands. 3. A surface (shallow well) provided with a pump: Number of organisms developing per c.c. at 37° C. = ii2. 21° C. = 7,700. " Excretal " B. coli isolated from i c.c, but not from o-i c.c. Atypical B. coli isolated from o-i c.c. Streptococci present in 40 c.c, but not in 10 or i c.c. The well is evidently polluted, and was, in fact, surrounded by manured ground, while the covering to prevent surface water gaining access was defective, and there was no internal rendering of the sides of the well. 4. A surface well provided with a pump : Number of organisms developing per c.c. at 37° C. — g. ,. .. 21° C.= 7io. B. coli absent in 50 c.c. Streptococci absent in 50 c.c. Here the examination showed no evidence of any harmful contamination. The well was situated in a town, and was surrounded by houses, but surface water was prevented from entering. From the topographical position it was impossible to say whether the water was polluted or not. 5. A spring used as a public supply: Number of organisms developing per c.c. at 37° C. = i. ,, 2i°C.=34. B. coli absent in 50 c.c Streptococci absent in 50 c.c. The bacteriological results are quite satisfactory, and, indeed, showed very little variation at each monthly examination. Careful investigation of the source showed no likely sources of contamina- tion, but remote sources of pollution were possible, and systematic bacteri- ological examinations were very valuable. N.B. — The foregoing information under " Bacteriological Evidence " is, in the main, summarized from a contribution by Dr. W. G. Savage to previous editions of this book. 9 CHAPTER XV SEA WATER The Rivers Pollution Commissioners found that sea water con- tains approximately: Parts per :oo,ooo. Total solids .. .. .. .. 38987 Chlorine . . . . . . . . 1975 "6 A specimen collected by the late Dr. Tidy during high water at Margate gave : Parts per ioo,ooo. Total solids 1. VJLCXl ;3W1H_IJ Chlorine .. 1770-5 Lime 35-1 Magnesia 205-6 Silica 0-4 Hardness . . 564-0 As edible sea shellhsh, reared or deposited round our shores, are sometimes exposed to dangerously contaminated sea water, and sea water contaminated with human excrement has on good grounds been helS to be responsible for the infection of enteric fever, it is desirable to learn what evidence is available of the sewage pollution of sea water. A summary of our present position with reference to bacterio- logical evidence of contamination will serve to indicate the value of the assistance of chemical standards. Although the Royal Commission on Sewage Disposal, appointed in 1898, reported that they were satisfied that bacteriology could not at present be relied upon to determine whether or not shellfish are polluted by sewage, typical Bacilli coli communis in shellfish are usually regarded as sufficient evidence of such pollution; and in sea water, Houston, Hewlett, Klein, and others would take B. coli communis in i c.c. as indicating contamination. Dr. Houston's 130 SEA WATER 13 1 work for the Commissioners, which was pubhshed in their Fourth Report, vol. iii., 1904, clearly shows that no sample of sea water remote from pollution contains either B. coli communis or the spores of B. enteritidis sporogenes, even when as much as 100 c.c. of the samples are used for test purposes. But birds and fish may contribute B. coli communis ; and he suggests the prudence of not pushing an extremely delicate test too far. He further demonstrates that B. coli, added to sea water, is no longer in evidence in i c.c. of the sample after a maximum period of nine days and a minimum period of five days. He therefore con- cludes that absolute standards cannot be laid down at present; but he maintains that B. coli present in 10 c.c. and absent in I c.c. should be viewed with some degree of suspicion, the water not necessarily to be condemned apart from topographical and epidemiological considerations. Such vegetable growths as Ulva latissima are not delicate indi- cators of sewage pollution, for they may not be in evidence in cases where sea water is exposed to the lesser degrees of con- tamination. Of course, gross contamination is unmistakable when sea water is judged from the results of either a bacteriological or chemical examination; but it is the evidence of previous and relatively slight contamination that may be ill-defined and elusive. Coast-tides, currents, and eddies are capable of con- veying sewage contamination for some distance from the actual outfall of sewage into the sea to parts where there is no local contamination added, and there is nothing to indicate such pol- lution. A period of twenty-four hours would suffice for this contamination, in a very dilute form, to reach several miles from its outfall; and laboratory experimentation indicates that the B. typhosus can survive several days in sea water. The writer and F. N. Kay Menzies find that the chemical evidence upon which opinions are based as to the purity or otherwise of the various classes of fresh waters is not wholly applicable to sea water after slight contamination with sewage, and they conclude as follows: While the chlorine figure is often a useful one for indicating animal contamination m fresh waters, it is useless in respect of sea water, for the reason that a relatively small amount of sewage (with an average chlorine figure of about 10 parts per 100,000), discharging into a large volume of sea water (with a 132 LABORATORY WORK chlorine figure which may vary from i,6oo to over 1,900 parts per 100,000), has not sufficient effect upon the chlorine figure of the sea water to furnish e^ddence of sufficient delicacy. The oxidized nitrogen figure is even more serviceable than the chlorine as a clue to the previous animal contamination of fresh waters ; for in fresh waters a little (say i per cent.) of sewage contamination leads to a rapid appearance of nitrates; but when sewage effluent with already formed nitrates is added to a fresh water, there is generally an initial reduction of the nitrates (often lasting for two or three weeks), or the figure may remain practically stationary for the first few days, and then a rise set in until a constant figure is arrived at. But it is remark- able how often in fresh waters no evidence is to be obtained of the presence of the intermediate or nitrite stage of the develop- ment of nitrates; and, when appreciable, how faint the evidence often is. In polluted sea water, however, the evidence of oxidized nitrogen may not be available; for under ordinary conditions, up to a period of several weeks, no oxidized nitrogen may appear in sea water as the result of sewage contamination; but it ultimately appears in amounts which give very definite reactions by qualitative tests, and more especially is this true of nitrites. When sewage effluent already containing nitrates and nitrites is added to pure sea water the nitrates disappear in a day or two, though a trace of nitrites may persist for much longer. Therefore, as nitrites may be in evidence in polluted sea water where nitrates are not, they furnish the better evidence of sewage contamination. Thus oxidized nitrogen in sea water will indicate contamination which is either very recent or very remote; but its absence is no guarantee of freedom from such contamination in a dangerous form, and may indeed be even more significant of danger than its presence in sea water which is believed to have been recently contaminated. Turning next to the free and albuminoid ammonia figures, which form such a useful indication of animal contamination in fresh waters. When small proportions (i per cent.) of sewage are added to fresh water, there is to be noted an increase of ammonia — generally slight — for the first few days; then a reduction sets in, and after several days or several weeks (accord- ing to season and dosage) this evidence of contamination has disappeared. When sea water is similarly contaminated a slight preliminary increase is to be noted for two or three weeks, SEA WATER 133 and then reduction generally sets in slowly. Therefore the free ammonia figure is a very valuable clue to contamination of sea water, forming invariably an item of evidence which, starting at the actual time of contamination, persists for several weeks. When small proportions of sewage are added to fresh water, an increase in the albuminoid ammonia figure is to be noted, which often persists for several weeks, so that the figure may reach one several times greater than the original figure; then an irregular fall sets in through many weeks. But in ^ea water we find that the preliminary increase is less rapid, the original figure being generally found after two or three weeks, and later there is some reduction. When this figure in pure sea waters is compared with the free and saline ammonia figure, it is found that not only is it always a much higher figure, but that it is subject to far greater variations. With these more indefinite characteristics the figure does not lend itself as a basis for com- puting the lesser degrees of sewage contamination. Pure sea water has a considerable reducing action upon potassium permanganate under the conditions of the processes employed in water analysis. The oxygen absorbed figure in pure sea water generally approximates to 0-5 part per 100,000 in four hours at the temperature of the laboratory, and careful analysis may furnish no appreciable difference when sea water is polluted with i per cent, of sewage effluent. It is obvious, therefore, that this process does not assist us in determining the presence of the lesser degrees of contamination of sea water. Nor does the dissolved oxygen in slightly contaminated sea water furnish reliable results, for the figure falls with the time which has elapsed since the sample was taken ; and the differences between pure sea water and sea water contaminated with i per cent, of sewage effluent are often so slight as to be unserviceable. Phosphates are absent from pure sea water, and their presence is valuable corroborative evidence of contamination. The evi- dence, however, may be obscure where the contamination is slight; but working with 100 c.c. of sea water, the ammonium molybdate reaction is generally appreciable when sea water is contaminated with quite small proportions of sewage. Interesting and suggestive as are the results of laboratory experiments extending over many weeks, one may not argue from the results of sewage contamination of stationary sea water employed in experiments to sea water moving under the in- 134 LABORATORY WORK iluence of tides, eddies, and currents. To keep to the practical issues of the problem it is necessary to observe the behaviour of sewage contamination during a period of at most several days instead of several weeks; and it is upon these considerations that the following conclusions upon the chemical evidence of slight sewage pollution of sea water are based: We may conclude that the free and saline ammonia figure furnishes the only reliable chemical guide to the lesser degrees of animal contamination of sea water; that delicate corrobora- tive evidence ma}' sometimes be obtained from the presence of nitrites and phosphates; but the complete absence of oxidized nitrogen is compatible with recent pollution, and it is not always easy to obtain a definite reaction for phosphates when the contamination is but slight. The free and saline ammonia figure remains of all the available tests hitherto suggested the most reliable and the most delicate; and an ammonia figure much exceeding 0-002 part per 100,000 is certain evidence of the sewage contamination of sea water. Samples of Pure Sea Water, obtained from Points at which the Water was judged to be Free from Pollution: (Parts per 100,000). Neighbourhood where Sample Free and Saline Albuminoid 1 Nitrogen as Nitrates and collected. Ammonia. Ammonia. Nitrites.* Aberdeen O'OOI 0-007 COO Bournemouth . o-ooi8 0-002 o-oo Carnarvon 0'Ooi5 0-006 o-oo Clacton o-ooi 0-004 o-oo Folkestone 0-002 0-004 o-oo Hastings 0'00i5 0-003 o-oo Ilfracombe 0'00i5 0-0075 o-oo Oban . . O-OOI o-oog o-oo 1 Scarborough . O-OOI 0-005 o-oo Ventnor O-OOI 0-016 o-oo There is also available the method of taking several samples for comparative purposes and of judging the presence and degree of contamination in any particular sample from any observed variations — more particularly, in the case of sea water, in the * Professors E. A. Letts and E. H. Richards find that a trace of nitrates (averaging 0-005 N per 100,000) may be demonstrated in pure sea water, by adding to 25 c.c. of the water a few drops of brucine sulphate solution and 25 c.c. of strong HgSOj. SEA WATER 1 35 ammonia and oxidized nitrogen figures. Tin's method will often be serviceable, and should always be availed of wlienever it is possible to obtain fair control samples from situations obviously more remote from any source of contamination. In analyses of sea water, the nitrates should be tested for, qualitatively, by brucine and sulphuric acid, and the nitrites by Ilosvay's method. The oxidized nitrogen should be esti- mated quantitatively by the wet copper-zinc-copper process. The oxidizable organic matter, by Tidy's process, conducted at the laboratory temperature; and the dissolved oxygen by Winkler's process. The phosphates should be tested in lOO c.c. of the sample by first adding strong nitric acid, and then evapor- ating to a solid residue; the residue to be then digested in strong nitric acid and the filtrate tested with molybdic solution. CHAPTER XVI ALKALIMETRY AND ACIDIMETRY— ICE— MINERAL WATERS —ANALYTICAL SCHEMES Alkalimetry and Acidimetry. For the purposes of estimating the degree of alkahnity or acidity of water, it is convenient to use standard solutions based upon the atomic or molecular weights of the different reagents; or made up so that equal volumes of the solutions are chemically equivalent to each other. Such solutions are "normal" when they contain in i litre at i6° C. chemically equivalent weights of the active reagents weighed in grammes, hydrogen being taken as the unit. There- fore the normal solution of hydrochloric acid must contain the molecular weight of the acid — i.e., 36-46, in grammes per litre, since HCl is a univalent substance. The normal solution of sodium carbonate (NagCOs) must contain the molecular weight of the salt {(23 x 2) +12 + 16 x 3)} divided by 2=53 grammes per litre; for the molecule of monobasic HCl can neutralize only half a molecule of the bivalent NaaCOg. But the hydrogen equivalent of some reagents is not so easily arrived at. Take, for instance, potassium permanganate; the molecular weight of the formula (KoMuoOg) is 316, and the normal solution is 31-6 grammes per litre. This is because 316 grammes of permanganate of potash liberate 80 grammes of oxygen, which are chemically equivalent to 10 grammes of hydrogen; and so 31 "6 grammes of the salt are equivalent to i gramme of Hydrogen. ^ ^^^^^ ^g^^ " Seminormal " and " decinormal " solutions are obviously those made up to h and yV, respectively, of the strength of 136 ALKALIMETRY AND ACTDIMETRY 137 the " normal " solutions. They are commonly expressed as ^ and J^f solutions. Thus each c.c. of a normal solution of HCl will contain j^Vo o^ t he molecular weight of t he acid in grammes {i.e.,o- 03646 gramme) , and each c.c. of a decinormal solution will contain 0-003646 gramme. Therefore measured quantities of normal and decinormal acids should exactly neutralize similar quantities of the normal and decinormal alkalies; and if, on titration, they are not found to quite correspond^ the difference must be ascertained and a simple calculation made in order to correct it. In estimating alkalinity the decinormal solution of hydro- chloric acid may conveniently be employed, and for acidity the decinormal solution of sodium carbonate. In either case one of these standard solutions would have to be added in measured quantity until the neutral stage is exactly reached, as indicated by a suitable reagent which is added to the solution. Methyl-orange (about i gramme to the litre) is a good " indi- cator " where the alkalinity of water is being tested. This sub- stance has the property of yielding a beautiful scarlet colour in the presence of acidity; but its solution, which is of a bright orange colour, must not be emplo3^ed where organic acids are concerned or where nitrites are present. In these cases phenol- phthalein may be substituted; but not if free carbonic acid is present. Phenolphthalein dissolved in 50 per cent, alcohol is a colourless solution which strikes a rose-red colour in the presence of alkalinity; but is colourless in acid solutions. The marked presence of ammonium salts would vitiate the results when phenolphthalein is employed. Litmus should not be used as an " indicator," for the COg, so commonly present in water, considerably masks its indications. A solution of cochineal is almost free from this drawback, and is a useful indicator; it is prepared by digesting the dried and powdered cochineal in warm water to which a little alcohol has been added, and then filtering; the solution has a yellow or yellowish-red colour, which is turned violet-red by alkalies, and the original colour is restored by mineral acids. A I per cent, solution of rosolic acid in dilute alcohol is a delicate indicator. Even COg and acid salts change the indicator to a yellowish tint, the colour being rose in alkaline solu- tions. 138 LABORATORY WORK Example. — It is desired to estimate the alkalinity of a water sample. A few drops of the methyl-orange " indicator " are added to 100 c.c. of the water in a white porcelain dish. A deci- normal (or centinormal) solution of HCl is then dropped in from a graduated burette until evidence of a scarlet tint appears, denoting all alkalinity to be neutralized. It took 6 c.c. of the decinormal acid to effect neutrality, therefore the alkalinity is equivalent to 6 c.c. of this acid solution. But 6 c.c. of the decinormal HCl is equivalent to a similar amount of deci- normal sodium carbonate solution; therefore the alkalinity is equivalent to 6 c.c. of decinormal sodium carbonate solu- tion. But I litre of the normal solution contains 53 grammes of sodium carbonate, therefore i litre of the decinormal solu- tion contains 5-3 grammes, and i c.c. of this contains 0'0053 gramme, and 6 c.c. contain 0-0318 gramme of sodium carbonate. Therefore the alkalinity of 100 c.c. of the solution is equivalent to 0-0318 gramme of sodium carbonate, or 31-8 parts of sodium carbonate per 100,000. In estimating the aciditj' of a peaty water, the " indicator " — one or two drops of an alcoholic solution of phenolphthalein — is added, and the sodium carbonate decinormal solution run in until a very faint pink tint is obtained, when the calculation is made as above. To prepare the normal HCl it is necessary to take about 181 grammes of liquid acid of the S.G. i-io, and dilute to a litre with water; then titrate the exact strength with normal sodium car- bonate. Ice. Both artificial and natural ice are liable to furnish on analysis considerable evidence of pollution, which, since pathogenic organisms can survive in ice for long periods, must be regarded as significant of danger. Under natural conditions the most super- ficial layer of the ice contains most impurity. The popular belief that water purifies itself by freezing is unfounded. It is certainly not borne out when the water obtained from the melted ice is subjected to chemical analysis and bacteriological examination. MINERAL WATERS 139 The results of many analyses performed in this country, America and the Continent, show the following variations: Parts per 100,000. Free and saline ammonia . . . . from O'OOi to 0'32 Albuminoid ammonia .. .. ,, "002 to 0*44 Chlorine as chlorides . . . . ,, o"i to 6-5 N as nitrates . . . . . . ,, nil to traces Total solid matter . . . . . . ,, i to 50 Bacteria per c.c, 40 to 2,000. Mineral Waters. The examination of mineral waters for public health purposes should be conducted on precisely the same lines as those of an ordinary water analysis; that is to say, an effort must be made to ascertain the freedom of the water from dangerous organic and metallic contamination. Artificial mineral " waters " consist of water into which car- bonic acid is forced under pressure. Lithia water is, in addition, charged with lithia; potass water with bicarbonate of potash; soda water is commonly sold without the addition of any soda, but when such is added it is usually to the extent of about 10 grains of bicarbonate to the pint. Natural mineral waters are generally the purest. Those which are chalybeate mostly contain the iron in the form of ferrous carbonate, held in solution by excess of carbonic acid, such as those at Tunbridge Wells, Spa, and Cheltenham. Instances of alkaline waters naturally charged with carbonic acid, and con- taining sodium carbonate and bicarbonate, are found at Carlsbad, Ems, Malvern, Nieder-Seltzers, and Vichy. At Harrogate and Aix-la-Chapelle waters are found naturally charged with sul- phuretted hydrogen. Those waters which possess a marked aperient action generally owe their properties to either sulphate of magnesia, as at Epsom and Leamington, or to sulphate of soda, as at Cheltenham and Scarborough. In Central Wales there is a deep spring containing 9 parts of barium chloride per 100,000. If the sample is collected from a well and it is desired to know the temperature of the water, the thermometer should be let down in a stout glass bottle ; this will come up fiUed with the water. 140 LABORATORY WORK and a reading of the thermometer when surrounded by the water can be taken. All artificial mineral waters should be tested for lead, iron, copper, zinc, and arsenic. Each of these metals has been found in samples of soda water, etc., to which the metal has gained access either by the apparatus used, the improper washing of the carbonic acid, or from the use of metal taps to the syphons. Sometimes a considerable quantity of lead is present and very impure water is used; on this account it is desirable that efficient supervision should be exercised over their manufacture. Lemonade and ginger-beer are also liable to contain traces of lead, derived from the apparatus, and, in the case of the former, from the impure tartaric acid emploj'ed. The sediment sometimes yielded by mineral waters after long storage generally consists of hydrated ferric oxide, alumina, silica, and calcium carbonate. The carbonic acid of aerated waters is unfavourable to germ life, and the bacteriological counts are generally low. In aerated waters the large amount of carbonic acid inter- feres with the estimation of the free and saline ammonia by Nesslerization, and must therefore be remo\-ed as follows: The ammonia should be fixed with 10 c.c. of normal sulphuric acid, then the water is heated to drive off the carbonic acid, and after neutralizing the acid with 10 c.c. of normal sodic hydrate, the ammonia may be distilled over and estimated. \\nien it is found necessary so to deal A\-ith carbonic acid, a blank experiment should be performed, in which any ammonia found in 500 c.c. of ammonia-free distilled water containing 10 c.c. of normal sulphuric acid and 10 c.c. of normal sodic hydrate is distilled over and estimated, and this is deducted in arriving at the figure of the free and saline ammonia in the sample. Scheme for Effecting an Analysis in the Quickest and Most Convenient Manner. Although a little confusion may be experienced at first, yet, after a little practice, the following plan will be found practical and expedient, the time required being about three hours. I. Start the process for the estimation of the oxidizable organic matter. ANALYTICAL SCHEME I4I 2. Start the evaporation for the estimation oi the total solids. 3. Start the concentration of the water for the purpose of testing for poisonous metals, etc. 4. Start the distillation for the estimation of the free and saline ammonia. 5. Start the water boihng for the estimation of temporary and permanent hardness. 6. Start the alkaline permanganate boiling in preparation for the second stage of Wanklyn's process. 7. Apply qualitative tests for nitrates, nitrites, sulphates, and phosphates. 8. Start the quantitative estimation of oxidized nitrogen (picric acid process). 9. Make the quantitative estimation of chlorine. 10. Make the quantitative estimation of total, temporary, and permanent hardness. (By this stage the free and sahne ammonia will be over, the alkahne permanganate may be added to the boihng-fiask, and the distillation for the albuminoid ammonia started.) 11. Estimate the free ammonia. 12. Care has been taken throughout not to disturb any deposit which may be present. Now collect and make an examination of any sediment or suspended matter. Note the physical char- acters in the 2-feet tube. 13. Estimate the albuminoid ammonia. 14. Complete the picric acid process. 15. Estimate the oxygen absorbed. 16. Test for poisonous metals, and estimate quantitatively if any one is present. 17. Complete the estimation of the total, volatile and non- volatile solids. PART II SEWAGE AND SEWAGE EFFLUENTS ,« collecting samples of sewage or sewage effluents for analyst the average sample should be obtained m each case. This may * done by mix,ng together the hourly samples taken through- out the day, and these should vary m bulk m proportion to the flow of sewage or effluent at the tune when each sample is taken, S::re,usingbottlesofonesi.eforcollectinga,lthe sample. the bottle is filled if the maximum flow of sewage is observed at the time the sample is taken; if half the maximum, the bottle is half ailed, and so on. When aJl these samples are mixed together, the mixture wiU fairly represent the average romposltTn; akd a part of this should be taken for the "itf of the greatest importance that the an^ysis should be perfo med as soon after the collection of the sample as possible 'or important changes may be rapidly brought about by the teer^ing micro-organisms present. About J litre is he amount reqS for a complete analysis. The bottle should be quite S with the sample to be analyzed and i more than a day passes before the analysis is undertaken it should be kept in a cold chamber in the interval. The estimations made in the analysis of sewage and sewage effluents include: The free and saline ammoma the albuininoid ammonia, the chlorine, the oxidized nitrogen, the to al -ohds m soteion volatile and non-volatile), and the suspended matter m estimation of the dissolved oxygen absorbed by an effluent is displacing that of the oxidizable organic matter which was fortriy so'genetally employed. In addition so '- as effluen are concerned, the physical characters are noted, and not infre nuently an incubation test (at about 27° C) is applied, m order ?o seelf the sample develops odour at the end of a day or two. 144 LABORATORY WORK As a general rule, a sample of effluent is allowed to deposit before the quantities are removed for analysis; and as in samples of Sewage the matter remaining in suspension is considerable and leads to variable analytical results, two or more analj^ses should be performed of the sample, and the mean taken of the different figures obtained; and the solids maj^be estimated both before and after the sample has been shaken, and the two results given. The results of the anah-sis should be expressed in terms of parts per 100,000 to the second place of decimals; analyses which record a third or fourth place of decimals give a fictitious ap- pearance of accurac}^ when such a changeable substance as sewage or as sewage effluent is concerned. The anah-sis proceeds upon similar lines to those of water analj-sis, but it is necessary to dilute the sewage matter to a considerable extent before commencing certain of the estima- tions. So far as the solids are concerned, these may be estimated from the original sewage or effluent, but for the calculation of the two ammonias 20 c.c. of sewage effluent and 10 c.c. of sewage should be made up to the litre with ammonia-free distilled water, and the results obtained will represent the amounts in 20 c.c. and 10 c.c. respectivel^^ In Nesslerizing the ammonias the contents of the Nessler glasses may be all mixed together in a beaker, and the colour of 50 c.c. matched; thus, supposing 200 c.c. of the distillate were collected before all the free " ammonia " had come over, then the ammonia estimate in 50 c.c. must be multiplied by 4. For Tidj^'s process 20 c.c. of effluent and 10 c.c. of sewage should be added to 100 c.c. of distilled water. The nitrates may be estimated from the original effluent by the phenol-sulphonic acid method. For the estimation of the chlorine the effluent should generally be first diluted with an equal quantity of distilled water. The amount of chlorine is practically unaffected by the usual methods of sewage purification, and if its quantity is not added to by the w^aste liquors from manufacturing processes (ferrous chloride — iron pickle, or salt) it furnishes a useful clue to the strength of the original sewage. About 10 parts per 100,000 may be taken to indicate sewage of average strength. It is generally necessary to fflter the sewage, and occasionally the effluent, before making this quantitative estimation. In performing Tidy's process it will often be noted that the solution becomes decolorized at the bottom first. The flask should, therefore, be shaken from time to time. Various coal- sewaCxp: and si<:wage effluents 145 tar products, indigo, logwood and other dyes, and such salts as thiocyanates, sulphites and sulphides, will also absorb oxygen in addition to oxidizable organic matter. In performing Wanklyn's process, if the ammonias come over very slowly, so that the water in the boiling-flask reaches below 150 c.c, then a further 100 c.c. of ammonia-free distilled water should be added to the boiling-flask. In the case of sewage or bad effluents it is sometimes difficult to get all the albuminoid ammonia over without adopting this expedient. In the estimation of N as nitrates the picric acid process is to be preferred, but this method may yield unsatisfactory results if the effluent contains waste gas liquors. In such a case the effluent should be diluted, and the wet copper-zinc couple method employed. The amount of suspended matter will indicate the amount of deposit hkely to take place in a stream. It may be estimated with sufficient accuracy by collecting it from 500 c.c. of effluent or 100 c.c. of sewage, on previously well-washed, dried, and weighed fine hard filter-papers. The albuminoid ammonia figure is a fair indication of the amount of organic matter, but the organic nitrogen, as estimated on Kjeldahl's principle, is a much more inclusive estimation than the albuminoid ammonia, and it is almost as easily arrived at Although it will be found that the organic nitrogen of Kjeldahl's process averages a little over twice the nitrogen of the albu- minoid ammonia (though sometimes showing marked departures from this average), the fact that the two analytical figures do not bear a constant ratio to one another is significant, and points to the desirability of adopting the more inclusive estimation. The presence of oxidized nitrogen in an effluent must not be regarded as insuring absence of odour, although if nitrates are found to persist in an inoffensive effluent for a few days after its collection, the effluent is not likely to become offensive. Nitrates are a measure not of that poUution which may be oxidized, but of that which has been oxidized, and their presence often gives little indication of what remains to be purified; but generally high nitrates are a good feature in a sewage effluent. 146 laboratory work Organic Nitrogen by Kjeldahl's Method (Modified). I., Place 20 c.c. of the sewage or sewage effluent into a small flask, and after adding about i c.c. of sulphuric acid evaporate over the water-bath to about 5 c.c. 2. Add 20 c.c. of pure concentrated H2SO4, close the mouth of the flask by a small glass funnel, and boil slowly for an hour or two until the solution is of a clear, pale yellow colour. 3. Let cool, and then transfer the contents of the small flask to a distilling flask, being careful to well wash out the small flask and to transfer the washings to the distilling flask. 4. Make up the bulk of liquid to about 500 c.c. with ammonia- free distilled water, and then neutralize the acid with excess of strong potassic hydrate solution. The amount necessary to add can be determined by previously ascertaining how much is required to neutrahze, sa3^ 22 c.c. of the concentrated acid in water. 5. Heat a piece of pumice-stone to bright redness, and drop it into the flask to prevent " bumping." Distil over about 400 c.c, receiving the first portion of the distillate into a flask containing 20 c.c. of ammonia-free water slightly acidulated with two drops of dilute sulphuric acid, the distillate being received direct into this solution. 6. Nesslerize the ammonia; and [i of this will be nitrogen. 7. Deduct the amount of nitrogen as " free and saline am- monia " in 20 c.c. of the original sewage or effluent (previously ascertained by Wanklyn's process) and the difference = the organic nitrogen. Calculate to parts per 100,000. 8. Always make a blank experiment to determine the amount of ammonia thus obtained from the reagents and distilled water emploj^ed, and deduct this amount. Note. — By the action of the sulphuric acid the nitrogen of the organic matter is converted into sulphate of ammonia. The potassic hydrate liberates the ammonia, which is then distilled over. The above is a simple method of estimating the organic nitrogen, giving results which vary but little from those obtained from working with the solid matter of a litre of water. In the latter case a few grains of yellow mercuric oxide and of potassium sulphate should be added to the solids along with the sulphuric STANDARDS OF PURITY OF SEWAGE EFFLUENTS I47 acid, and, before distillation, 50 c.c. of potassium sulphide solution (40 grammes to the litre) should also be added. The composition of sewage from the same district varies very greatly from time to time, and this is true, though to a less extent, of the effluent. The effluent, moreover, varies according to the method and degree of treatment ; but the following mean results of many analyses will serve to give the reader a general idea of the composition of domestic sewage and sewage effluents: Free and saline ammonia Organic ammonia . . O absorbed in two hours at 27° C. Nitrogen as nitrates and nitrites Chlorine Suspended matter . . Solids in solution (a) Volatile (6) Non-volatile Sewage as it leaves the Outfall Sewer. Parts per 100,000. 6-5 1-4 9-6 CO ii-i 50 90 50 40 Effluent after the .Sewage has been chemically treated and then passed through Filter-beds. Parts per 100,000. 1-50 0'i4 1-35 I'20 IO-5 2'0 75 32 43 Often it is desired to obtain an expression of the amount of purification obtained by a given method of treatment ; it is then usual to calculate the percentage amounts of purification effected from the differences between the " albuminoid ammonia " and " oxygen absorbed " figures of the original sewage and those of the effluent. For this purpose the figures of the original sewage are taken as 100. Example. — ^The albuminoid ammonia of the original sewage = 0-8 part per 100,000, and that of the purified effluent = 0-2. The percentage purification would therefore be — ^ =75- Standards of Purity of Sewage Effluents. A satisfactory sewage effluent must be without faecal odour, and should possess little colour or turbidity. It has been sug- gested that pearl type should be readable through a column of effluent 10 inches in depth. In parts per 100,000 the figure of albuminoid ammonia should not exceed o-i or 0-15, nor the oxygen absorbed by oxidizable organic matter in two hours at 27° C. a figure of 1-5. The final effiuent must not be liable to 148 LABORATORY WORK putrefaction or secondary decomposition; and, if satisfactor}^ all frothing will disappear in tlu-ee seconds after a half-lilled bottle is shaken vigorously for one minute. The Royal Commission on Sewage Disposal, appointed in 1898, found as follows: The harm caused b}^ allowing unpurified, or imperfectly purified, sewage to flow into rivers and streams may be placed under one or more of the following headings: The de-aeration of the water of the river, and consequent injury to fish; the putrefaction of organic matter in the river to such an extent as to cause nuisance; the production of sewage fungus and other objectionable growths; the deposition of suspended matter, and its accumulation in the river-bed or behind weirs, which will draw upon the oxygen in the supernatant water; the discharge into the river of substances, in solution or suspension, which are poisonous to fish or to live-stock drinking from the stream; the discoloration of the river; and the discharge into the river of micro-organisms of intestinal derivation, some of which are of a kind liable, under certain circumstances, to give rise to disease. The extent to which the purification of a sewage need be carried varies with the particular circumstances of the town and river concerned, and they recommend that local circumstances should be taken into account. The effect of an effluent on a stream does not generally depend on the absolute amount of organic matter contained in it, but rather on the nature and condition of that organic matter; and the important thing to ascertain in examining an effluent is the extent to which the contained organic matter has undergone fermentation. In the Eighth Report of the Royal Commission on Sewage Disposal the Commissioners deal with the question of the stan- dards to be applied to sewage and sewage effluents discharging into rivers and streams, and the test which, in their opinion, should be used in determining those standards. The Commis- sioners reiterate a previous recommendation that Local Authori- ties should not be required to purify their sewage more highly than is necessary to obviate the risk of actual nuisance (from odour, growths, putrefying solids, and detriment to fish life) arising from its discharge. They express the view that an effluent ought not to be considered alone, but that the nature and volume of the recipient waters should always be taken into consideration ; and that any standard laid down ma}- be either a STANDARDS OF PURITY OF SEWAGE EFFLUENTS I49 general standard or a special standard which will be higher or lower than the general standard, as local circumstances require or permit. They find that the nuisance-producing power of an effluent is broadly proportional to its deoxygenating power on the stream; and it is recommended that an effluent, in order to comply with the general standard, should not be permitted to contain more than 3 parts per 100,000 of suspended matter, nor with the suspended matters included should the effluent take up more than 2 parts per 100,000 of dissolved oxygen in five days when it is maintained at a temperature of 18° C. The Commissioners classify rivers as follows: " Very clean " if no more than o-i part of oxygen is absorbed. " Clean " if no more than 0-2 part of oxygen is absorbed. " Fairly clean " if no more than 0-3 part of oxygen is absorbed. " Doubtful " if no more than 0-5 part of oxygen is absorbed. " Bad " if more than i-o part of oxygen is absorbed. The above figures represent the parts of dissolved oxygen taken up by 100,000 parts of the water in five days at 18° C. " If 100,000 c.c. of river water (containing effluent) do not take up more than 0-4 gramme of dissolved oxygen in five days, the river will ordinarily be free from signs of pollution, but above this figure it wih almost certainly show them"; and "if the mixture should yield a figure exceeding 0-4, then the effluent would have to be improved." " This figure (0-4) we term the ' limiting ' figure, and in our opinion it should be the foundation upon which any scheme or standard should be constructed in order to ascertain the minimum degree of purification which would be sufficient to obviate risk of nuisance." The albuminoid ammonia and acid permanganate tests only yield empirical results which do not sufficiently indicate the con- sequences which follow when effluents are discharged into streams; acid permanganate is too vigorous an oxidizing agent; and the amount of dissolved oxygen taken up from water by an effluent is doubtless a test which possesses an advantage as an indication of what naturally takes place. In the experience of the Commissioners, if the dilution of the effluent, while not falling below 150 volumes, does not exceed 300, the dissolved oxygen absorption test may be omitted, and the standard for suspended solids fixed at 6 parts per 100,000. 150 LABORATORY WORK When the dilution, while not falling below 300 volumes, does not exceed 500, the standard for suspended solids may be further relaxed to 15 parts per 100,000. Lastly, with a dilution of over 500 volumes, the Commissioners conclude that all tests ma}' be dispensed with and crude sewage discharged, subject to such conditions as to the provision of screens or detritus tanks as may appear necessary to a Central Authority. It is not possible, nor is it necessary, to laj^ down fixed standards with reference to tidal waters. Some effluents, especiall}^ those containing iron salts, while free from visible suspended solids at the time of sampling, are capable of yielding considerable deposits on standing. There- fore, as suspended matter is liable to separate out of true and FIG. iS. APPARATUS FOR ADENEY'S PROCESS. colloidal solution on standing, anj^ delay in making the estima- tion is important. Furthermore, by delay, the suspended matter may lead to a material reduction of the oxygen in the effluent. To determine the amount of oxygen which a sewage effluent will absorb in a given time, a definite quantity of the effluent is allowed to remain in contact either with atmospheric air or with water containing a high figure of dissolved oxygen. For the former test the following method is satisfactory : Adeney's Method for the Determination of the Rate of Absorption of Dissolved Oxygen in Polluted Waters. — A known volume of the effluent or polluted water — 100 to 250 c.c, according to its strength — is decanted into the bottle B, into which a little freshly precipitated magnesium hydrate has been previously added for SEWAGE AND SEWAGE EFFLUENTS 15^ the purpose of fixing the carbonic acid in the water. A similar volume of the distilled water is poured into the bottle A. Similar volumes of air will thus have been left in the two bottles. The corks with the connecting tube and stopcocks in position arc then fitted into their respective bottles, care being taken pre- viously to open both stopcocks. A slight rise of water from capillary action will of course occur in the portion of the con- necting tube in the bottle A; the height to which it rises in this way should have been previously marked by a writing diamond. This mark serves as an index for subsequent measurement. The two bottles so connected, and both stopcocks being still open, are completely immersed in the water-bath for a few minutes to allow of their contents assuming a common temperature. Both stopcocks are then closed, and at the same time the temperature of the bath and the height of the barometer are noted. The apparatus is taken out of the water-bath and dried, especiaUy the corks, which when completely dry are coated with shellac varnish to prevent diffusion of air through them. It is then put into a mechanical shaker, by means of which the contents of the two bottles are kept in gentle motion. As oxygen is absorbed by the polluted water from the atmo- sphere in B, the pressure of the atmosphere wih be reduced relatively to that of the atmosphere in A, which will be unaffected by the distilled water. Consequently, the water from A wiU rise in the connecting tube b in proportion to the volume of oxygen ■absorbed by the polluted water from the atmosphere in B. ^ The volume of oxygen which is indicated by the rise of the water in the connecting tube can be measured at any time by attaching, by means of a flexible tube, a burette containing distihed water at the temperature of the laboratory. As the water from the burette is cautiously allowed to flow into B, the water in the connecting tube wiU gradually sink back to the index, when the stopcock to the bottle B is closed; and the volume of water which has flowed from the burette is equal to the volume of oxygen which has been absorbed from the atmosphere of B at the temperature and pressure of the atmosphere obtaining at the commencement of the experiment. It may here be noted that the distilled-water bottle A acts as a reference pressure bottle. If a comparatively rapid absorption of oxygen occurs during the first hour or two, and this is followed by a slower and regular absorption, it may safely be taken to be due to the poUuted water 152 LABORATORY WORK being de-aerated to start with, and possibly also to the presence of easily and directly oxidizable substances in it; the subsequent slower and regular absorption being due to indirect oxidation accompan}'ing the fermentation of the polluting matters. This apparatus may also be employed for determining the strength of a crude sewage or of any other polluted water in the crude state, after the solids in suspension have been separated. The sewage should, liowever, be diluted with four to nine N'olumes of tap water. The errors affecting the estimation are negligible, provided care be taken at the commencement that the \\'ater and air in the two bottles are at a common temperature and that the apparatus is airtight in all parts. Under many conditions of work it will be found preferable to employ some of the polluted water to be examined instead of dis- tilled water for use in the reference pressure bottle A, after steril- izing it by the addition of a few drops of a concentrated solution of mercuric cliloride and shaking it with air in order to thoroughly aerate it. It is sometimes advantageous that the bottles employed should be larger than those indicated in the diagram, and should be of about 1,000 c.c. capacity. The Dissolved Oxygen Absorbed in Five Days. The method adopted by the Royal Commission for this deter- mination is that of Winkler, as modified by Rideal and Stewart, the particulars of which are as follows: Reagents required : 1. Concentrated sulphuric acid. 2. Concentrated hydrochloric acid (free from chlorine). 3. ^ permanganate (3*94 grammes KMNO4 per litre). 4. Potassium oxalate (2 per cent, of the crystallized salt). 5. Manganous chloride (33 per cent, of the crystallized salt). 6. A mixed solution of caustic potash and iodide of potassium, containing 70 grammes KOH and 10 grammes KI per 100 c.c. 7. Sodium thiosulphate solution, containing about 12 grammes of the salt per litre. The Process. I. Well shake the effluent, in order to bring its dissolved oxygen content to something near that of the diluting water; wait a few minutes, and then measure 300 c.c. and gently mix with four SEWAGE AND SEWAGE EFFLUENTS 153 times its volume of well-aerated tap water at i8° C. (This water will contain in solution about 7 c.c. of oxygen per litre, or i part by weight of oxygen per 100,000 of watei.) 2. Quietly and quickly fill four small bottles (capacity 340 or 360 c.c.) with this mixture, the bottles being allowed to stand full to the mouth for five minutes, and then stoppered. 3. Place two of the bottles in an incubator at 18" C. for five days. 4. Determine at once the dissolved oxygen in the water of the other two bottles, as follows: {a) First add o-g c.c. of sulphuric acid and then sufficient of the per- manganate to still provide a pink colour after twenty minutes. I to 2 c.c. of |- permanganate are generally sufficient for this purpose. Mix the contents, and let stand for twenty minutes. [b) Remove the excess of permanganate by the addition of about i c.c. of the oxalate solution ; restopper and mix. (c) When the liquid has become colourless,* i c.c. of manganous chloride solution is run into the bottom of the bottle, followed imme- diately afterward by 4 c.c. of the solution containing potassium hydrate and potassium iodide. {d) Insert stopper, and turn over the bottle once or twice; let stand for a few minutes; again turn over once or twice, and then allow the hydroxides of manganese to settle. {e) Add 5 c.c. of the hydrochloric acid, restopper and let bottle stand in shade for five to ten minutes, with occasional rotations. (/) Twenty c.c. of the liquid are now pipetted out and rejected, and the remainder is titrated with thiosulphate, as described on p. 88. {g) The second bottle is treated in the same way, and the mean of the two results is taken. {h) At the end of five days the dissolved oxygen in the incubated bottles is determined, and the mean of the two results is subtracted from the first mean. The difference multiphed by 5 gives the amount of dissolved oxygen absorbed by 100,000 parts of the effluent in five days. Example : The dissolved oxygen in the mixhtre at the start. The capacity of the bottle=34i c.c; subtract 20 c.c; and volume of mixture tested=32i c.c. The thiosulphate used=6-95 c.c; but each c.c. = 0-0003773 gramme oxygen; .-. 6-95 c.c. = 0-0003773 X 6-95 grammes of oxygen; and the dissolved oxygen in parts per , . , 0-0003773x6-95X100,000 ^.Q^„ 100,000 of mixture= ^^^-^^ ^^ = 0"bi7. 321 * In the case of poor effluents and tank liquors a bro^^^l precipitate may form, and this must be given time to disappear. 154 LABORATORY WORK The dissolved oxygen in the mixture at the end of incubation similarly calculated =0-488 part per 100,000. Thus 0-817- 0-488 = 0-329 gramme of dissolved oxygen is taken up by 100,000 parts of the mixture; and as 1 of this mixture is effluent, 0-329 x 5= 1-65 is the absorption of dissolved oxygen by 100,000 parts of effluent in five days at 18° C. Notes. — It is recommended that all samples should be allowed to stand at 18° C. for forty-eight hours after sampling and before carrying out the test ; and if it is not possible to commence the tests at the end of that time the sample should be meanwhile kept on ice. For practical purposes no corrections are necessary for the few c.c. of added reagents. The temperature of 18° C. (65° F.) represents the maximum temperature likely to be reached in river water. The tempera- ture of the incubator should not var}^ more than one degree on either side of this standard temperature. A few control determinations with tap water alone should be done from time to time, to make sure that the tap water itself does not take up any appreciable quantity of oxygen. Effluents conforming to a satisfactory standard may cause considerable growths of organic life, which may subsequently produce a nuisance. There is much to learn of the causation of these objectionable growths, but certain of them are doubtless promoted by the presence of nitrates. Carchesium (constituting whitish masses of fflamentous growth, characterized by bell- shaped heads on thread-like stems), green growths of Oscillatoria nigra and Spirogyra, and coarser growths of water-weeds may grow in well-purified effluents ; and imperfectly purified effluents may foster such grey growths as Leptoniitus and Sphcerotilus, or even Beggiatoa. A large thin-leaf seaweed, of cabbage-green colour and known as sea-lettuce {Ulva latissima), like different species of the grass- like Enteromorpha, etc., flourishes in association with the sewage pollution of sea water. The ulva grows most extensively in those estuaries where the water is shallow and the tidal movements are slow and ineffectual in carrying all the sewage pollution out to sea at each ebb of the tide. We are indebted to Professor E. H. Letts for much information with reference to this sewage seaweed. He finds that it absorbs nitrogen from the ammonia and nitrates of sewage origin, and that mussels attach themselves to it by SEWAGE-POLLUTED MUD 155 their byssus threads. After the ulva reaches a certain size, wave action detaches most of it, and the small retained pieces are capable of continuing the growth of the plant. If the detached ulva is not swept out to sea, it accumulates on the shore, where, exposed to the sun and air, fermentative decomposition sets in, and a micro-organism reduces the sulphates (which are abundantly present in the tissues of the ulva) to sulphides; and eventually sulphuretted hydrogen is liberated and an intolerable nuisance results. The weed is also found in some places where the sea water is free from sewage pollution, in sheltered and shallow waters with sluggish currents, and where the means for its anchorage exist. In sewage-polluted mud the N could be determined in lo grammes of the mud by the Kjeldahl method. Information may also be obtained as regards the deoxygenating qualities of the mud deposited in the bed of a stream. Twenty-five grammes of the mud should be mixed with 500 c.c. of tap water and 10 c.c. of the mixture (containing 0-5 gramme of the mud) made up to 100 c.c. with distilled water, when Tidy's permanganate process may be performed. The dissolved oxygen absorption at 18° C. in twenty-four hours may be determined on i or 5 grammes of wet mud (accord- ing as the mud is foul or otherwise) by allowing the mud to remain in an airtight bottle in contact with a relatively large volume of water containing oxygen in solution; and Winkler's process, as modified by Rideal and Stewart, may be employed (pp. 152 and 153). A highly nitrogenous mud will usually take up much oxygen, and fine sulphide of iron undergoes oxidation very readily. The absorption of dissolved oxygen by polluted muds goes on for very long periods, but the twenty-four hours' test serves for comparative purposes; indeed, as there are no recog- nized standards, the presence and amount of contamination by animal matter are best proved by comparing the results obtained with those furnished by the mud collected at other parts of the river bed, which are obviously remote from possible contamination. PART III SOIL EXAMINATION THE ANALYSIS OF SOILS The sanitarian will not often find it necessary to make a chemical analysis of soil; he may want to classify soils and to examine them for f cecal pollution; but generally for his purposes laboratory results are of very secondary importance to those of observations made upon the soil in situ. Although the power of absorbing and retaining moisture is a consideration of the first importance from a health view- exercising as it does an important influence upon the health of whole communities — yet it is of litcle practical value to perform any tests in this connection upon small quantities of soil which are collected and brought to a laboratory. The amount of moisture retained is so largely dependent upon local factors that the most reliable and valuable information is always ob- tained by observations of the soil in situ. The amount of moisture in a sample of soil would be ascertained by drying 50 grammes on the water-bath, and then placing in the hot-air oven at 95° C. until the weight is constant. The depth of the ground water and the extent of its fluctua- tions are often of great importance. The digging of trial-holes will enable the height of the ground water to be ascertained, and the fluctuations in the level of the ground water may be determined by some arrangement similar to that shown in Fig. 19, which sufficiently explains itself. The method of testing the capacity which the soil possesses for holding water is obvious: The dried soil is weighed in a cylinder and then saturated with water (this may take hours in the case of clay) ; the water is allowed to drain off through very 157 158 LABORATORY WORK fine muslin until no more drops fall, and the soil is then reweighed; the difference in the two weighings represents the weight of water the known weight of soil is capable of holding. In collecting samples it must be borne in mind that the char- acters of the soil may vary within small areas and at different depths, so that many samples may have to be collected and analyzed before one can speak with accuracy of the composition of the soil of a small area. These samples may be taken by a long, narrow spade, or by means of an ordinary i:^- to 2 inch auger screwed into the soil ; and 2 or 3 kilogrammes must be collected FIG. 19. ARRANGEMENT FOR REGISTERING THE VARYING LEVELS OF THE GROUND WATER. A, a float; B, a pulley; C, an index; D, a graduated scale. Opposite the scale D a narrow glass window is provided so that the scale can be seen without disturbing the arrangement. for analysis. The depth of the surface soil varies considerably in different localities. In uncultivated grounds it generally occupies only a few inches in depth on the surface, and in culti- vated grounds its depth is generally the same as that to which the implements used in cultivation have penetrated; which is gener- ally from 9 to 12 inches. Soil is composed of laj'ers of more or less disintegrated rock and quantities of organic matter resulting from the decay of plants and animals. Entering into its composition are: The earths — silica, alumina, lime and magnesia; the alkalies — soda. THE ANALYSIS OF SOILS 159 potassa and ammonia; the acids — sulphuric, hydrochloric, car- bonic, nitric, phosphoric, silicic and humic; oxide of iron and small portions of other metallic oxides; a considerable propor- tion of moisture (chiefly as a liquid film enveloping the particles) ; several gases, and micro-organisms. In addition to the so-called " nitrifying organisms," there are some which reduce nitrates to ammonia, and others that will fully oxidize ammonia in the pres- ence of air, but will reduce nitrates to ammonia in the temporary exclusion of air. The variable amount of vegetable and animal matter gaining access is either partially or wholly decomposed, and is ultimately converted into water, carbonic acid and nitric acid by the action of micro-organisms. Of the mineral matters, either silica, silicates and double silicates, or calcium and mag- nesium carbonates, generally predominate. All soils contain, though in different proportions, the chief mineral constituents which are found in the ash of the plants which grow upon them ; and an examination of such ash will often afford a rough-and- ready clue to the constitution of the soil. The colour of soil depends mainly upon the amount of humus, oxide of iron and moisture. A dried soil is always much lighter in colour than when the moisture is present. The less weathered stratum, which lies immediately under the soil, is called the subsoil, into the composition of which com- paratively little organic matter enters. Sometimes this subsoil is porous sand or gravel; sometimes light and loamy and closely similar to the superimposed soil; sometimes stiff (clayey) and more or less impervious to water. The subsoil is generally lighter in colour than the soil, and its depth is usually limited by deposits of undecomposed or partly decomposed rock matter, or by deposits of clay, sand, or gravel. The Classification of Soils. The sample having been collected, the coarser stones should be removed, and all lumps broken up so far as possible with a wooden pestle. The mechanical analysis of soils, or the sorting of the con- stituent particles into groups, is effected either bj^ a stream of running water or by allowing the turbid mixture of soil and water to settle during varying periods of time, after the coarser particles have been removed and sorted by sieves of different- sized meshes. l6o LABORATORY WORK Schloesing has insisted upon the neeessity of iiist treating the soil with dilute acid and subsequently washing it, and of adding ammonia to the water in which it is afterwards to be suspended. The acid dissolves calcium carbonate and decomposes " humates," and the liberated humic acid is dissolved by the ammonia. Other- wise the humates (if abundant) and the chalk tend to bind together the finest particles, which flocculate into loose aggregates which ma^' not get disintegrated. The acid emploj^ed is too weak to dissolve any appreciable amount of mineral constituents other than calcium carbonate. The groups of particles obtained in a mechanical analj^sis do not possess any definite chemical individuality. The coarser fractions may contain fine grains of quartz, particles of clay, ferric oxide, etc. It is likely that the phj^sical properties of the soil depend rather on the size than on the chemical composition of the constituent particles. Special apparatus has been devised, both for thoroughly crushing and also for washing and separating the various soil constituents seriatim. Knopp's set of sieves is useful for the purpose of classifying the coarse constituents of some soils. The soil is first dried, and then the lumps of soil are crushed up with the fingers and placed upon the top sieve with the coarsest meshes ; no hard pestle must be used for the crushing, or miineral particles would be disintegrated or broken. After thorough shaking, the particles all separate out on one of six sieves, and the very fine material collects on a tray at the bottom of the apparatus. This latter material may be classified by means of elutriation, or washing. Particles collecting on the top sieve are more than 7 milli- metres in diameter, and = coarse gravel ; those collected on the second, between 7 and 4 millimetres, Rnd= medium gravel; on the third between 4 and 2 millimetres, and = fine gravel ; on the fourth, between 2 and i millimetres, Sind= coarse sand ; on the fifth, between i and 0-3 millimetre, 3.nd= medium sand; on the bottom, finer than 0-3 millimetre, smd—fine sand. What remains upon each sieve is weighed, and the results are expressed as percentages of the total weight. Kiihn classifies everything coarser than 5 millimetres as stones ; between 5 and 3 millimetres, as coarse gravel ; between 3 and 2 millimetres, as fine gravel ; between 2 and i millimetres, as THE ANALYSIS OF SOILS i6i pearl sand ; finer than 0-5 millimetre, as fine sand ; and the portion separable by elutriation, as eaYth. Elutriation may be performed by the washing cylinder of Knopp. This consists of a glass cyhnder 55 centimetres high, to which are attached four glass tubes fitted with taps at intervals of 10 centimetres. The soil material which passes through a 0-3 millimetre sieve is placed in the cyhnder, and this is filled with water to 10 centimetres above the highest tap. The whole is well shaken for five minutes, and then allowed to stand for another five minutes, when the top tap is opened and the cloudy water allowed to escape into a weighed dish. The material drawn off from the second and third and fourth taps is similarly ilill 4 '^ iHii^ FIG. 20. KNOPP'S SOIL-WASHING CYLINDER. collected. The separate cloudy waters are then evaporated and the residues weighed; and the fine sandy residue at the bottom of the cyhnder is also collected and weighed. In this manner the clayey matter in the fine soil can be further classified and compared with similar observations on other soils. To constitute pure clay the particles should not exceed o-oi milfimetre, and the material should be previously treated with sufficient hydrochloric acid to dissolve out any carbonates; the washed son should then be boiled for half an hour with 10 per cent, ammonia to dissolve humus, and the residue, washed, dried, and ignited, may be weighed as clay. This method is suffi ciently exact for practical purposes. Poquillon advocates the following method of estimating clay: l62 LABORATORY WORK Ten grammes of the soil are rubbed up with 25 c.c. of water, and the hquid mixed with 100 to 120 c.c. of a o-i per cent, solution of ammonium chloride and left for five minutes. The supernatant liquid is then decanted; the operation is repeat-ed six to eight times until the washings are clear, when the residual sand is washed, first with dilute hydrochloric acid and then with water, dried and weighed. The turbid wasliings are mixed, •acidified with hj'drochloric acid, left for two or three hours, when the precipitated clay is collected on a filter, washed with water, dried, and weighed. The amount of sand in clay is usually estimated as follows : Heat a weighed quantit}' of the dried fine material with sul- phuric acid; then boil with water, collect the insoluble matter on a tared filter, dr}^ and weigh; remove and boil an aliquot part of this insoluble matter with a strong solution of sodium carbonate, and weigh the insoluble residue as sand. Lime may be estimated by treating the earth in a litre shaking- flask for half an hour with 500 c.c. of standardized hydrochloric acid (containing about -J per cent, oi hydrogen chloride), and titrating an aliquot part with soda, using phenolphthalein as indicator. The simplest apparatus for a silt estimation takes advantage of the relative rates of descent of the various soil particles through water at rest. The chief disadvantage of silt methods is the tendency of the fine particles to aggregate and form small lumps which act as larger particles. The only means of obtaining a perfect separation of the soil particles are by boiling for one hour with water, and by wet pestling. Even then it has been demonstrated that particles of the different sizes are represented throughout the entire deposit. Doubtless the finest particles are best separated by elutriation. An}' soil containing less than 5 per cent, of chalk, which is not so rich in vegetable matter as to constitute a " peaty " one, and which contains not over 10 per cent, of clay and excess of sand= a " sandy soil." If such soil contains 10 to 40 per cent. of clay and excess of sand = a " sandy loam." If 40 to 70 per cent, of clay= a " loamy soil." If 70 to 85 per cent, of clay= a " clay loam." If 85 to 90 per cent, of clay= a " strong clay soil." Sand makes a soil friable, gives it a low specific heat and the power of draining quickly. A clay soil containing no sand at all= a "pure agricultural THE ANALYSIS OF SOILS 163 day'' which is essentially silicate of alumina mixed with small quantities of organic matter, lime, magnesia and ferric oxide. The different varieties of clay are mainly due to the varying amounts of these latter substances. Strong clays absorb and retain nearly three times as much water as sandy soils, while peaty ones absorb a still larger pro- portion; and the same remarks broadly apply to the relative readiness with which water is lost by evaporation from these soils. If there is more than 5 per cent, of chalk, the remainder con- sisting mainly of clay, the soil is called a " marl " ; and if there is more than 20 per cent, of chalk, " calcareous." " Peaty " soils generally contain from 60 to 80 per cent, by weight of organic matter; " rich cultivated soils," from about 5 to 20 per cent. ; and " stiff clayey " ones, from 2 to 10 per cent. By means of vegetation, and owing to the fixation of free nitrogen by soil micro-organisms and plants, even a sandy soil may in time become productive. To ascertain the substances which a water will extract from soil and hold in solution, Schulze's method is recommended by Fresenius. The necks of several middle-sized funnels are closed with small filters of strong filter-paper; these are moistened, and the paper pressed close to the sides of the funnel; the air- dried soil is then introduced in small lumps ranging in size from a pea to a walnut (but not pulverized, or even crushed) until the funnels are filled to about two-thirds. Distilled water is now poured on in quantity sufficient to cover the soil. If the first portion of the filtrate is turbid, it must be poured back into the funnel, and the filtration allowed to proceed quietly ; the funnels are again filled with water, and this process of extraction is continued until the combined filtrates weigh twice or three times as nmch as the soil used. The several filtrates are next mixed, and the necessary analysis performed to obtain the desired information. Alumina was never found by Schulze in the aqueous extract. In most soils the phosphoric acid exists as a basic ferric phos- phate, and hence the great insolubility of soil phosphates. The smaller part into which the non-concentrated aqueous solution was divided is finally tested for nitric and nitrous acids and ammonia. 164 LABORATORY WORK As, however, the solvents which act naturally on the soil are something more than distilled water, it is desirable to examine those substances which are soluble in carbonic acid water, as by saturating distilled water with carbonic acid and allowing this to act upon the soil for several da^-s in a closed flask, which should be well shaken from time to time. Water containing both carbonic acid and ammonium chloride (about 0-05 per cent.) should also be allowed in a similar manner to act upon the soil and the substances then taken up should be examined. Probably the best solvent for extracting from soil the " avail- able " (as distinguished from " total ") mineral constituents for lant food, is a i per cent, solution of citric acid (Bernard Dyer). The total phosphoric acid in soils should be determined in the manner recommended by Hehner, as follows: The soil is incinerated and digested with hj^drochloric acid, evaporated to dryness to render silica insoluble, redigested with acid, filtered and washed. The filtrate and washings are concentrated to a small bulk, and treated in the cold with excess of a solution of ammonium molybdate in nitric acid. After standing forty hours in a warm place, the liquor is decanted through a filter, the precipitate is washed several times by decantation (first with dilute nitric acid, then with very small amounts of distilled water), and finally transferred to the filter and washed free from excess of acid. The ammonium phospho-molybdate is then dissolved in ammonia, evaporated to dryness in a platinum capsule, and dried to constant weight at 100° C. The residue contains 3I per cent, of its weight of phosphoric acid. Or the ammonium phospho-molybdate may be dissolved in ammonia, magnesium mixture added, and the precipitate collected, ignited and weighed as MgaPgO^, which xo-64=P2^5- To estimate nitric acid, first rapidly dry the sample at about 60° C, so as to stop nitrification ensuing after the collection of the sample ; extract 1,000 grammes of fine soil with 2,000 grammes of distilled water for forty-eight hours, with frequent shaking; and then filter 1,000 c.c. of the extract (corresponding to 500 grammes of the dry soil). A small quantity of pure sodium carbonate should be added to the filtrate, which is next evapor- ated to about 100 c.c. Any precipitate which forms during evaporation should be filtered off, when the nitric acid may be estimated in the filtrate. Sulphur exists in soil as sulphates (generally calcium sulphate). THE ANALYSIS OF SOILS 165 in organic compounds, and as sulpliides (iron pyrites). To estimate the sulphates in soil, heat along with dilute hydro- chloric acid for a short time, then filter, and precipitate from the filtrate with barium chloride solution. An examination for the peaty acids may be made thus: Some of the washed soil is dried and sifted, to separate any straw, roots and stones; what passes through a fine sieve is digested for several hours at about 30° C, with a solution of carbonate of soda, and filtered. The filtrate is then slightly acidified with hydrochloric acid; and if brown flakes separate, these consist of the peaty acids — i.e., ulmi'c, humic, or geic. The more ulmic acid is present the lighter is the shade of brown; a dark shade indicates a preponderance of humic or geic acids. These flakes may be collected upon a weighed filter, washed until the water begins to be coloured, dried and weighed. Then burn the dry mass, deduct the weight of the ash (after sub- tracting the filter ash) from that of the dry mass, and enter the difference as " acids of humus." The total nitrogen of soil would be best determined by Kjel- dahl's process: Five to ten grammes of the fine air-dried soil are placed into a small hard Jena glass flask, and 30 c.c. of pure sulphuric acid are poured over the soil, so as to thoroughly wet it. When all the frothing has subsided, 15 grammes of potassic sulphate are added (to raise the boiling-point), and about ^ gramme of colour- less (anhydrous) cupric sulphate (as an oxidizer), and the mix- ture is heated until the liquid is a yellow colour. Then 50 per cent, caustic potash solution (recently boiled to expel any ammonia) are added until the liquid becomes alkaline, as in- dicated by the circumstance that the copper is precipitated as blue cupric hydroxide. The remainder of the process is carried out in the manner described on p. 146. Example. — 10-55 grammes of soil were taken. The distillate received into 50 c.c. ^^ sulphuric acid required 27 c.c. ~ sodic hydrate to neutralize it. .-. 23 c.c. of the /o acid have been neutralized by the ammonia in the distillate. But I c.c. of the y'^ acid= 0-0017 gramme NH3 or 0-0014 gramme of N. .-. there are 23 x 0-0014= 0-0322 gramme N in 10-55 grammes of soil= 0-3 per cent. l66 LABORATORY WORK The collection of ground air and the estimation of carbonic acid are dealt with in Air Analysis. It is occasionally desirable to know whether the soil has been recently polluted with excremental matter. The filtered aqueous extract (obtained b}^ acting upon a known weight of dried soil with distilled water for forty-eight hours, with frequent stirring) can in these cases be examined for oxidized nitrogen, chlorine, and organic matter, and the amounts thus obtained compared with those procured from similar soil in the neighbour- hood. In 100 parts of soil dried in the air Krocker found that clayey soils, before manuring, yielded o-i to 0-45 of ammonia; loamy soils, 0-13; sandy soils (ne\er cultivated), about 0-05; and marls, 0-004 to o-og of ammonia. Ferrous sulphide is always in evidence in foul sewage deposits and in mud exposed to gross sewage contamination. Its presence has been explained by fermenting organic matter reducing ferric oxide or hydrate to ferrous compounds; then some of the sul- phuretted hydrogen from the decomposition of organic matter forms the ferrous sulphide; and carbonic acid, acting on ferrous sulphide, is capable of producing the sulphuretted hj^drogen which may cause an offensive nuisance (vide pp. 122 and 155). As would be inferred, the soil of graveyards above the burial level does not materially differ, as regards the amount of organic matter and its products, from similar soil (unmanured) elsewhere; but that taken on the level of the cofhns and from a short dis- tance below, is relatively richer in organic matter. Such soil is found to be somewhat richer in bacteria than other unmanured soils, and more especially is this the case with that lying around the top of the coffins (Reiners, Fraenkel, Young). The various manures with which the soils under cultivation are dressed necessarity effect considerable changes in the con- stitution of the original soil, besides yielding abundance of soluble matter to the water which comes in contact with them. The commoner manures are — Farm^'ard and animal excrement and "guano"; bone dust and other phosphatic manures (calcium phosphate), etc. ; vege- table manures — sawdust, soot, charcoal, peat, and seaweed; ammonia salts, especially the sulphate; sodium salts, especially the nitrate; potassium salts, especially the chloride, nitrate, and phosphate; and gypsum. THE ANALYSIS OF SOILS 167 The following are some of the recommendations of a Com- mittee of the Agricultural Education Association : Taking Samples. — Under ordinary conditions the sample shall be taken to a depth of 9 inches, but in case of shallow soils to such lesser depths as mark a natural line of demarcation. The committee approves of the use of the auger as one method that may be adopted for taking samples. Several cores should be taken and mixed for analysis. Drying. — The sample shall be air-dried for analysis. The drying may be accelerated by heating to a temperature not ex- ceeding 40° C, but in all cases the soil should be finally left, for a day or two, spread in a thin layer and exposed to the air at the ordinary temperature of the room. Sifting. — A sieve with round holes 3 millimetres in diameter shall be used to separate the fine earth for analysis from the stones and gravel. Gentle pressure with a wooden or caoutchouc pestle, or other means, may be adopted to break up aggregates of clay and silt, but care should be taken not to crush any of the stones or lumps of chalk. For determination of the " available constituents " the " fine earth " is used without grinding. For the other determinations 100 grammes or more of " fine earth " is sifted through a woven sieve of 40 meshes to the inch, or a sieve with round holes of I millimetre in diameter. What is retained by the sieve is ground till it will pass through, and the whole mixed. Determination of Carbonate of Lime. — The carbon dioxide evolved on treatment of the fine earth with acid is calculated as carbonate of lime. This is regarded as a convenient measure of the " available basicity " of the soil, without discriminating between carbonates of lime and magnesia. Determination of Total Mineral Constituents. — The fine earth is boiled with strong hydrochloric acid in an open flask for a short time in order that the acid may attain constant strength, and then digested at the ordinary water-bath or steam-oven temperature for forty to forty-eight hours, the flask being loosely stoppered. In this solution the phosphoric acid and potash are determined, and other mineral constituents as desired. The object is to obtain as thorough an extraction of the soil as is possible, short of ultimate analysis. The period for the extraction is made sufficiently long to minimize errors due to l68 LABORATORY WORK variations in the actual time, the strength of the acid, or the temperature. Unignited soil is taken, since ignition effects a drastic and variable alteration of the constitution of the soil — e.g., no con- stant proportion is found between the potash extracted from ignited and unignited soil. H3'drochloric acid is taken as the most generally effective solvent; even peaty soils are found to yield as much phosphoric acid to hydrochloric acid as to nitric acid or aqua regia. The Bacteriological Examination of Soil. From the point of view of an examination to obtain results immediately available for public health purposes, the bacterio- logical examination of soil has only a limited usefulness. Its utility from this aspect is chiefly in connection with the con- tamination of water from surface washings. The examination of soil for B. typhosus, B. pestis, and other pathogenic organisms is, apart from the pathogenic anaerobes, a matter of great difficulty. The examination is made on the general lines laid down for the isolation of these organisms. Thus, to detect the typhoid bacillus in soil, the soil is mixed thoroughly with sterile water, and the water examined for this organism by methods similar to those used for its isolation from contaminated water. A number of investigations have been carried out upon the vitality of typhoid bacilli in soil. The results show considerable discrepancies. They indicate that under experimental con- ditions the typhoid bacillus will survive for many weeks (ten to twelve) in soil, but that under natural conditions probably not more than one week; and that the factors influencing its vitality are many and varied, the antagonism of other microbes, and the physical conditions of moisture and temperature being the most important. Soils which have been comparatively recently contaminated with organic matter in quantity — for example, by sewage or manure — show evidence of this when bacteriologically examined in the total number of aerobic organisms, the number of spores present, the number of B. coli, B. enteritidis sporogenes, and streptococci. The following statement is by W. G. Savage: THE ANALYSIS OF SOILS 169 " In collecting soil for bacteriological examination the depth from which it is obtained is of fundamental importance. If the surface soil is to be examined, scrape up with a sterile spatula, and transfer to a sterile receptacle. To obtain soil from a given depth either a fresh cutting must be made and the soil collected at the required depth, or, preferably, some form of borer may be used. For this purpose Fraenkel's borer is convenient, its chief drawback being that it holds only a small quantity of soil. " If Fraenkel's borer is used, it is advisable to collect at least eight samples from spots about a foot apart, and to mix together to obtain a representative sample. Also in this way sufficient soil will be obtained for a concurrent chemical examination. By means of this borer the exact depth of the soil taken can be ascertained. Owing to its length it cannot be sterilized in the hot-air oven, but it can be conveniently and sufficiently sterilized by pouring in methylated spirit and igniting. After steriliza- tion wrap the lower portion in a sterile cloth and secure with string. This plan is very convenient when a number of samples have to be taken in one day, and at, perhaps, a long distance from the laboratory, since the borer can be resterilized at once before each sample is taken, it being only necessary to carry a bottle of spirit and a number of sterile cloths in a metal box. The soil is removed by a sterile spatula from the interior of the borer to the steriHzed tin or other receptacle used for the soil. " The examination should.be commenced as soon after collec- tion as possible. " To estimate the total number of bacteria, and for some other steps of the examination, very extensive dilution must be prac- tised. As an example of a convenient method of dilution the following procedure is given: other methods of dilution will readily suggest themselves. It is important to remember that owing to a number of inherent difiiculties (such as the difference of coherence of different soils) numerical estimations are only relatively accurate, and in any case the same method should be used throughout for each investigation: " Accurately weigh a small sterile glass-stoppered bottle con- taining 100 c.c. sterile water. Quickly introduce, with a sterile spatula, I gramme of the soil (previously well mixed together) into the bottle. With a little practice i gramme can be quickl}- and sufficiently accurately added. Mix very thoroughly by re- peated shaking, if necessary breaking up the soil by a pointed 170 LABORATORY WORK sterile glass rod. Call this solution ' Dilution A.' Allow the soil particles to settle, then add i c.c. or more, according to the sus- pected contamination of the soil, to a sterile flask contain- ing 100 c.c. (or, more accurately, 99 c.c.) sterile water. Mix thoroughly and label ' Dilution B.' Van»dng quantities of Dilutions A and B are used for the examination. ' ' To obtain the total number of aerobic organisms make gelatine plates from these dilutions. Thus, 0-2, 0-5, i-o c.c. Dilution B are convenient amounts to add to the gelatine tubes. For the number of organisms developing at 37° C. use in the same way agar plates. "For the number of spores, present as such, add varjdng amounts FIG. 21. FRAENKEL'S borer. Lower end shown with open and closed soil-chamber. of the dilutions to gelatine tubes. Heat to 80° C. for ten minutes, then plate, incubate, and count in the ordinary way. "For B. coli various fractions of the dilutions are added to tubes of lactose bile salt media. These are incubated at 37° C, and those which produce acid and gas are used to inoculate solid media, and the organism isolated exactly in the same way as for the isolation of this bacillus from water. " Streptococci and spores of B. enteritidis sporogenes are exam- ined for by methods identical with those used for water." According to Houston and Savage, B. coli is absent, or present in small numbers only, in uncontaminated soils, and is not readily isolated even from soils polluted with objectionable THE ANALYSIS OF SOILS 171 animal matter, unless the contamination is gross in amount and of recent sort. Houston found that when sewage containing large numbers of B. coli is added to soil tlie coli organisms relatively rapidly disappear, Houston regards the presence of the spores of B. enteriiidis sporogenes as an indication of contamination, but not necessarily recent, and the presence of streptococci as indicating very recent contamination. Experimenting with soil to which sewage (con- taining numerous streptococci) had been added, he found that the addition of sewage to a soil might be detected by the presence of streptococci even in a minimum amount of the soil thus polhited, but that their disappearance seems to be extremely rapid. The results, obtained by the writer, of a mineral analysis of a few common soils are given below. It must be understood that soils which are called by the same name may vary con- siderably in the nature and amounts of their less characteristic constituents. The main purpose of the following analyses is to afford an approximate idea of the amounts of the more charac- teristic substances which enter into the composition of a few of the more common soils: Clay [Stourbridge) Silica . . Alumina Organic matter Iron (oxide) . . Lime . . (carbonate, 1-4) (sulphate, o-i) Magnesia, etc. Phosphoric acid Water . . 68 15 4 3 1-5 traces 0-5 lOO-O Calcareous [Sussex). Lime . . . 90 (carbonate, 89-5) (sulphate, 0-35) (phosphate, 0-15) Organic matter 3 Oxide of iron and alumina . . . . 2-5 Silica 0*55 Magnesia (carbonate) • 0-5 Water • 3-45 100-00 172 LABORATORY WORK Peaty {Devonshire). Organic matter Silica . . Alumina Lime . . Oxide of iron . . Sulphuric acid Magnesia Soda and potash Phosphoric acid (Air-Dried Soil.) 90-5 7-5 074 0-5 0-46 0-2 0-05 0-03 0-02 100-00 Garden (Vegetable) Mould. Silica . . • 49-25 Organic matter • 13-5 Oxide of iron . . • 9-25 Carbonic acid. . 7-12 Lime . . • 5-13 Alumina • 274 Soda and potash • 2-5 Chlorine • 1-5 Sulphuric acid • 1-3 Phosphoric acid . 0-4 Oxide of manganese . 0-25 Magnesia o-i6 Water . 6-9 100-00 PART IV AIR ANALYSIS CHAPTER I THE NORMAL CONSTITUENTS OF AIR- EUDIOMETRY -oxygen- Composition OF THE Atmosphere (freed from aqueous vapour) In loo Volumes. 78-07 20-95 0-95 0-03 V traces. Nitrogen Oxygen Argon . . Carbonic acid Hydrogen Ammonia Nitric acid Helium, krypton, neon, xenon The amount of aqueous vapour is variable, the average in this country being 1-4 per cent. Suspended matter is also present in variable nature and amount. In the air of towns the carbonic acid may vary from 0-03 per cent, to o-o8 per cent, and over during the prevalence of a fog. Oxidized sulphur and sulphuretted hydrogen are frequently present in traces, as are also ammonia, marsh gas, and organic matter. During dense fog in large town districts the amount of sulphurous and sulphuric acid present is very much increased. The air of large towns is generally slightly acid, owing to the sulphurous and sulphuric acids which are derived from the sulphur compounds contained in the articles used for com- bustion ; and the air of occupied rooms in which coal gas is burning 173 174 LABORATORY WORK may be slightly acid. A piece of delicate blue litmus-paper, moistened \vith neutral distilled water, commonly denotes this acidit}' by changing in an hour or less to a faint, though dis- tinct, red. Oxygen. We haye seen that the amount of oxygen in the external atmosphere may be taken to constitute a normal percentage of about 20*95. After a careful consideration of the number of inyestigations which have been made, it seems that it may reach a slightly higher limit oyer large expanses of open countr3% and that, eyen in the atmosphere of occupied rooms, it rarely falls below 2075 per cent. In some mines the oxygen has been estimated considerably below 20 per cent. (Angus Smith found i8-2y per cent., and some Continental observers have estimated it even lower.) The Estimation of the Amount of Oxygen in THE Atmosphere. Eudiometry. Apparatus required. — An eudiometer {evSloti, good, and fiirpov, measure) is the instrument emplo^^ed for measuring the volume of a gas or gaseous mixture. One of the most simple and convenient forms is Hempel's gas burette. From the accom- panying figure it will be seen to consist of two glass tubes sup- ported on fiat stands and connected together at their lowest points by wide india-rubber tubing; the tube which is seen in Fig. 22 to be held up (and which will be subsequently referred to as tube A) is plain, and is continued full bore to the top, where it generally ends in a slightly trumpet-shaped mouth ; the other tube (which will be referred to as tube B) is graduated into c.c, and narrowed above so as to fit inside of a short piece of small india-rubber tubing, which serves to connect it with an " absorption pipette " containing the absorbing solution. This apparatus, as shown mounted upon a wooden stand, consists of two glass bulbs blown in a fine glass tube, bent in the manner portrayed in the figure; the lower globe has a larger diameter than the upper, and is capable of holding about 150 c.c. of the reagent employed, while the upper one should be of at least 100 c.c. capacity. EUDIOMETRY 175 To charge the absorption pipette tlie hquid reagent is poured into the upper bulb, and the air is then sucked out of the lower bulb through the capillary tube rising from it, until the lower bulb is filled and the reagent reaches into the syphon bend of the capillary tube, but leaves the upper bulb almost empty. The single " absorption pipette " is shown in Fig. 22; but for the absorption of oxygen it is necessary to use " a double pipette." Reagents that are affected by oxygen (such as pyrogallic acid and cuprous chloride) should not be employed in a single pipette FIG. 22. HEMPEL's GAS BURETTE AND ABSORPTION APPARATUS. because the reagent in the lower bulb is exposed to the general atmosphere. Hempel's double pipette permits of the use of such reagents without this exposure. The double pipette is shown in Fig. 23. The first bulb is the largest (150 c.c), and is filled with the absorbent; the next is empty; the third contains water; and the fourth is empty. Thus, when any gases are passed over into the " pipette," the water in the third bulb passes into the fourth to make room for the gas, and thus shuts out the atmosphere, the absorbing reagent only coming in contact with the small amount of air originally in the second bulb. 176 LABORATORY WORK Reagents Employed. — A solution of pyrogallic acid and caustic potash — i.e., 15 grammes of the acid and 50 of caustic potash, to the htre of distilled water. The Process. 1. A certain amount of atmospheric air is first measured in the gas burette in the following manner: Both tubes are placed upon a level surface, and distilled water, which has been thoroughly shaken up in the air and thus mechanically saturated with it, is poured down the plain tube A until each tube is about half-full. Now if the tube A is raised the height of the water in B will ascend until it fills the tube B; and when the tube A is subsequently lowered the atmospheric air of the compartment will pass into the graduated tube B, where it can be imprisoned by turning the greased cock at its mouth. The volume of air thus collected is next exposed to the same atmospheric pressure as obtains in the room by adjusting both tubes so that the water stands at the same level in each. An accurate reading is then taken of the volume of air collected. It is convenient to take about 100 c.c. 2. Connection is then made, as shown in the figure, by fine, stiff india-rubber tubing, between the burette and the " absorp- tion pipette " ; the latter being raised close to the top of the tube containing the air, since it is desirable to have as short a length of tubing as possible. 3. A background of white enamel serves to make the coloured absorbing liquid which rises in the capillary tube distinctly visible, and enables the precise height to which it reaches to be carefully marked. 4. Next, by liberating the clasp upon the tubing and opening the greased cock above referred to, the gas burette and absorp- tion pipette are put into communication with each other; when by raising the tube A the air is forced over into the absorption pipette. The cock on the burette is then closed, the india-rubber tubing is pinched, and a firm clasp applied, after which the pipette may be disconnected and gently shaken. 5. When absorption has taken place, by reconnecting the absorption apparatus with the burette, opening the cock and removing the clasp, the residual air can be brought back into the burette by lowering the tube A; care being taken that the EUDIOMETRY I77 absorbing solution does not pass beyond where it originally stood in the connecting capillary tube, as indicated by the mark on the porcelain. 6. The air may be thus treated several times, in order to give the solution time to absorb all the oxygen. 7. Before making the final reading the height of the water in the two tubes is brought to the same level, just as it was at the commencement of the process (and for the same reason), and then the volume which the air now occupies is read off. A constant reading should be obtained. 8. The volume remaining is broadly due to nitrogen, and the difference between it and the original volume represents the oxygen FIG. 23. — HEMPEL'S double ABSORPTION PIPETTE. and CO2 absorbed. The solution of pyrogallic and caustic potash will also absorb sulphuretted hydrogen, sulphurous acid, and hydrochloric acid (if present). From the percentage loss in the volume of the air by this treatment the percentage amount of CO2 (estimated by Pettenkofer's process, and calculated to the same temperature and pressure) must be deducted, and the remainder will represent the percentage amount of oxygen at the current temperature and pressure. Example. — The volume of air collected in the burette at the current temperature and pressure was 50*6 c.c. The volume of the residual air after treatment was 40*0 c.c. Therefore the loss is I0'6 c.c. in 50'6 c.c. = approximately 20-95 per cent. Assuming that the COg in the air of the room has been found to be 0"05 per 12 T78 LABORATORY WORK cent., the ox3'gen would amount to approximately (20*95 - 0'05) = 20-90 per cent., at the current temperature and pressure. Notes. — At temperatures of about 15° C. the last trace of oxygen is thus removed in about three minutes of shakmg (Hempel). Since the conditions of temperature must remain the same throughout the estimation, the gas burette, after it has been charged, should not be handled except by its iron or wooden stand, and the apparatus must not be moved about from one spot to another. The absorbing reagent should always first be saturated at the current temperature and pressure by shaking it up with gases that are but slightly soluble in it, otherwise errors of estimation result; the necessity of always thus saturating the water in the burette with the gas under examination has been mentioned. With this precaution this method of eudiometry gives results but little inferior to those obtained by working over mercury and using solid absorbents. Mr. J. F. Spencer has designed a four- way tap and connections, fitted to the top of the burette, which permits of the absorbing liquid being brought right up to this tap, so that on turning the tap the gas in the burette can be brought directly in contact with the absorbing hquid, without any intervening air space. The oxygen may also be estimated by Dumas' process, in which the air, having been freed of its carbonic acid by aspirating through a strong solution of caustic potash, is passed through a combustion tube containing a length of pure spongy metallic cop- per. The copper is kept ignited, and becomes tarnished by oxidation; the difference in weight of the original copper and the tarnished metal represents the oxygen taken up from the volume of air experimented upon. CHAPTER II CARBONIC ACID The estimation of the carbonic acid in the atmosphere is of great value. This is not because the carbonic acid is hable to exist in injurious amounts even under the worst conditions of ventilation commonly obtaining, but because this gas, when furnished by respiration, may afford an important clue to the general condition of the atmosphere. It is therefore the knowledge of the amount of carbonic acid which has been added to the atmosphere by respiration which is generally required. So inert is carbonic acid per se that it may exist to the extent of 2 to 3 parts per lOO without serious consequences, and fatal results would not accrue with less than from 5 to 10 per cent. The amount of carbonic acid which is present in a pure atmo- sphere, and which may be termed " normal," is 0-033 per cent, by volume. The lowest estimation of carbonic acid made in any atmosphere was 0-02 per cent., in air at a very high altitude. The external atmosphere of cities, during fogs, often contains '0-07 per cent., and may contain as much as o-og per cent, or even more. In ill-ventilated sitting-rooms, well lighted by gas, the carbonic acid often reaches 0-2 per cent. Where there is " overcrowding " it has been estimated as high as 07 per cent., and it is commonly under these circum- stances 0-3. Angus Smith found in the worst parts of theatres 0-32 per cent. We are indebted to Pettenkofer for a method of estimation which, owing to the facility of its performance and the accuracy of its results, is very generally adopted. 179 l80 LABORATORY WORK The Collection of Samples for the Estimation of Carbonic Acid by Pettenkofer's Method. For the estimation of carbonic acid, the samples of air may be conveniently collected in wide-mouthed, glass-stoppered bottles of about 4 litres capacity, which, when used for this purpose, are termed " air-jars." These must be thorouglily cleansed in every case before use, and the last washings should be with ammonia- free distilled water. After the collection of the sample the stoppers should be tied dowTi, and hermetically sealed with pre- pared lard in those cases where they have to be removed. Lastly, a label is attached, on which should be written a statement of the current temperature and pressure, the date, the hour, the place from which the sample was taken, and the nature and extent of the non-human sources of COg (gas-burners, candles, and lamps) . Following out the principles ad\'ocated with regard to water samples, a sample of air should be collected — whether it be vitiated by respiration, combustion, trade processes, or by pro- ducts of decomposition, etc.- — at the time when, so far as can be judged, the atmosphere will afford its maximum evidence of pollution. In investigating the respiratory contamination of the air of a bedroom, for instance, the sample should be taken shortly before the first riser quits the room — that is to say, after the room has been occupied by its customary number of occupants for the greatest number of consecutive hours. The immediate proximity of gas, lamp and candle lights, stoves and fireplaces should be avoided; and samples should always be taken at the mean height at which the air is being respired. For purposes of calculating the added COg a comparison sample of the external atmosphere should always be taken, since the COg in the external atmosphere in towns varies materially from time to time. The air is made to occupy the jar by either of the following methods : I. An air-jar may be accurately filled with clean water — which can, with rare exceptions, be got upon the premises, and should have been previously boiled — and then emptied and allowed to drain in the compartment the air of which is to be examined. A sample of the air then rushes in to fill the place of the escaping CARBONIC ACID l8l water. At the time of use the water should be at the tempera- ture of the room. 2. The air may be forced in by bellows, which are provided with a long nozzle, which reaches well down into the jar to within an inch of the bottom. This ensures that the air which originally occupied the jar will be completely displaced from below up- wards. 3. The original air in the jar may be pumped out by means of a small air-pump. Angus Smith drew the air out of the bottle by a flexible bellows-pump, shown in Fig 24. For filling small bottles, J. S. Haldane suggests a long piece of rubber tubing reaching from the bottom of the air-jar to the operator, who sucks in a deep breath of air through the tube. FIG. 24. -THE FLEXIBLE BELLOWS-PUMP EMPLOYED BY ANGUS SMITH TO DRAW OUT AIR FROM THE AIR-JAR. 4. A jar may be accurately filled with mercury, and emptied in the compartment where the sample is to be collected. Al- though this plan is theoretically the best, it is practically in- applicable on account of the large amount of mercury required, and the difficulty of conveying this (from its great weight) from place to place. The first method is recommended; for it is easy of execution, and furnishes satisfactory results. Whenever it is possible, there is a slight advantage in making the analysis at once in the compartment in which the sample has been taken, since the general atmosphere is that of the jar; and this can sometimes be done, although it is often very l82 LABORATORY WORK incon\'enient. There should be as Httle loss of time as possible in commencing the analysis, and in the meantime the jar should not be exposed to temperatures varying much from that at which the sample was collected. The room in which the sample is analyzed must be free from draughts and of a uniform temperature, or, at least, not liable to frequent changes in temperature. The error which would be introduced by breathing into the air-jar and reagents, or by handhng the jars more than is abso- lutely necessary with warm hands, is obvious. It is desirable to hold the breath for the few moments during which the sample of air is being collected in the air-jar, the wide mouth of which permits of a rapid escape of its contained water. Pettenkofer's Alkalimetric Method of Estimating THE Carbonic Acid in the Atmosphere. The rationale of the process is as follows: Clear baryta water combines with carbonic acid with great readiness, thereby becoming turbid (BA(0H)2 + CO2 = BaCOg + H2O), and the carbonic acid taken up reduces the alkalinity of the bar\'ta water. If, therefore, the degree of alkalinity of a measured quantity of bar^'ta water is estimated, and then this reagent is made to take up all the carbonic acid of a sample of air, the amount of carbonic acid taken up will be in proportion to the reduction of alkalinity of the baryta water. Special Apparatus : I. An air-jar of about 4 litres (4,000 c.c.) capacity. It is necessary to know the exact capacity of the bottle in order that the amount of air which it will hold may be accurately known. This can be ascertained by filling the bottle with as much water as it will hold when the stopper is inserted, and then measuring the water as it is emptied out; the volume of the water which the bottle held will correspond to the volume of air which takes its place. Chemical Reagents : I. Pure clear baryta water (4'5 grammes of the crystallized hydrate to the litre) to which about i gramme of baric chloride should be added to counteract the influence of small quantities of alkalies which may be present. In order that the barj-ta water may be kept quite pure it will be necessary to remove the carbonic acid from the air which enters the store bottle when some of its contents are withdrawn, by making it pass through soda CARBONIC ACID 183 lime. Fig. 25 shows how this can be readily effected : A large glass store bottle is represented, fitted with a syphon tube to draw off the clear baryta water. Any air which enters must pass through the tube which is packed with soda lime. 2. A standard solution of oxalic acid (crystallized) made to such a strength (i.e., 2*819 grammes to the litre) that i c.c. corresponds to 0-5 c.c. of carbonic acid at the standard temperature and pressure. 3. A solution of phenolphthalein (i part in 250 parts of alcohol). The Process. 1. A sample of the air is collected in the air- jar. 2. Fifty c.c. of perfectly clear baryta water are then placed in the jar, and the liquid is made to flow round the sides of the jar FIG. 25. STORE BOTTLE FOR BARYTA WATER. by rolhng it about on its side for a minute or two from time to time. About three-quarters of an hour should be allowed for all the carbonic acid present in the sample to combine with the baryta. 3. The alkalinity of another 50 c.c. of the clear baryta water is meanwhile tested in the following manner: The 50 c.c. are placed in a small flask with a very narrow neck, and are tinted pink by a drop of phenolphthalein ; then the standard solution of oxahc acid is cautiously run in from a graduated burette until the alkalinity of the baryta water has just been neutralized by the acid, when the amount of acid employed is read off from the 184 LABORATORY WORK burette. The exact neutral stage is indicated by the disappear- ance of the pink colour. 4. At the end of tliree-quarters of an hour 25 c.c. of the baryta water are removed from the air-jar, and the causticity estimated as above. In effecting the removal the precipi- tate of barium carbonate which has settled should not be disturbed. 5. The difference in the number of cubic centimetres of acid solution required to neutralize {a) 50 c.c. of the original baryta water, and {b) the baryta water which has taken up the carbonic acid of the air in the jar, represents the amount of carbonic acid in the air. 6. When 50 c.c. of baryta water were added to the jar an equivalent bulk of air was displaced. The amount of air experi- mented upon, therefore (assuming the capacity of the jar to be 4,000 c.c), represents 4,000 — 50 c.c. = 3,950 c.c. 7. The result must be returned as the amount of carbonic acid per cent, of the air at the " standard temperature " — i.e., 0° C. — and the " standard pressure " — i.e., 760 millimetres of mercury— since the \'alue of the oxalic acid solution is in respect to CO2 at the standard temperature and pressure. Example. — The causticity of 50 c.c. of the original and clear baryta was tested by cautiousl}^ running in the standard acid solution. Twenty c.c. of the acid solution were required to effect the neutralization. Twenty-five c.c. of the baryta water from the air- jar required only 8-5 c.c. of the acid solution to neutralize its diminished causticit}^ Therefore, the whole 50 c.c. required 17. .•. 20-17 = 3 c.c. of the acid solution represents the carbonic acid taken up by the barj^ta water from the sample of air. But I c.c. = 0*5 c.c. of carbonic acid. .•. 3 c.c. = 1*5 c.c. of carbonic acid. .'. the carbonic acid originally in the air sample amounts to 1-5 c.c. at the standard temperature and pressure. The capacity of the jar was 4,000 c.c, and the air examined is 4,000 c.c. -50 c.c. = 3,950 c.c. at the current temperature and pressure — sa}', 742 millimetres and 17° C. Now, this volume of air has to be converted to its \-olume at the standard temperature and pressure. B}' Bo\'le's law the volume of air varies inversely as the pressure. By Charles' law the air contracts on cooling CARBONIC ACID 1 85 ^1-^ ( = 0-00366) of its hulk at 0° C. for every degree Centigrade down to 0°. The volume of air experimented with wiU therefore represent, at the standard temperature and pressure : 3 950x 742 ^ ^930900 ^ 3,630 c.c. 760 X {i + (o'00366 X 17)} 76oxi"o6222 .-. there were 1-5 c.c. COg in 3,630 c.c. of air, or 0-0413 per cent., at the standard temperature and pressure. The sample of the outside air is similarly examined, and the difference between the COg in the inside and outside air will represent the added COg impurity of the inside air. Notes. — One c.c. of COg at 0° C. weighs 1-96633 milhgrammes at 760 millimetres pressure; the relation, therefore, between the volume and the weight = Y.-y6-V¥¥ = o '508. The writer prefers to perform this process in a long cylindrical jar of the capacity of about 3 litres and fitted with a doubly per- forated cork, one perforation transmitting the drawn-out point of a graduated burette with stopcock, and the other perforation a small piece of glass tubing carrying externally a small, close- fitting, india-rubber cap. The air is collected as directed; 50 c.c. of baryta water (standardized at the time of use by decinormal oxalic acid) is added through the glass tube, which is then sealed by its cap. After one hour, the bottle being gently rolled along a table from time to time in the meanwhile, a few drops of phenolphthalein are added through the small glass tube (the cap of which is now removed), the biirette is charged with the decinormal oxalic acid solution, and this is allowed to discharge cautiously into the bottle. So soon as the pink colour commences to weaken, the acid is added slowly in drops, and the bottle is gently shaken and stood upon a white porcelain slab. The neutral stage can be readily noted when the experimenter looks down from above on to the white porcelain slab. This method avoids exposing the baryta to the atmosphere and the breath of the worker. The above method, in which the air is collected in a large bottle, is preferable to the method of slow aspiration of air through barium hydroxide contained in tubes, for it is more con- venient in practice, and the sample represents the state of the air at a particular time ; whereas by the latter method the air is being continuously collected for an hour or more. iS6 LABORATORY WORK Letts and Blake have pointed out that the action of the BaHaOo on the glass leads to a slight contamination with alkalies and silica, and therefore some error of experiment. They recom- mend that the air-jar and the barjla stock bottle should therefore be coated with paraffin wax; but the \\Titer finds that this pre- caution may be disregarded in the case of bottles which have been previously well exposed to BaHgOg. The Process of Lunge and Zeckendorf is quicker and simpler but not so precise and reliable as Pettenkofer's, and it is least satisfactory where the amount of CO2 is very low; but it is a useful means of readily making a closely approximate estimation. In this method a decinormal solution of sodium carbonate is prepared (5-3 grammes of anhydrous carbonate to the litre), to FIG. 26. — THE APPARATUS FOR THE LUNGE AND ZECKENDORF PROCESS. which a little phenolphthalein is added, a purple solution resulting which keeps well. A small caoutchouc bag of 70 c.c. capacity, and fitted with two valves working opposite to each other (so that on pressing it the air is forced in one direction onlv), is connected by means of an india-rubber tube with a glass tube, and this passes through a perforation in an india-rubber stopper fitted tightly into a small flask of 150 c.c. capacity. The ball is first pressed several times, so as to fill it and the bottle with the air to be tested; 2 c.c. of the decinormal soda solution are then added to 100 c.c. of freshl}^ distilled ammonia-free water, and 10 c.c. of this -g^ normal solution are emptied into the small flask. The ail of the bag is then slowly pressed over into the reagent, the flexible tube is tightly compressed with the fingers, and the flask well shaken. This is repeated with another bagful of air, and so on until the purple colour is discharged. C.\RB0NIC ACID 187 Lunge found, by direct experiment, the percentage amount of carbonic acid which corresponds to the number of times the air-content of the ball must be discharged into the bottle. His table is as follows: Number of Pressures of India- rubber Ball. 7 0-135 8 0-115 9 O'lO 10 O'og II 0-087 12 0-083 13 o-o8 14 0-077 15 0-074 Parts per Cent, of Carbonic Acid in the Air. 0-30 0-25 0-21 0-18 0-155 Number of Pressures of India- rubber Ball. 16 17 18 19 20 22 24 26 28 30 35 40 Parts per Cent, of Carbonic Acid in the Air. 0-071 0-069 0-066 0-064 0-062 0-058 0-054 0-051 0-049 0-048 0-042 0-038 0-030 As suggested by S. H. Davies, a pump delivering 50 c.c. per stroke is better than the caoutchouc bag employed by the authors of the process. Haldane's Rapid Method of Determining Carbonic Acid in Air. In this process the apparatus shown in Fig. 27 is required. The gas burette A, which is enclosed in a water-chamber, consists of a wide ungraduated and a very narrow graduated portion. It holds about 20 c.c. from the tap to the bottom of the scale. The graduated part, which is about 4 inches long, is divided into 100 divisions, each of which corresponds to xowo^h part of the capacity of the burette. The lowest division is marked 0, and the highest 100. Any difference between a reading at or near zero and a second reading is thus shown by the scale in volumes per 10,000. In using the apparatus the air is first expelled from the gas burette by opening the three-way tap B to the outside and raising the mercury bulb C. The tap is then closed, and the mercury bulb replaced in its stand. On opening the three-way tap again a sample of the air is drawn in, and the level of the 1 88 LABORATORY WORK mercury falls to near the zero mark. The tap is now opened towards the absorption pipette D, which is filled up to a mark at E with a 20 per cent, potash solution, and the sample measured vvitli the precautions to be described below. It is then passed over (by raising C) into the absorption pipette, the potash being displaced into bulb I. The air is dri\'en backwards and forwards for a minute, and then again measured after the absorption of the carbonic acid. The difference between the two readings FIG. 27. HALDANE S APPARATUS. gives directly the number of volumes of carbonic acid per 10,000 in the sample of air. It is evident that the correctness of the analysis depends entirely on the a\oidance of errors of various kinds in the two determinations of the volume of the enclosed air. Mistakes might be caused by slight variations in the temperature of the water, or the pressure under which the sample is measured, or in the degree of saturation with moisture of the sample. A variation of o-i° C. in the temperature of the water in the jacket CARBONIC ACID 189 would, for instance, unless corrected, cause an error of fully 4 volumes per 10,000 in the analysis. In order to have a sharp index of the pressure under which the air is measured, the level, not of the mercury, but of the potash solution in the narrow bore tubing of the absorption pipette, is taken as the index of pressure. For the first measure- ment, the level of the potash solution is accurately adjusted to the mark at E by raising or lowering the mercury by means of the rack-and-pinion arrangement shown in Fig. 27. For the second reading the potash level is again adjusted in the same way. To correct for variations in temperature of the water-jacket a control tube G is employed, of a size and shape approximately the same as the gas burette. The control tube communicates with the potash through the narrow-bore glass tube H, and before the first measurement is made the level of the potash in H is adjusted to the mark by lowering or raising the reservoir I, which slides up and down in a loosely fitting cork. At the second measurement the same precaution is taken, so that the air in the control tube occupies exactly the same volume as at the first measurement. As an alteration of temperature or of barometric pressure would affect the pressure to an equal extent in the gas burette and control tube, it is evident that the adjust- ment of the level of the potash reservoir compensates exactly any error which the alteration of temperature or of barometric pressure would cause in the reading of the gas burette. Before the adjustments of the potash levels are made, the water in the jacket is thoroughly mixed by blowing air through it by means of the tube K. This manipulation is essential to an accurate result. In order to obviate error due to variations in the saturation of the air, both the burette and the control tube are left with a little visible moisture inside. If the burette has once been wetted inside, and as much as possible of the water expelled by raising the mercury, it remains moist for a very large number of analyses, but a little moisture should always be visible. If by any mishap potash should be sucked over into the burette, it and its connection must be washed out with dilute acid intro- duced by the tap. At the end of an analysis the taps must be turned so as to close the communication between the potash and the burette and igO LABORATORY WORK control tube: otherwise potash may be sucked in if there is any great fall of temperature or rise of barometric pressure. 1 he manipulations required during an analysis may be re- capitulated as follows: (i) Open the tap of the control tube to the air for a moment, and then turn it so as to connect the con- trol tube and potash-pressure gauge. (2) Turn the tap of the burette so as to connect the burette and the potash pipette. (3) See that the level of the potash alters sharply and about equally in the tubes when the potash reservoir is raised. (4) Blow air through the water-jacket. (5) Raise or lower the potash reservoir till the potash is exactly at the mark in tube H. (6) Raise or lower the mercury reservoir by means of the rack and pinion till the potash in E is exactly at the mark. (7) Read off the mercury level on the scale of the burette to 0"2 of a division. (8) Raise the mercury to the upper hook, so as to drive the air into the potash bulb, then lower it a little and raise it again twice so as to \\ash any carbonic acid in the connecting tubing into the potash bulb. (9) Return the air to the burette. (10) Blow air through the water-jacket. (11) Adjust the two potash levels as before, and read off the mercury level. The first reading subtracted from the second gives the result in volumes per 10,000. (12) Close the two taps. The advantages of the method are that an estimation can be made in two or three minutes, only a very small volume of air is required, and the apparatus, being fitted up in a small box, is exceedingly portable; the disadvantages are that it is rela- tively costly, it only makes an approximate estimation of the amount of COg, and (to a beginner) it is rather difficult in manipu- lation. The process is, however, sufficiently accurate to suffice for most of the practical purposes of hygiene. Conclusions to he Drawn from the Amount Estimated. — From the writer's experiments, no " stuffy " odour is appreciable in the atmosphere until the respiratory impurity reaches at least 0-03 per cent, in those cases where samples have been collected from rooms occupied under ordinary conditions. This stuffiness is mainly due to exhalations from the skin, and the degree of personal cleanliness largely determines the rapidity of its appearance; and recent experiments have demonstrated that the physical changes in impure air are at least mainly respon- sible for the ill-effects of the air of overcrowded rooms. 'Ihese experiments included a number of tests made in a CARBONIC ACID IQI specially constructed glass chamber in which the physical and chemical qualities of the air could be rigorously controlled. It was found that with a respiratory impurity of carbonic acid exceeding any recorded up to that time as having been found in the air of a crowded room — e.g., from i-o to 1-5, or even 17 per cent. — no injurious property of the air could be demonstrated so long as the temperature and humidity were kept low; and that under these circumstances the absence of any disturbance was so complete that even the power of cerebration remained intact. On the other hand, as soon as the temperature and humidity were increased to beyond a certain point, there appeared, both in normal and in diseased persons who were submitted to experi- ment, the usual symptoms that occur when people are crowded together in one room — i.e., feelings of discomfort, oppression, lassitude, giddiness, nausea, etc. These symptoms, however, could be relieved at once by reducing the temperature and humidity of the air to normal. The extent to which gas burned in a common gas-burner may furnish carbonic acid to the atmosphere is between 2 and 3 cubic feet per hour (for i cubic foot of gas produces from 0-5 to o-6 cubic foot of CO2) ; and the amount of sulphur compounds thus yielded in a gas well purified is not important hygienically. The air over burial-grounds, especially when these are crowded, has been said to contain an abnormally high amount of carbonic acid, but the writer's experiments have failed to confirm this. CHAPTER III THE ORGANIC MATTER IN THE AIR The organic matter in the air includes that given off from the lungs and skin; its composition is very imperfecta understood, but it consists partly of volatile fatty acids and their ethers, and partly of vaporous and suspended matters (epithelial and fatty debris). It is certainly largely oxidizable and nitrogenous, since it will reduce solutions of the permanganate of potassium, and will yield ammonia. It quickly putrefies; and when air containing it is aspirated through sulphuric acid, the organic particles are charred and darken the solution. When collected in large amounts in water it can be precipitated by silver nitrate. Probably the major part is molecular and suspended, since it does not diffuse equally about a room, and tends to fall and settle; and there is no doubt that it is mostly in combination with watery vapour, for substances absorb it according to their hygroscopic powers — i.e., it is absorbed chiefly by wool, feathers, etc., and least by horsehair. It gives a foetid, " stuffy " odour to the atmosphere, and from the persistence of this odour it is doubtless burnt off but slowly by the atmospheric oxygen ; and in smaU quantities it gives odour to water. The processes which we may employ for the estimation of this matter in air are preferably those which ser\^e to detect the same matter in water. A large measured volume of the air is made to slowly pass through distilled ammonia-free water, which will retain all the soluble and suspended organic material. The water is then tested by Wanklyn's method as to its nitrogenous organic matter, and by Tidy's process, as to its oxidizable organic matter — it being borne in mind in the latter test that either nitrous acid, sulphurous acid, sulphuretted hydrogen, or tarry matters, will, if present, also reduce permanganate of potassium. A convenient method of performance (Fig. 28) is to take a small 192 THE ORGANIC MATTER IN THE AIR T93 wash-bottle, partially fill with 250 c.c. of distilled ammonia-free water, and then tightly fit with a doubly perforated india-rubber stopper. Into one perforation a glass tube bent at right angles, with one trumpet-shaped extremity, is accurately fitted, while the other end is made to dip well down into the distilled water ; the second perforation conducts another bent glass tube, the end contained within the flask being above the surface of the water, and the other connected directly by india-rubber tubing to a second wash-bottle similarly fitted, and containing another 250 c.c. of the distilled water. This second bottle is connected by india-rubber tubing to the aspirator. The capacity of the FIG. 28. — -APPARATUS FOR COLLECTING THE ORGANIC MATTER IN AIR. aspirator being known (and a convenient size is that of 20 litres), it is filled with tap water ; the tap is then turned so that the water passes slowly out, when air enters the trumpet-shaped mouth of the bent glass tube to take the place of the escaping water; such air is washed in the distilled water in the two bottles before it reaches the aspirator, and so parts with its organic matter. Example.— li it is desired to make an estimation of the nitro- genous organic matter, the aspirator is five times filled and allowed to empty; 100 litres of air will then have been drawn through the 500 c.c. of distilled water; therefore this 500 c.c. of water will contain the nitrogenous organic matter of 100 litres of air. 13 194 LABORATORY WORK Suppose, after distilling off the free ammonia, the 500 c.c. of water are found by Wanklyn's method to contain 002 milli- gramme of albuminoid ammonia, then there will be 002 milli- gramme of such ammonia in 100 litres of air. But in dealing with air such results are generally expressed in terms of " milli- grammes per cubic metre." Therefore, if there is 0'02 milligramme of albuminoid ammonia in 100 litres, there will be 02 milligramme of albuminoid ammonia in i ,000 litres — or a cubic metre — of air. Outside air contains albuminoid ammonia up to ot milli- gramme per cubic metre, and averages about 0"o8. In a hospital ward this ammonia has been estimated as high as 1-3. The oxidizable organic matter may also be collected in the same way, and estimated by Tidy's process, as in Water Analysis. The estimation of the oxidizable organic matter can also be performed by very slowly aspirating 100 litres of air through two wash-bottles containing standard potassium permanganate solution and dilute sulphuric acid kept at about 27° C. The strength of the permanganate may be milHnormal, as in the process next to be described. In line, bright weather the oxi- dizable organic matter in town air will absorb about 3 c.c. of oxygen per cubic metre or 3 volumes of oxj-gen in 1,000,000 volumes of air, but in stagnant and fogg>' air the amount is considerably higher. Another method of approximately estimating the organic matter in air is that of Camell}^ and Mackie. In this process from 3 to 4 litres of air are shaken with 50 c.c. of millinormal solution of potassium permanganate for five minutes, and the amount of decomposed permanganate is deduced, on colorimetric principles, from the loss in colour sustained by the original solution; and from the extent of this loss, as estimated b\' the amount of standard solution required to restore the colour, the amount ot oxygen absorbed can be calculated. One c.c. of the millinormal solution =0008 miUigramme 0=0-0056 c.c, at standard tem- perature and pressure. To each litre of the 17^777 solution are added 50 c.c. of dilute sulphuric acid (i in 6). Henriot and Bouissy have suggested that the moisture of a measured volume of the vitiated air of crowded apartments should be collected by condensation, and the reducing substances estimated in the moisture obtained. The amount of organic matter in air is found to be closely related to the amount of dust. CHAPTER IV AMMONIA— MARSH GAS— CARBON MONOXIDE— SULPHUR COMPOUNDS— NITRIC, NITROUS AND HYDROCHLORIC ACIDS— PHOSPHURETTED AND ARSENIURETTED HYDROGEN Ammonia. Traces of ammonia are present in every atmosphere. In towns it generally amounts to about o*o6 milligramme per cubic metre. Such traces are derived almost entirely from the combustion of coal and coal gas. The ammonia generally exists in combina- tion with an acid, as the carbonate and chloride, or less commonly as the nitrate or sulphate. Ammonia may be found in consider- able quantity near ground where decomposing organic matter is deposited. Traces of ammonia do not appear to affect health, but it is a constant ingredient of the most impure airs. A considerable amount of ammonia in the atmosphere may be detected by moistening strips of filtering paper with Nessler's reagent, and hanging these up for some time in the air of the com- partment. But when present only in faint traces, large quan- tities of air must be aspirated through distilled ammonia-free water rendered slightly acid with sulphuric acid, and the ammonia tested for by Nessler's reagent; and if a measured quantity of air is employed, the ammonia may be estimated quantitatively by " Nesslerization." It has been found as high as 0-8 milligramme per cubic metre in certain hospital wards. Marsh Gas (CH4). This gas probably exists in air in many circumstances, but owing to the difficulties of its detection, it is not easy to prove its presence when in traces only. There are certainly traces in the atmosphere of towns, and over districts of abundant vegeta- tion (especially when such districts are marshy) it may exist in con- siderable quantities. As evolved from strata in which coal mining 195 • iq6 LABORATORY WORK operations are progressing it is known as " fire-damp," and its power of exploding when ignited in the presence of carbonic acid has been often disastrously exemplified. There is little doubt that after a while marsh gas may create symptoms of poisoning, and, being inodorous and non-irritating, its presence would not be detected by the senses. Any escape of coal gas, containing as it does some 35 per cent, of marsh gas, will charge the atmosphere with considerable and dangerous amounts, but fortunately in these cases the strongly-smelling ingredients of the coal gas give timely warning. Carbon Monoxide (CO). The affinity of this gas for haemoglobin is about three hundred times greater than that of ox^-gen. Owing to its properties of entering into combination with the haemoglobin of the red cor- puscles, displacing their oxj^gen and thus paralyzing their oxj'gen-carrying functions, it destro^'S life by cutting off the oxygen supply to the brain and tissues; and its dangers are enhanced from the circumstance that it gives no indication of its presence to the sense of odour. Symptoms of poisoning are e\-ident when the haemoglobin is about one-third saturated with CO, and death results from some 70 or 80 per cent, of saturation. Sir T. Oliver thus describes the symptoms: In acute intoxica- tion the individual feels dizz\', and complains of headache, noises in the ears, throbbing in the temples, a feeling of sleepiness, and a sense of fatigue. There may be a feeling of sickness which culminates in vomiting, a sense of oppression at the chest, with quickened or irregular breathing, palpitation, and an inability to stand or walk straight. Convulsions ma}' or may not come on, or there may be only a few muscular tremors. There is a peculiar fixed look about the eyes, the pupils of which are dilated and their reaction slow. Consciousness is lost by degrees, or it may be retained for some time, and yet, owing to the great loss of motor power, the individual, although aware of the danger, is often unable to escape from it. When a man has recovered from the effects of carbon monoxide, his Ufe is still imperilled for some days to come. Not only does he run the risk of dying as late as eight days after the accident, but he has still to face the risk of secondary maladies developing — such, for example, as gl}-cosuria. It may be necessary to examine the air for this gas in the CARBON MONOXIDli I97 atmosphere of compartments where iron or copper stoves are employed, and especially when the material is cast-iron and when the fuel is coke; where coal gas (which contains some 5-5 to 6-5 per cent., but may contain from 4 to 12 per cent.) is incom- pletely burnt or escapes; or where there is a possibility of some of the products of combustion from leaky furnace-flues, etc., escaping into a compartment — for the air in furnace-flues has been found to contain over 20 per cent, of carbonic oxide, and that of ordinary flues from domestic fireplaces as much as 4 per cent. The carbonic oxide of the air of flues is always the product of imperfect combustion — that is to say, the carbon is either not fully oxidized to carbonic acid (COg) owing to the supply fresh air being insufficient, or else the carbonic acid, being formed low down in the furnace, gets reduced to carbonic oxide in subsequently passing through the rest of the furnace. Of the gases generated from the explosion of gunpowder, carbonic oxide forms about 7-5 per cent., but in mines it is usually only formed in dangerous amount by an extensive fire-damp explosion, and especially by an explosion in which coal-dust is involved. A serious drawback to the adoption of " water gas " as a source of heat and light is the fact that it contains (before combustion) from 25 to 35 per cent, of this very dangerous ingredient. Considerable quantities of CO may exist in the atmosphere near coke-ovens, brick-kilns, and cement works. The carbonic oxide in the atmosphere of stove-heated rooms is derived from either or several of the following sources : 1. Red-hot cast-iron may transmit the gas from the fire, either through its substance or through minute fissures, and the hot iron may even reduce CO2 to CO. , 2. The carbon which enters into the formation of the cast-iron may get oxidized, and reach the external atmosphere, as CO. 3. Particles of suspended organic matter in the atmosphere may get charred and partially oxidized by coming in contact with a highly heated stove. 4. Currents may pass down the smoke flue under certain con- ditions, and thus introduce the gas. In the Lancet of February 7, 1914, a useful means of testing as to whether the products of combustion of a gas stove are or are not entering a room is described. When traces of chloroform or carbon tetrachloride are introduced into the mixture of air 198 LABORATORY WORK and gas drawn into a Bunsen burner, the products of combustion then contain hydrochloric acid, which, although present in small proportion, is readily detected by the dense white fumes of am- monium chloride which it forms in coming into contact with ammonia. All that is necessary, therefore, in testing a gas fire as to its ventilating function is to place near the air inlet of the gas hre an absorbent substance (a chalk pencil) containing carbon tetrachloride, and then to hold on a wire a piece of sponge con- taining strong ammonia solution at different points along the rim of the canopy of the gas lire. Any leakage or escape of combustion products is thus instantly indicated by the produc- tion of visible fumes of ammonium cliloride. The advantage of this test is that any point of leakage can be located from a gas fire fixed in position in the house. CO is present in traces in tobacco smoke. Qualitative Tests. Vogel's test is suffiiciently delicate for all practical purposes. Into a wash-bottle 100 c.c. of distilled water are poured, and then a little defibrinated blood is added to the bottle, which is afterwards connected to an aspirator. At least 10 litres of air are then drawn through the faintly reddish liquid. The test is obscured if too much blood is present. The bottle is then rolled about for half an hour and allowed to stand for a short while, C I - F 1 ii 1 I'i^ liii 1" ^^ 1 'if r 1 1! Ill 1 1 FIG. 29. SHOWING THE CHARACTERISTIC DISPOSITION OF THE ABSORPTION BANDS IN THE SPECTROSCOPIC PICTURE OF OXY- AND REDUCED KJEMO- GLOBIN. The upper scale represents oxy-hasmoglobin and the lower reduced haemoglobin. when some of the contents are removed and examined by the spectroscope. Oxy-haemoglobin shows two well-marked bands with sharp edges, in the yellow and in the green parts, respectively, of the solar spectrum, both lying between Fraunhofer's lines D and E. The spectroscopic appearance of hemoglobin in combination with carbonic oxide is almost identical, but the left-hand band CARBON monoxidp: ^99 (at blue end) of the carbonic-oxide-hEemoglobin lies a lit^e nearer to the right (yellow end) than in the case of oxy-ha.moglobm and the edges of the band are not so sharply defined Ihe blood also takes up a more or less marked bright pmk or cherry- red tint, which may also be observed in the cadaver. Two drops of a colourless solution of ammonium sulphide are next added, the bottle is well shaken, and the liquid is gently warmed and re-examined. If no marked change m the spectro- scopic appearance of the fluid has ensued, ^^^^^onic oxide hemo- globin is present ; otherwise the ammomum sulphide will deoxi- dize or reduce the oxy-ha^moglobin, and the two bands will be replaced by a single broad band shaded off at the borders^ and occupying a position almost intermediate with regard to the oridnal two bands. ,, . Delicate results are obtained where the CO is not oo small n amount, by placing a mouse in a wire cage and aUowing it to breathe the air for several hours. The mouse may be subse- quently drowned in its cage, and the blood then examined by the spectroscope, to see if the two absorption bands of CO- hemoglobin are present. A control test may be made fro- the blood of a mouse which has not been thus exposed to carbon "^^T^Xfhas devised a delicate chemical test for CO-hemoglobin : To 10 cc of the solution of blood he adds 15 c.c. of a 20 per cent, solution of potassium ferrocyanide and 2 c.c. of acetic acid (I volume of glacial acetic acid to 2 volumes of water) ; the pre- cipitate very soon becomes reddish-brown if CO-hemoglobm be present, but greyish-brown with oxyhemoglobin, the difference slowly disappearing. Chemical Tests upon the Air. I If air is aspirated through a tube filled with a solution of paUadium chloride (containing i milhgramme of palladium and 2 drops of hydrochloric acid), a portion of the paUadmm is reduced to the metallic state and a dark precipitate results. After an hour's action with frequent shaking, the palladium which has deposited owing to the reducing action of the carbon monoxide may be collected and ignited with the usual precau- tions- I gramme of metallic palladium =0-2624 gramnie or -.10 cc of CO The air should first be aspirated through lead 200 LABORATORY WORK acetate solution, and also through dilute sulphuric acid, to remove any SHg or NH3 which may be present. 2. A known quantity of the suspected air, freed from COa by its passage through potash bulbs, is passed over periodic acid contained in a U-tube kept at a temperature of 80° C. The carbon monoxide, if present, decomposes the periodic acid, setting free iodine. From the amount of liberated iodine the quantity of carbon monoxide is deduced. The method is accurate, and no other element likely to be present in the atmosphere will reduce periodic acid; but the air should first be freed from dust. Quantitative Estimation. The subchloride of copper (made by exposing copper turnings and the oxide of copper to the action of strong hydrochloric acid — S.G. 1-124) has the property of absorbing CO, and advantage may sometimes be taken of this fact to estimate the quantity present by the method of eudiometry, where the amount is considerable and exceeds o-i per cent. It is necessary that the Og and CO, of the air be first removed by means of potassium pyrogallol mixed with a considerable excess of potassic hydrate, before the residual air is slowly passed over into a double absorption pipette charged with the solution of subchloride of copper. It is also necessary to use an absorption apparatus with large bulbs, in order that a good quantity of the copper solution may be employed; and time must be allowed, for complete absorption takes place slowly. Two or three treatments should be repeated until a " constant reading " is obtained, to ensure that all the gas has been absorbed. If there is a marked amount of carbonic oxide present, the loss in the original volume, taken under the same conditions of temperature and pressure, is appreciable, and represents the amount of CO; or the cuprous chloride solution may be transferred, under suitable precautions, and boiled in vacuo, and the expelled gas collected. Quite 98 per cent, of the carbonic oxide actually present will be obtained by the latter method. The union of cuprous chloride with carbonic oxide is very feeble, and the solution readily parts with the carbonic oxide to the atmosphere on shaking. The solution will also absorb acetylene and ethylene. CAKBON MONOXIDE 201 /. Haldane's method of testing for the i)resenf:e and for esti- mating the amount of carbon monoxide is as follows: For the detection of CO he places in a dry and clean bottle about 5 ex. of a dilute blood solution, and then, after aspirating some of the suspected air through the bottle, stoppers it and shakes for ten minutes. During the shaking the bottle should be protected from light, which has a most powerful dissociating action on CO-hamoglobin. On pouring out the solution into a test-tube and comparing its tint with that of some of the original solution in another test-tube, the presence of the carbonic oxide is indicated by the pink tint of the former. To measure accurately the extent to which blood is saturated with CO, Haldane's method depends upon the fact that normal blood, when sufficiently diluted with water, has a yellow colour, whereas blood saturated with carbonic oxide forms a pink solu- tion when similarly diluted. A solution of about i of normal blood to 100 of water is made; also a solution of carmine, dis- solved with the help of a little ammonia and diluted till its depth of tint is about the same as that of the blood solution. Two test- tubes of equal diameter (about | inch) are then selected. Five c.c. of the solution of normal blood are measured into one of the test- tubes, and a drop of the suspected blood is placed in the other test-tube and cautiously diluted with water till its depth of tint is about equal to that of the normal solution. If carbonic oxide be present in the hemoglobin, a difference of quality in the tints of the two solutions will now be clearly perceptible. Carmine solution is then added from the burette to the normal blood, and water (if necessary) to the abnormal blood, till the tints are equal in both quality and depth. The carmine is added in about 0-2 c.c. at a time, the points being noted at which there is just too little and just too much carmine, and the mean taken. The solution of abnormal blood is then saturated with coal gas by thoroughly shaking up with coal gas for a few seconds, and the addition of carmine to the other test-tube continued until equality is again established, the amount of carmine used being noted. The percentage saturation of the abnormal blood with CO can now be easily calculated, since we know how much carmine solution its saturation represented as compared with what complete saturation represented. The method of calculation is illustrated by the following example: To 5 c.c. of normal blood solution 2-2 c.c. of carmine 202 LABORATORY WORK is required to be added to produce the tint of the blood under examination, and 6-2 c.c. to produce the tint of the same blood fully saturated. In the former case the carmine was in the proportion of 2-2 in 7-2, and in the latter of 6-2 in 11 -2. The percentage saturation (.%') of the hemoglobin with carbonic oxide is therefore given by the following proportion sum: 6-2 2-2 : — : : 100 : x ; II-2 7.2 X therefore =55 -2. As the compound of CO with haemoglobin is to a slight extent dissociated when the blood is diluted with water, the value found is a little too low. The corrections needed are as follows : Add 0-5 if 30 per cent, saturation be found, i-i if 50 per cent., i-6 if 60 per cent., 2-6 if 70 per cent., 4-4 if 80 per cent., 10 -o if 90 per cent. Thus, in the above example, we must add 1-3, so that the true saturation is 56-5 per cent. In comparing the tints the test-tubes should be held up against the light from a window, but bright light should be avoided as much as possible, as it increases the dissociation. For the detection and determination of small percentages of CO in air the sample of air is collected in a clean and dry bottle of about 4 ounces capacity. The cork of the bottle is removed in the laboratory under a 0-5 per cent, solution of blood, and about 5 c.c. of the air allowed to bubble out, a corresponding volume of the blood solution entering. The cork is then re- placed, covered with a cloth to keep off the light, and shaken continuously for about ten minutes, when the haemoglobin will have reached the point of saturation corresponding to the percentage of CO present. The solution is then poured out into a test-tube, and the saturation determined with carmine solution in the manner described above. It is evident that as in eacli case the saturation found corresponds to a definite percentage of CO in the air, it is easy to calculate this percentage. The method furnishes good results with very small percentages of CO, but becomes less and less accurate as the amount exceeds 0-2 per cent. Sulphur Compounds. Sulphurous and sulphuric acids, sulphuretted hydrogen, and ammonium sulphide may all be present in the atmosphere of large towns, the first two invariably so, in traces; the two latter are, however, less often appreciable. SULPHUR COMPOUNDS 203 Is The external atmosphere of towns obtains sulphur compounds from the combustion of coal and gas. Their presence may be deleterious to health, and the oxy- acids of sulphur are unfavourable to vegetation and destructive to stone-work and mortar, upon which a scale of soluble calcium sulphate forms. Angus Smith considered sulphuretted hydrogen " one of the most deadly of gases," and held that, in traces even, " it lowers the tone of health." Sulphuretted hydrogen, in large quantities, has been ascribed as the direct cause of death among sewer-men by Stevenson, Haldane, and others. It is certain that an atmosphere con- taining from 07 to 0-8 of SHg per 1,000 of air is dangerous to human Hfe. The gas acts upon the nervous system, and causes a functional arrest of the respiratory centre in the medulla. Sulphurous acid in large quantities appears to favour the development of— even if it does not induce— bronchitis, asthma, anemia, conjunctivitis, etc. It may be estimated by aspirating . a known quantity of the air through a dilute solution of bromine in water, precipitating the sulphuric acid thus formed by barium chloride solution, and calculating the SO2 from the amount of BaS04 obtained. From 10 to 17 milligrammes per cubic metre of air have been estimated under the worst atmospheric conditions of London and Manchester. A cubic metre of air weighs 1,293,200 milhgrammes, and so this amount represents from o-oo8 to 0-013 part per 1,000. Sulphuric acid may be collected by aspirating a large volume of air through distilled water, and the estimation may be made by precipitating it as BaS04, as described under Water Analysis. Sulphuretted hydrogen and ammonium sulphide may some- times be detected by exposing to the air strips of filtering paper moistened with a solution of lead acetate. Any faint evidence of darkening about the borders of the previously white paper win prove the presence of these gases in the atmosphere. If the darkening is due to ammonium sulphide, filtering paper moistened with a solution of the nitro-prusside of sodium will show evidence of violet coloration if the gas is present in sufficient quantity. A quantitative estimation of sulphuretted hydrogen may be made by aspirating a measured volume of air through a little freshly prepared decinormal solution of iodine in iodide of 204 LABORATORY WORK potassium, to which some starch paste has been added. The operation is stopped as soon as the solution becomes colourless (H2S4-T2= 2HI + S). Each c.c. of the iodine solution employed X 1 7 ^milligrammes HjS in the volume of air examined. The vapour of CSg may be absorbed in a strong solution of potash in 06 per cent, alcohol; the contents of the flask are then acidulated with a little acetic acid. A small amount of calcium carbonate is next added, in order to nearly neutralize. The faintly acid solution is then mixed with an amount of water similar to the potash solution employed, and a little fresh starch solution is added. A solution of iodine in potassic iodide, con- taining 1-666 milligrammes iodine per litre, is then run in until a faint blue colour appears. Every c.c. of iodine solution re- quired = I milligramme CS., in the volume of air examined. In testing for hydrochloric, nitric and nitrous acids, large volumes of air must be taken. The acids may advantageously be absorbed in 10 per cent, pure soda lye. The amount of nitric acid in the air is very small. It is most marked after thunder- storms and in the air of towns. The above-mentioned acids may be estimated bv the methods described in Water Analysis. Chlorine and bromine may be absorbed in pure 10 per cent, colourless solution of potassium iodide (2C1 + 2KI= 2KCI-1-2I), and the hberated iodine titrated by decinormal sodium thio- sulphate with starch (12-69 milligrammes 1 = 7-99 Br = 3-54 CI). Traces of arseniuretted hydrogen in the atmosphere may be detected by aspirating air through a solution of cuprous cWoride in hydrochloric acid, and then causing it to impinge on a paper impregnated with mercuric chloride, the depth of tint of the yellow strain produced serving to indicate the amount of AsHg present. Phosphuretted Hydrogen. — This gas has in recent years been shown to be given off by ferro-silicon (employed in the manu- facture of steel) when this material is exposed to the action of water or moist air. There is very little danger from the low- grade ferro-silicon, but the danger is very great in respect to high-grade ferro-silicon containing about 40 to 60 per cent, sihcon. The phosphuretted hydrogen is deri\-ed from the calcium phosphide (CagPa) impurity in the ferro-silicon, which in contact with water or moist air is decomposed with the evolution of PH3 (phosphuretted hydrogen). The gas is intensely poisonous, experiments ha\ing demon- NITRIC, NITROUS AND HYDROCHLORIC ACIDS 20.5 strated a fatal effect on animals when the air contains but 0-25 per thousand of the gas. To test for the presence of the gas, air may be aspirated over filter-papers, one moistened with a solution of nitrate of silver and the other with a solution of the acetate of lead. If PH, is present, the nitrate of silver filter-paper is darkened, but not the acetate of lead paper; while SHg darkens both papers. It is, however, preferable to separate any SHg before testing the action of the air on the AgNOg paper. The amount of PH3 may be calculated by estimating the silver which is thereby precipitated from a standard solution of AgNOg (3Ag=PH3). Along with PHg a relatively small amount of AsHg may be liberated from ferro-silicon, and thus this equally poisonous gas may also gain access to the atmosphere. Table of the Amounts of Various Gaseous Impurities which have BEEN SHOWN TO INJURIOUSLY AFFECT HuMAN BeINGS. {Compiled from the Investigations of Lehmann, Matt, Gruber, Ogata, Fried- lander, etc.) Chlorine . . ^ Bromine • • I ^ , ,.,,.,>.. i. .. 0-002 — 0-005 per 1,000. Carbon bisulphide -> r . Iodine . . J Phosphuretted hydrogen! Hydrochloric acid . . |- . • . . o-oi — 0'05 ,, Sulphurous acid . . I Ammonia . . "| Sulphuretted hydrogen - . . . . 0-2 — 0-3 Carbon monoxide ) Carbonic acid . . . . . . 30 — 50 CHAPTER V OZONE— PEROXIDE OF HYDROGEN Ozone (O3). This gas is an allotropic oxygen, in which tlie molecule contains 3 atoms of ox3'gen instead of the 2 present in ordinary atmo- spheric oxygen. It is a gas with a peculiar phosphorous odour, and possessing marked irritating properties upon the mucous membrane of the eyes and nose and upon the respiratory tract. In nature it oxidizes oxidizable matter, and thus purifies air. It is best prepared artificially by passing electrical discharges through moist air, and hence it will be readily understood that it exists naturally in greatest quantities during and after thunder- storms, when it is also generally associated with nitric and nitrous acids and peroxide of hydrogen. The peroxide of hydro- gen is also a powerful oxidizing agent, by parting with some of its oxygen and becoming water (2H202=2H20-i-02). Nitrous acid also parts with its oxygen with great readiness. There are good grounds for doubting whether ozone ever exists in air in appreciable quantity, and whether it ever exceeds I part in 700,000. Certainl}' most of the observations of ozone hitherto recorded have included peroxide of hydrogen. According to Tid)^ — 1. Most ozone is found after thunderstorms, and least in damp and foggy conditions of the atmosphere. 2. More is found on the coast than inland, especially when sea- breezes are blowing. 3. More is found at high than at low levels. 4. More is found in country than in town districts. 5. More is found in wintei (especially after heavy snowstorms) than in summer. 6. More is found during the night than the day, and most at dawn. 206 OZONE — PEROXIDE OF HYDROGEN 207 7. The western winds in Great Britain contain more ozone than the eastern. Houzeau points out that the manifestation of ozone upon ozone papers is affected chiefly by the intensity of the winds in most cases, except where these blow directly off the ocean. 8. It is rarely, if ever, found in the air of occupied dwelling- rooms. Test. — A test for ozone which has been much employed is that of exposing to the atmosphere a white porous paper (filtering or blotting) previously soaked in a solution of potassium iodide and starch, and allowed to dry. Ozone will free the iodine, which then combines with the starch to form the blue iodide of starch, and FIG. 30. THE thus a blue colour is created (O3 + 2KI + H20= 2KHO + Ig + O2). The papers are exposed in a cage, and observations are taken at least every twelve hours. The cage aids in protecting the papers from direct sunlight, dust, and rain, each of which may lead to a subsequent fading of the colour; it consists (Fig. 30) of a double cylinder of very fine wire gauze; and projecting downwards from the under part of the lid is a small hook, to which the ozone papers are attached. The above-mentioned papers lead to errors of estimation from the following causes: I. Nitrous oxide (N2O3), peroxide of hydrogen and chlorine (each of which may also be present from electrical discharges in the atmosphere), and some volatile organic acids, produce 20S LABORATOKV WORK similar results upon the papers, and sulphurous acid and sul- phuretted hydrogen tend to destroy the blue colour. 2. The freed iodine is partially volatilized, and thus its effect is lost, while some of it may return to the potash and form inert iodide and iodate. 3. It is impossible to get uniform conditions—?'.^., the amount of light, moisture, temperature, and wind \'ary, and make results incomparable; and the purity and strength of the starch vary. .1 better test (Houzeau) is the Ijluing of faintly reddened litmus- paper previously moistened with a i per cent, solution of potas- sium iodide and dried, when the ozone liberates the iodine, and the alkaline potash formed gives the paper a blue tint. In the absence of hydrogen peroxide, ammonia is the only other gas in the atmosphere which can produce the same effect, and, conse- quently, another piece of the litmus-paper, not treated with potassium iodide, is exposed at the same time. Then any difference in the shades of the two papers must be furnished by ozone, which can be estimated by means of the ozonometer. Perhaps the best papers for general use are those saturated with a mixture of 15 per cent, solution of KI, and a sufficient quantity of a i per cent, alcoholic solution of phenolphthalein to render the liquid opalescent. These papers are coloured a fugi- tive red with ozone, while chlorine, bromine, or nitrous acid only give a blue or browTi coloration (Arnold and Mentzel). Ozone papers must be kept preser\'ed from the air in a tightly closed bottle, for to air containing the merest trace of ozone the papers react. But immediately before using them for test purposes they should be moistened, and the tint matched as quickly after the test as possible. Hydrogen peroxide reacts similarly to ozone upon all these papers, and any such ozone estimations are, in consequence, vitiated by this gas. Schone has pointed out that ozone blackens a bright piece of silver foil, but hydrogen peroxide has no such effect. More- over, chromic acid, whether in the solid form or in solution, de- composes even the most dilute peroxide of hydrogen, while it has no action on ozone. Engler and Wild find that the best test for ozone is by means of the chloride of manganese, which not only yields an extremely delicate reaction, but, in consequence of its hygroscopic character, keeps the prepared paper of the requisite moisture. Ozone turns such paper brown by the formation of manganese dioxide. OZONE — Peroxide of hydrogen 209 Hydrogen peroxide and nitrous acid have no such effect, but since ammonia and its carbonate turn these papers brown, the colour should be further tested by moistening with tincture of guaiacum, when, if the papers have been acted on by ozone, a blue colour will be developed even before the brown has had time to dis- appear; whereas no blue forms if the browning is due to ammonia. The method by which a quantitative estimation of ozone {i.e., " ozonometry ") is usually made is colorimetric. The intensity of the colour created by the ozone when the prepared papers are exposed to the atmosphere, generally for two hours, is matched against a standard scale (i to 10) of tints, each tint having been originally produced by exposing similar papers to known amounts of ozone. The greater the movement of the air the greater the quantity brought to act upon the paper, and hence less quantities of ozone present in the atmosphere on windy days may create more colour than greater quantities on still days. The only way, therefore, by which an accurate comparison of the ozone in different atmospheres can be made is by lining a dry glass tube with the ozone papers, and then aspirating similar quantities of air through such tubes. Peroxide of Hydrogen. H2O2 is generally present in traces, but it exists in con- siderable quantities during and after thunderstorms. It has been seen that it has similar properties to those of ozone; it may be distinguished, however, by certain of the tests given under " ozone," and also by the fact that it is only after the lapse of several hours that it reddens potassium iodide paste. A good test for the presence of hydrogen peroxide is the following: To some distilled water that has been made to take up the vapour add a drop of a i per cent, solution of potassium chromate, followed by a little ether and a few drops of dilute sulphuric acid. On shaking, the ether takes up the blue colour of per- chromic acid. The test is fairly delicate. 14 CHAPTER VI SUSPENDED MATTER IN THE AIR The nature of the suspended matter found in the atmosphere must necessarily ^■ary widely with the place and the circum- stances of its collection; and it would not be going too far to say that particles of ahnost everything the observer can see about ' him may be included. Obviously, the amount increases according to the extent to which the atmosphere departs from its state of greatest purity, high mountain air on the one hand containing few, and low town air containing many. It is more especially in factories and workshops that the examination of suspended matters is important, since both the nature and number of the particles have been shown to determine the prevalence of lung disease. These minute particles have a tendenc}^ to settle when the air is still, and the collection and microscopical examination of the dust which settles in a closed room furnishes a rough means of qualitati\'e examination. The Methods of Collection : Undoubtedly the best method is to aspirate large volumes (loo litres) of air slowly through small amounts (lOO c.c.) of dis- tilled water, placed in one or two small wash-bottles ; the bottles are then connected together and with the aspirator by rubber tubing, as shown in Fig. 28. The waters may then be mixed and evaporated to about 20 c.c, when drops may be mounted and examined by the microscope. This method can be made a quantitative one by aspirating a measured quantit}- of air through one or two w^ash-bottles, mixing the waters, and then counting the number of particles in an aliquot part ; or the water may be evaporated to dryness in a weighed platinum dish, and the weight of residue collected will be the weight of suspended matter in the volume of air aspirated, and after ascertaining the loss on ignition, this may also be SUSPENDED MATTER IN THE AIR 211 expressed as " volatile " and " non- volatile." This method is the most suitable for collecting trade dusts. If lead dust is so collected the lead may be dissolved out from the ignited total solids by nitric and hydrochloric acids, diluted, and then esti- mated colorimetrically. The results should be expressed in terms of milligrammes per cubic metre. Another method is by means of Pouchet's aeroscope. This instrument consists of a vertical glass cylinder, capable of being hermetically closed at either end by a copper ferrule. The ferrule at the upper extremity of the cylinder is a permanent fixture, and gives passage to a vertical copper tube which is partly outside and partly enclosed within the cylinder; of this tube the extremity of the part which is outside the cylinder is expanded into a trumpet-shaped mouth, and the end of the part which is inside the cylinder is gradually drawn to a very fine point, not more than 0-5 millimetre in diameter. The ferrule at the lower extremity of the cylinder is tem- porarily removed, so that a circular glass slide — which has been previously smeared with pure clean glycerine — can be placed with its centre immediately under the finely drawn point of the copper tube. The whole apparatus is then made air-tight and connected with the aspirator. The air which is thus sucked in falls in a spray upon the glass slide, and the glycerine retains the suspended matter. Subsequently the slide can be removed and examined by the microscope. A slight modification of Pouchet's aeroscope is the instrument of Marie Davy. The accompanying figure sufficiently explains itself- FIG. 31. M. MARIE DAVYS MODIFICATION OF POUCHET S AEROSCOPE. . Hesse's apparatus is seen by Fig. 32 to consist of a long glass tube connected at one end to the aspirator ; the small india-rubber cap which closes the other end is removed just before use, and 50 c.c. of pure glycerine is poured into the tube, which is then 212 LABORATORY WORK turned round so as to make the gl}'cerine coat the whole interior. As the air is subsequently aspirated through the tube, the suspended matter is caught up by the glycerine, which can be removed by a clean spatula and examined micro- scopically. But methods in which glycerine is employed are somewhat unsatisfactory, for the reason that the original glycerine will generally contain solid particles. A preliminary microscopic examination of the glycerine, however, does not entail much additional labour or time, and thereby the nature and amount of the foreign matter it contains can be previously noted. A fourth plan entails the use of a pure sugar filter through FIG. 32. HESSE's apparatus FOR THE COLLECTION OF SUSPENDED MATTERS IN THE AIR. a, The extremity connected with the aspirator; b, a removable cap. which the air is slowly drawn; the sugar is then dissolved in a sufficiency of pure water, when the suspended matters caught up in it are retained in suspension in the water, and may be collected and examined microscopically. The filter is best arranged as a glass tube, at least an inch in diameter, disposed horizontally, and packed (not too tightly) for se\-eral inches with the sugar crystals. One end of the tube is left open for the entrance of air, and the other connected by india-rubber tubing with an aspirator. The filter dissolved, the suspended matter may also be separated by filtration through a weighed Swedish filter-paper, then thoroughly washed and dried at a low tempera- SUSPENDED MATTER IN TITE ATE 2Tf; ture and woiglicd. If tlio amount of air aspiratcfl has l)een measured, the weighed quantity of its original suspcndf^d matter can be expressed quantitatively. It is obvious that the amount of the dust in town air must vary considerably; commonly the extent of this variation is between 5 and 25 milligrammes per cubic metre; it has been estimated as high as 224 in cement works during work. J. Aitken has devised an ingenious method of enumerating the particles of suspended matter in the atmosphere. A full description of the elaborate apparatus may be studied in the Proceedings of the Royal Society of Edinburgh, 1889. In this method a measured quantity of air is taken and passed into a receiver, where it is mixed with a large measured quantity of filtered (dustless) air, and saturated with water. The air in the receiver is then expanded by means of an air-pump; a shower of rain is thus produced which carries down the sus- pended matter, and the number of particles which fall on a measured area are then counted. From 10,000 particles per c.c. of air to over 2,000,000 may thus be obtained from the outside air, and in occupied rooms near the ceiling they may reach many millions. In the centre of London the average is from about 250,000 to 500,000 per c.c, while at the top of lofty mountains and in mid-ocean the dust particles may number only from 200 to 300. On the Terrace of the House of Parhament, London, the air was found to contain about 40,000 particles per c.c. Aitken has demonstrated that it is the dust in the atmosphere which determines mists and fogs, inasmuch as the condensation of aqueous vapour is not determined only by reduction of tem- perature but also by the presence of particles of dust, each particle becoming shrouded with a covering of moisture. By his method of enumeration it is assumed that each droplet has for its nucleus a dust particle. Vomer employs a blackened resinous substance which on cool- ing presents a uniform and polished surface. This surface may be protected from dust access prior to the experiment by means of a watch-glass sealed down with vaseline. For the purpose of an experiment the watch - glass is removed and an exposure of ten minutes is made; when, by means of an electric lamp, the number of dust particles per square centimetre may be counted. 214 LABORATORY WORK Dust collected from the top of a wardrobe in a bedroom yielded the following results on analysis: ]\Ioisture .... . . . . . . 4-4 Organic matter . . . . . . . . 52'6 Silica and insoluble silicates .. .. 21-0 Oxide of iron and alumina . . 97 Lime (CaO) 6-2 Carbonic acid, witli traces of sulphuric and phosphoric acid . . . . . . Tvi TOO-O The inorganic mattei was mostly amorphous, wliile the organic was for the most part organized. Among the commonest con- stituents of the latter were vegetable and animal fibres derived from fabrics, such as linen, cotton and wool. In addition there were a few feather barbs and particles of carbon ; squamous epithelial cells from the skin, starch granules, and a few pollen spores were also identified. The dust from 20 square yards of glass roofs at Kew and Chelsea was found to contain nearly 5 per cent, of SO3. equal to about 2 per cent, of S. The soot-fall from the atmosphere over industrial centres can be estimated by collecting it in a hopper of known collecting area, which terminates below in a small tube connected with a capacious bottle. The apparatus is similar to a rain-gauge, and it collects, of course, both rainfall and deposit. It has been found that in the city area of London some 650 tons of soot fall upon every square mile each year. There are marked differences between domestic and boiler soot. The latter is little more than dust or ash, practically all the hydrocaibons having been burnt; whereas the former possesses a high content of tar and volatile substances and a low content of ash. The Nature of the Suspended M.vtter of Air. The following substances may be found: Animal. — Debris from wear and tear of clothes, etc. ; wool and silk fibres ; human hair ; particles of feather ; debris of dried epithelial cells, and epidermic scales from skin; fragments of insects — i.e., scales from wings, legs; particles of the spider's web ; dried faecal particles from horses' SUSPENDED MATTEK IN THE AIR 215 dejecta; minute ova; amnebifoim organisms; mohjcular debris in considerable quantity. Vegetable. — Particles of carbonaceous matter ("soot"); molecular debris in large quantity; vegetable fibres, hairs and cells; cotton and linen fibres; starch grains; portions of plants, and pieces of woody fibre; pulverized straw; moulds, fungi, diatoms, and bacteria and their spores; pollen grains; algae, notably Protococcus -pluvialis and also the small oval cells of other unicellular algai. The spores and mycelium of A chorion Schonleinii and Tricophyton tonsurans have been found in the atmosphere of skin wards. Mineral.- — Especially numerous when the ground is dry. Minute particles of every chemical constituent of the soil may be raised up into the atmosphere — e.g., silica, silicate of alumina chalk, peroxide of iron, etc. Sodium chloride is invariably present, and is in greatest quantities at the seaside. Lead, arsenic, and zinc may be furnished by the wall-papers, paint, and " dryers" employed upon the walls of rooms; arsenic also from artificial fruit, flowers, curtains, etc., used for ornamenta- tion; coal dust, etc. There are certain trade dusts which vitiate the air of the immediate neighbourhood in which the trade processes are carried on; and particles of a great variety of substances may thus find their way into the atmosphere from coal, tin, stone, slate, cement, wood, clay, steel, flour, textile fabrics, glass, etc.; while poisonous matter may get into the atmosphere where lead, arsenic, copper, chromium, phosphorus, and mercury are being used for trade purposes. CHAPTER Vll THE CHARACTERS OF THE AIR COLLECTED FROM VARIOUS SOURCES— BACTERIOLOGICAL NOTE Marsh Air. The air collected over marshy regions is contaminated by the products of \'egetable decomposition. Such air contains excess of carbonic acid, commonly 0-05 per cent.; marsh gas may be markedly present; sulphuretted hydrogen is also sometimes appreciable ; watery vapour in large amount; ammonia in traces; phosphuretted hydrogen in faint trace. The suspended matter is found to mainly consist of vege- table debris, algse, diatoms, fungi, and other micro-organisms. In many cases where the presence of sulphuretted hydrogen is appreciable the marshy waters contain soluble phosphates, which become deoxidized to sulphides by reducing agents (chiefly organic matter) , and the sulphuretted hydrogen doubtless results from the action of vegetable acids upon these sulphides. Sewer Air. Sewer air varies in composition with the sewage and the state of the sewerage system. Its reaction is generally alkaline. Its temperature practically never falls below 9° C, and it is always saturated with moisture, or nearly so. Oxygen is variously diminished, according to the efficiency of the sewer ventilation; it is sometimes in normal proportions. Carbonic acid is variously- increased from the same cause; it probablv does not average in a good modern system of sewerage more than three times the amount normal to the atmosphere, but it may be ten times as great, or even more. Ammonia is somewhat in excess of the external air, and it may be greatly in excess. 216 SEWER Airs. 2T7 Sulphuretted hydrogen ^ may be present in variable quantities. Ammonium sulphide - but usually only in traces. If, how- Carbon bisulphide J ever, the sewage stagnates in the sewer, the sulphuretted hydrogen may be present in consider- able amount. Marsh gas is in traces, or absent; and traces of nitrous dioxide and phosphuretted hydrogen may be present. The fcetid and putrid organic vapours of sewage are, according to Odling, aUied to the compound ammonias, and are probably carbo - ammoniacal and contain traces of animal alkaloidal substances. The odour of sewer air is not usually due to sulphuretted hydrogen, but to minute quantities of a variety of volatile substances, such as indol, skatol, the mercaptans, and com- pound ammonias. The micro-organisms are almost exclusively moulds and micrococci, and these, together with animal and vegetable debris, appear to constitute the very sparse suspended matter. The micro-organisms in the sewer air are related more to the micro-organisms in the air outside than to those of the sewage (Andrewes and Law) ; they are generally fewer in number than those of the outside air at the same time. Splashing may, however, disseminate sewage bacteria in sewer air, and possibly also the bursting of bubbles or the ejection of minute droplets from flowing sewage (Haldane, Carnelly, Horrocks, Andrewes). The organic matter in sewer air probably averages from two to three times the amount in the outside air. The Air in Coal Mines. The oxygen is diminished, and the reduction may be extremely faint or so considerable that the total oxygen does not much exceed i8 per cent. The carbonic acid is increased, and may reach 1-5 per cent, or over. A trace of carbon monoxide is generally present. The considerable variations in the amounts of oxygen and carbonic acid in different mines are mainly de- pendent upon the ventilation provided. Marsh gas is sometimes in large amount, but it may be only in traces in some mines. A little sulphuretted hydrogen may be present, and in some mines high percentages of " black damp " or " choke damp " get into the atmosphere. " Black damp " is a mixture of nitrogen with a relatively small proportion (generally from 10 to 15 per cent.) of carbonic acid. 21 8 LABORATORY WORK Tlie marsh gas, or methane, mav be estimated in the following manner: The volume of carbonic acid present in a sample of the air is first determined by absorption in a lo per cent, solution of caustic potash, then the methane is burnt bj^ passing the air to and fro over a spiral of incandescent platinum wire. The contraction resulting from the combustion of methane is exactly double the CO2 formed. Methane on combustion produces its own volume of carbonic acid; and the carbonic acid so produced may be absorbed in the solution of caustic potash, and its volume thus measured. The contraction on combustion and the volume of carbonic acid formed would bear a different ratio to one another if the combustible gas were any other than methane. For useful particulars of mine air anaJyses the reader should consult " Methods of Air Analyses," by J. S. Haldane, M.D., F.R.S. Town Air during Fogs. Reactibn acid; oxygen is slightly diminished; carbonic acid very much increased — may even exceed 0-09 per cent. ; sul- phurous acid and sulphuric acids markedly present; carbon bisulphide in traces; maybe carbonic oxide, ammonia, am- monium sulphide or carbonate in traces; sulphuretted hydrogen generally in faint traces; watery vapour excessive; fine sus- pended particles of carbon and tarry matters, together with an increase of the commoner form of suspended matter in air. Ground Air. Ground air mav be drawn from considerable distances into a house, especially during periods of frost, owing to the aspirat- ing effect of the warmed and expanded air of the house itself; and the foul air of a leaky drain or cesspool may, under favour- able circumstances, be sucked through the earth into a dwelling for a distance of many yards. When it is borne in mind that many houses contain cellars built and ventilated considerably below the ground level, it will be realized that " ground air" must enter materiallv into the constitution of the atmosphere of such cellars. Ground aii contains an enormously high percentage of car- bonic acid, and the maximum amount of this impurity is always found between July and November, when the prevalent tem- perature and moisture favour the rapid decomposition of GROUND AIR 219 vegetable matter. The ground air of sandy soils contains relatively little COg. Ground air commonly contains traces of ammonia, sulphuretted hydrogen, and hydrocarbons. It is very free from micro- organisms. The CO2 increases with the depth of the soil and it is sometimes as much as 5 per cent, in deep soil a few feet from the surface. The entrance of ground air into ground-floor rooms, base- ments and cellars may be detected by comparing the carbonic acid found in these rooms with that in the external atmosphere, when any considerable excess of this gas (not otherwise accounted for) points to such pollution. The source of other impurities v- 8— --- FIG. 33. HESSE S APPARATUS FOR COLLECTING GROUND AIR. present may also be traced by collecting samples of the ground air in the vicinity of the house and comparing the results of such examination. A sample of ground air may be conveniently taken in the following manner: A sharp-pointed narrow steel cylinder, with numerous perforations just above its point, is driven into the earth to depths varying from i to 4 feet. The upper end of the tube is connected with a large air-jar, which is again connected to an aspirator. The connection between the jar and the steel cylinder is shut off, and the jar is first emptied (by means of the aspirator) of the air it contains ; the connection is then re- established and the sample collected by aspiration. 220 LABORATORY WORK Bacteriological Note. Bacteria are always present in air, but, unless the air con- tains a large number of dust particles, in comparatively scanty numbers. The determination of the number of organisms is of greatest use as a means of comparing methods of ventilation. Pathogenic organisms are not readily detected in air, the organisms usually found being moulds and saprophytic bacteria. Fliigge has shown that in inhabited rooms the air is liable to be contaminated with bacteria derived from the presence of droplets of mucus extruded from the buccal cavity in the acts of sneezing, coughing, and loud speaking. Gordon has greatly extended our knowledge of such particulate pollution, and has shown that certain bacteria, which can be detected and estimated, furnish means whereby these different kinds of pollution can be recognized. According to Gordon, pollution of three separate kinds can be recognized by bacterial tests. 1. Pollution from Material derived from the Uj)per Respiratory Passages. — Gordon has shown* that certain streptococci are present in enormous numbers in human sali\\a, and tliat their presence serves as a means whereby the addition of saliva to air can be detected. The organism specially characteristic of such pollution is the Streptococcus salivarius. 2. Pollution from Material detached from the 5Am.— The Staphylococcus epidermidis albus is constantly present on the human skin, and by its detection in air the presence of particles detached from the skin may be deduced. 3. Pollution by Material brought in from the Street on Boots. — Such material consists largely of horse-dung, and may be recog- nized by the presence of B. coli, spores of B. enteritidis sporogenes and Streptococcus equinus. * Local Government Board, Medical Ofiiccr's Report, 1902-03, p. 4 CHAPTER VIII SCHEME FOR THE DETECTION OF GASES WHEN PRESENT IN LARGE QUANTITIES Whereas for the detection of various gases whicli may con- taminate tlie atmosphere it is necessary to pass large volumes of air through distilled ammonia-free water containing agents with which tlie gases will combine, and then to apply the necessary tests, yet, when these gases exist in considerable quantities (as in the atmosphere of chemical manufactories, or those manufactories in which chemicals are employed), they may often be discovered by tests applied to small quantities of the air collected in air-jars. 1. Moisten two pieces of delicate red and blue litmus-paper in neutral distilled water, and catch these between the stopper and the neck of the bottle in such a way that they hang down into the bottles free of the sides. Note any change in the colour of these papers after waiting two or three minutes. 2. If the reaction is acid or alkaline pour rapidly into the jar a small quantity of distilled ammonia-free water {i.e., about 10 c.c), and replace the stopper at once; then shake vigorously, so that some of the gas may be taken up by the water. A. // the blue litmus-paper turns red {i.e., the gas is acid), it is either carbonic acid, hydrochloric acid, sulphurous acid, nitric or nitrous acid. Add a drop or two of a solution of silver nitrate to some of the water poured from the air- jar into a test-tube. {a) A white precipitate denotes the presence of either — I. Carbonic Acid. — Very slight precipitate, insoluble in nitric acid; acidity also very faint. Clear baryta water added to the jar becomes turbid after shaking, and the tur- bidity is increased by adding ammonia. 222 LABORATORY WORK 2. Hydrochloric Acid. — Marked precipitate, insoluble in nitric acid, but soluble in ammonia and potassium cyanide; acidity also marked. 3. Sulphurous Acid. — Marked precipitate, soluble in nitric acid; the precipitate on being heated clears up and the solution darkens (AggS). The water from the jar will decolorize iodide of starch solution; and if it be warmed after the addition of hydiochloric acid and zinc, a piece of lead acetate paper held over the test-tube becomes darkened (S02 + 3H2= SH2 + 2H2O). Odour character- istic. {b) There is no precipitate, and this fact denotes the presence of either — 1. Nitric Acid. — Add brucine and sulphuric acid to some of the water from the jar, and note the appearance of the pink zone changing to yellow and brown; or add a crystal of ferrous sulphate and then sulphuric acid to the water, and note the brown coating of the green crystal. 2. Nitrous Acid. — Add a drop of a solution of starch and potassium iodide, and then a drop of sulphuric acid. A blue colour forming at once denotes the presence of this acid ; or the Ilosva}^ test may be applied. B. If the red paper is turned blue {i.e., the gas is alkaline), it is either — 1. Ammonia. — Add a drop or two of Nessler's reagent to a Uttle of the water from the jar, when a yellow to orange colour appears. Odour characteristic. 2. Ammonium Sulphide. — Nessler's reagent causes a black colour to appear when added to some of the water from the jar; and a solution of nitro-prusside of sodium pro- duces a violet colour. Odour characteristic — i.e., that of rotten egg predominates, but it is easy also to detect the presence of ammonia. C. // the litmus is not affected {i.e., the gas is apparently neutral), it may be — Sulphuretted Hydrogen. — Lead acetate papers placed in the jar are darkened, as are also solutions of lead, iron, or copper salts. Odour characteristic. IDETECTION OF GASES 223 D. // the litmus-pap^.vs are firsL reddened and then slowly bleached, the gas is — Chlorine. — Filtering paper moistened in a solution oi potassium iodide and suspended in the jar is first dark- ened by liberated iodine and then bleached. Odour characteristic. Furnishes a red colour with a mixture of sulphocyanide of potassium and a proto-salt of iron. Note.' — Sulphuretted hydrogen has many reactions in common with ammonium sulphide — e.g., both gases will darken lead acetate papers and solutions of lead, copper, or iron salts, and their odours are closely similar ; but they may be readily dis- tinguished if attention is paid to the subjoined differences: Ammonmm Sulphide. — Alkaline reaction; produces a violet colour with a solution of nitro-prusside of sodium ; odour of rotten eggs and ammonia. Sulphuretted Hydrogen. — Neutral reaction; no effect upon the nitro-prusside; odour of rotten eggs alone. PART V FOOD EXAMINATION The following list shows the results (expressed as a per- centage of adulterated samples) of analyses of samples taken under the Sale of Food and Drugs Acts, during the year 1912, in England and Wales : Milk, 10-9 per cent. Wine, 10-6 Spirits, g-6 ,, Butter, 6-0 ,, Confectionery and jam, 5-5 per cent. Beer, 5-5 per cent. Sugar, 5-1 Coffee, 4-8 Cocoa, 3-4 per cent. Flour, 2-6 Cheese, 2-5 ,, Mustard, 2.5 ,, Margarine, 2-i ,, Pepper, o-8 Lard, 0-4 ,, Bread, o-o ,, Tea, 0-0 CHAPTER I MILK The Composition of Cow's Milk and of Other Milk. Milk consists of water, proteids, milk-sugar {C-y2^22^i-i) ^^^ mineral salts (chiefly phosphates of calcium, magnesium and potassium, and chlorides of sodium and potassium; very small quantities of sulphates are present, and traces only of carbonates, if any). The fat is suspended as minute globules, and since it forms the lightest element in the milk it tends to rise to the surface in the form of "cream," the largest globules being the first to separate. The milk early undergoes a souring, followed b}^ a natural separation into a solid " curd " and a liquid " whey." 225 13 226 LABORATORY WORK This change is caused by the fermentative conversion of the milk-sugar into lactic acid by the agency of micro-organisms which gain access to the milk, the chief of which is known as Bacilkts acidus ladicus. Acetic, succinic, and carbonic acids are also produced in small amounts. The " curd " is fomid to consist of casein, albumin, and traces of other kindred nitrogenous substances; the "whey," of the water, milk-sugar, and salts. A still later change is characterized by the appearance of a bluish tint, which is ascribed to the growth of another micro- organism, called Bacillus syncyamis ; and ultimately the casein decomposes owing to the access and development of putrefactive bacteria. The average composition of piire cow's milk is as follows: Water, 87-40. Solids, 12 -60, consisting of Sugar, 475. Fat, 3-65. Proteids, 3-48. ,]\Iineral salts, 072. \'ieth's ratio of the sugar, proteids and ash in cow's milk is 13 : 9 : 2. The milk of individual cows, collected imder circumstances which preclude the possibility of any sophistication, has, how- ever, been found to var}- considerably in its composition. There is no parellelism between rainfall and composition of cow's milk; and from experiments undertaken by the Board of Agriculture it appears that the excessive drinking of water by cows has no direct bearing on the composition of their yield. The circumstances which mainly determine the variations are : {a) The breed of the cow. Alderneys give most fat, and Longhorns most casein. {b) The time which has elapsed since the last milking. The longer the interval between the milkings the poorer the quality. Richmond finds that evening samples generally show from 0-3 to 0-4 per cent, more fat than morning samples. (c) The stage of milking. That which is first drawn (" fore- milk ") contains very little cream (under 0-5 per cent.); but towards the end of the milking the cream is very high in amount ; and the very last quantities drawn from the udder (" the strip- pings ") are almost pure cream. MILK 227 (d) The health of the animal. {e) The age ; young cows secreting milk of a poorer quality. (/) The time of year. The lowest fat occurs in April, May and June, and the highest in October; and during July and August the solids-non-fat are below the average. (g) The period which has elapsed since last calving, affecting the presence or absence of colostrum and the richness of the milk. The total solids (especially the fat) increase with the advance of the lactation period. (h) The food taken ; beet and carrots throw up the sugar. Where the amount of fat is very high the solids-non-fat are frequently below the average in pure milk, the deficiency being due to deficiency in milk-sugar and not to proteids or ash; and a cow yielding a high figure of fat in the afternoon may give below the average in the morning. It is very rare, however, that in dairy samples containing the mixed milk of several cows the non-fatty solids fall below 8-5 per cent., or the fat below 3 per cent. The following table (after D. Richmond) shows a comparison between the milk of various animals : Casein, Water. Albumin, etc. Fat. Sugar. Salts. Human . . 87-80 2-20 3-30 6-40 0-30 Cow 87-20 3-57 3-76 475 0-72 Ewe 79-46 6-68 8-63 4-28 0-97 Goat 86-04 4-35 4-63 4-22 0-76 Mare 89-80 1-84 I-17 6-89 0-30 Ass 90-12 1-66 1-26 6-50 0-46 It will be seen from this table that the proteid material varies from 2-20 in human milk to 6-68 in the ewe; that the fat is lowest in the mare's milk (1-17), and highest in the ewe's (8-63); that' sugar ranges from 4-22 in the goat to 6-89 in the mare; and that the water is least in amount in the ewe's milk (79-46), and greatest in that of the ass (90-12). Thus, adopting the cow's milk as a standard for comparison, human milk (though var5nng greatly with the period of lactation, etc.) shows an increased quantity of sugar and a slightly in- creased quantity of water; but all solid constituents, with the exception of sugar, are materially less, the total proteids amount- 228 LABORATORY WORK ing to less than one-half. Marc's milk is also richer in sugar and water; but the fat, casein, albumin and ash are considerably less. Goat's milk is richer in the solid constituents except sugar, and therefore contains a less percentage of water. Ewe's milk is characterized by the very high amount of fat, casein and albumin; the ash is higher than in cow's milk, but the sugar and water are less. Citric acid is a normal constituent, in minute quantity, of cow's milk and of human milk. The Milk of Diseased Cows. Although the milk secretion is in abeyance during some diseases, it is not so in all, nor is it so in all cases of the same disease. In a few conditions the milk presents somewhat definite chemical and microscopical characters; to the naked eye it may be all that is desired. It is in cattle plague and foot and mouth disease that the changes are most marked. 1)1 cattle plague the sugar is markedly diminished; the fat is increased, together with — to a less extent — the casein and salts (Gamgee) . Blood and pus are also commonly detected in the milk. In foot and motUh disease the milk commonly contains pus, blood, or mucus {i.e., in those cases where there is ulceration of the teats or abscesses in the udder) ; and, as in cattle plague, the milk corpuscles are seen under the microscope to display a tendency to aggregate into grape-like clusters. M'hen the disease is advanced, bodies resembling pus cells (though a little larger), and large yellow granular bodies, together with pus and blood cells, are also present. In this disease the results of chemical analyses vary so considerably as to be of no value for diagnostic purposes; but the milk separates remarkably quickly on the application of a gentle heat into curds and a pale blue whey; and this feature alone is considered as almost diagnostic by some Continental observers. In tuberculosis the milk is not at first appreciably affected, but the fat, lactose, and casein diminish toward the later stages of the disease. In garget the milk from the inflamed quarters of the udder is often thin and poor in solid constituents, and blood and pus may be present. CHAPTER II THE ANALYSIS OF MILK The Physical Characters: 1. Consistence. — The milk should be quite opaque when placed in a narrow glass tube; otherwise it has probably been watered, and haf^ a bluish tint. Sometimes it is thick and viscid (" ropy "), and on pouring has an appearance somewhat akin to mucus. Such milk has the property of imparting, when added in small quantities, its own peculiar quaUty to large bulks of good milk. This condition may be due to inflammation of the udder or to the growth of certain micro-organisms, whereby a mucinous substance results from changes in the milk-sugar and casein. It is usually accompanied by considerable acidification. " Colostrum " (the milk yielded during the first few days after the birth of the calf) coagulates on heating, owing to the larger quantity of albumin it contains; it is also more yellow than ordinary milk, shows flocculi, and has a slightly insipid saline taste. Some milks coagulate shortly after being drawn; these are, of course, very acid, and are generally yielded by cows in febrile conditions while suffering from inflammation of the udder. 2. Colour. — ^The colouring matter of milk (lactochrome) is in association with the fat globules and to a slight extent with the casein ; and as soon as the fat is separated in the form of cream, the original colour largely disappears. A good milk should be white with the faintest possible suspicion of yellow, although such food as buttercups, carrots, mangel-wurzel, etc., tend to increase the yellow colour. A marked yellow may also occur naturally in milk containing considerable quantities of colostrum corpuscles, where the animal is jaundiced, or where there are certain congestive conditions of the udder. The colour is also 229 230 LABORATORY WORK artificialh- furnished by dairymen b}^ means of the addition of colouring agents. Rarely, the freshly drawn milk is of a faint blue hue, or even green, or reddish. The cause of these colorations has been ascribed to the food consumed and also to micro-organisms. Such milks have been known to cause severe gastro-intestinal irritation; and more especially is this the case with "blue" milk. When these colours form slowly after the milk has been drawn, it seems probable that they are due to micro-organisms (such as B. cyanogemts). A pinkish hue is sometimes created by the presence of blood, but generally the blood tends to deposit. 3. Taste and Odotir. — Milk has the power of absorbing any odorous gases with which it comes in contact, and of acquiring and retaining distinct flavours from the food consumed by the animal secreting it; thus cows which have been feeding upon turnips, garlic, fennel, damaged ensilage, distiller}'- grain, etc., yield a milk which tastes of these articles ; and when bitter medi- cines have been administered, or chestnut or vine leaves have been eaten, or when the cow suffers from some forms of liver disease, a bitter flavour is imparted to the milk. A bitter flavour may also be produced by certain micro-organisms. 4. Reaction.— This may be neutral or alkaline when freshly drawn from the udder, but the milk is commonly amphoteric in reaction, turning red litmus-paper blue and blue litmus-paper red. This amphoteric reaction with litmus results from the presence of two salts with opposite reactions. The acid reaction is due to the primary sodium phosphate (NaH2P04), and the alkaline to the secondary phosphate (Na2HP04). Milk is generally faintly acid by the time it comes to be analyzed; if markedly acid, lactic acid fermentation has well set in; and if markedly alkaline, some alkaline salt (such as sodium bicar- bonate) may have been added. The acidity may be calculated by running into 50 c.c. of the sample ,F^ sodium hydrate (phenolphthalein being used as the indicator), each c.c. representing " i degree " of acidity; this multiplied by 0-009 gives the percentage expressed as lactic acid (Richmond). (A degree of acidity represents i c.c. of N. acid to the litre, or i c.c. of -^ acid to 100 c.c.) Milk as sold to the consumer should not curdle when shaken up in a test-tube with an equal bulk of spirit containing 70 per cent, of alcohol by THE ANALYSIS OF MILK 23I volume. On inclining the test-tube and bringing it back to the vertical position flakes or films adhere to the sides if the acidity is above 8 degrees. 5. Sediment.— Any foreign suspended matter is generally readily seen on the white background which the fluid itself pre- sents; and any such matter may be detected at the bottom of the cream tube after the milk has stood in this for several hours. Dirt (cow-dung, dust, grit, hairs, textile fibres, pus, blood, epithelium, etc.) will deposit on standing, especially if the milk is well diluted with water. The dirt in mflk may be estimated by taking 100 c.c. of the sample, centrifugaHzing this, and decanting the fluid portion. The sediment is then shaken with 15 c.c. of 10 per cent, ammonia, the mixture being diluted after the lapse of one hour with water and again centrifugalized. The opalescent liquid is then de- canted, the sediment washed with water into a weighed platinum crucible, and further washed successively with alcohol and ether. The crucible and its contents are then dried at 100° C until constant in weight (Fendler and Kuhn). Useful comparative data as to dirt may be obtained by filtering samples through cotton discs, when the discs (which should be supported on wire gauze) are i inch in diameter and a pint of each sample is so filtered. If a vacuum pump is employed the milk is quickly drawn through the disc. Eau de javelle completely dissolves such cellular elements as leucocytes, leaving dirt which has gained access since the milk left the udder, so that it may be separately collected. To make this preparation, 20 grammes of good bleaching powder are rubbed up in a mortar with lOO c.c. of water, and the emulsion mixed with a solution of 20 grammes of anhydrous potassium carbonate dissolved in 100 c.c. of water. After thorough mixing, the gruel-like mass is allowed to stand for an hour, and then filtered under pressure. The clear yellowish liquid keeps well in the dark, but prior to use it should, if necessary, be again filtered. A. W. F. Lowe has suggested a test for the presence of bile- salts in order to prove that the dirt contains dung. A little grape-sugar is dissolved in a watch-glass containing the sediment, the liquid is then removed as closely as possible by decantation, the sediment is dried at 100° C, allowed to cool, and a drop of pure sulphuric acid is run over the particles, when a fine cherry-red 232 LABORATORY WORK crimson colour appears in the presence of bile-salts. The colour develops around the particles, and it is well to employ a magnify- ing glass for their examination. " Incows clinically tuberculous the faeces contain large numbers of living tubercle bacilli " (Third Interim Report of the Royal Commission, appointed in 1901, on Human and Animal Tuber- culosis), and this circumstance is probably responsible for the bulk of the infection of milk by this germ. The milk must be fresh at the time of analysis, as after lactic acid fermentation of the milk-sugar has set in there is a slight loss in the non-fatty solid matter. One or two drops of formalin will keep a sample fresh for several days. FIG. 34. THE CREAM TUBE. The sample should in every case be thoroughly mixed before any part is removed for analysis. The Cream. — Some of the milk is poured into a " cream tube." This is a glass cylinder, the upper part of which bears markings that show the proportion which the cream on separation forms to the total volume of the milk. The milk is made to stand exactly up to the level of the zero of the scale (due allowance being made for capillarity), and it is then set aside for twenty-four hours. Supposing the cream is found to reach down to 10 on the scale, then it is 10 per cent. ; and so the volume of cream is read off against the graduated scale. Generally the cream will have separated in twelve hours, but the separation is not complete in all samples until twenty-four hours; if the time is protracted beyond twenty- four hours, partial drying ensues, and the resulting contraction will eventually leave THE ANALYSIS OF MILK 233 a space between the lower surface of the cream and the upper surface of the milk. Good milk throws about lo per cent, of cream; but the milk of an Aldemey cow may yield between 30 and 40 per cent. Fresh cream contains a very variable proportion of fat. Gener- ally the amount falls within 40 and 50 per cent. ; but there may be 25 per cent, or less when the cream is considerably diluted with milk. Clotted cream generally contains from 45 to 60 per cent, of fat. Before the specific gravity is taken, any frothing (from shaking or pouring into the glass cylinder) must first be allowed to pass off. Westphal's balance is a rapid and a more exact method of obtaining the specific gravity of milk than a hydrometer (lacto- meter) {vide p. 8). The specific gravity of distilled water at 15-5° C. being taken as 1,000, that of pure milk at the same temperature is commonly about 1,032. The fat is so much lighter than the remainder of the milk that with its removal the specific gravity rises even higher still. A specific gravity much above 1,032 will therefore create suspicion as to the removal of the lighter element (cream) from the milk. The addition of water will lower the specific gravity again, for it is obvious that the specific gravity of water being 1,000, the more of this is added the nearer will the specific gravity of the mixture of milk and water be reduced to 1,000. An abundance of cream will also account for a low specific gravity of milk, so that a low specific gravity may mean either abundance of cream or the addition of water. It follows, then, that the specific gravity and cream tests con- sidered together afford a valuable clue as to the nature of the sample. The total solids are estimated in the following manner: Pipette 5 c.c. of milk into a weighed platinum dish. Curdle this by adding to it a few drops of a mixture of one part acetic acid to nine parts of methylated spirit; this will prevent any skin forming on the surface, and will greatly hasten the drying. Dry first on the water-bath, then for two hours inside the water- oven at a temperature not exceeding 105° C. Let cool and w^eigh. From this total weight subtract the weight of the dish. Multiply the remainder by 20 to bring it to 100 parts. Now convert the 100 c.c. of milk to weight, by deducing this from the specific gravit}/, and then calculate the percentage of solids by weight. 234 LABORATORY WORK Example. — Solids + dish weighed 12-903 grammes. Deduct weight of dish 12-203 0-700 gramme in 5 c.c. Multiply by 20 = 14-00 grammes of solids in 100 c.c. of milk. Now, if the specific gravity of the milk is 1,030, the weight of milk as compared with that of distilled water is as 1,030 is to i.ooo; and as 100 c.c. of water weigh 100 grammes, 100 c.c. of tlie milk will weigh 103-0 grammes. And 14-0 grammes of solids in 103-0 grammes of milk = 13-59 P^r cent. Richmond's formula for total solids is: 0-25 G+i-2 F + 0-14; where F==fat and G = the last two figures of specific gravity and any decimal. The result maybe rapidly obtained by means of Richmond's Slide Rule. Mineral Ash. — Ignite the total solid residue until all dark specks, etc., have disappeared, and nothing but a perfectly clean whitish ash remains. The ignition must be effected slowly and at as low a temperature as possible; Bell recommends that an Argand burner should be used in preference to a Bunsen, on this account. The ash is then weighed, and its percentage amount by weight ascertained in a similar manner to the total solids. Too large a proportion of ash (that is, above 0-75 per cent.) points to the addition of mineral matter. A milk may have, on the other hand, a paucity of ash, due to the copious admixture of water. Effervescence on the addition of hydrochloric acid denotes adulteration by a carbonate, which will generally be sodium car- bonate. The ash of a pure milk does not effervesce when hydro- chloric acid is added to it. The Fat. — There are many methods in use at the present day for the extraction and estimation of fat. The student is recom- mended to employ Schmidt's process, and, where necessary or desirable, to corroborate results by Adam's process. The Werner-Schmidt Process. This process has become a favourite one, for by it a very accurate estimation of the fat can be made in a short space of time. The process is as follows : I. A specially graduated tube, as shown in Fig. 35, is employed to receive 10 c.c. of milk, to which 10 c.c. of strong hydrochloric THE ANALYSIS OF MILK 235 acid is added; the milk and acid thus standing to the mark ot 20 c.c. on the tube. 2. The mixture is boiled, with frequent shaking, until it turns a brown colour (from the conversion of milk-sugar into maltose and caramel) . 3. Let stand for about three minutes, then cool by immersion in a stream of water. 4. Fill up to the 50 c.c. mark with ether; cork the tube and invert it three times; then set aside for fifteen minutes, when the ether will have separated. 5. Accurately pipette off 20 c.c. of the clear supernatant ethereal solution of fat into a weighed flask, and evaporate off FIG. 35. — stokes' tube FOR THE WERNER-SCHMIDT PROCESS. the ether, until the last small bubble disappears. A naked flame must not be brought near to the ether, so it becomes necessary to drive off the ether by placing the flask in hot water. The flask can be attached to a condenser and the ether recovered for subsequent use. 6. Dry in air-bath at 100° C, and weigh the residual fat. 7. Next notice how many c.c. of ethereal solution remain in the tube; then from the fat estimated in the 20 c.c. calculate the amount of fat in the whole of the ether. Example. — Ten c.c. of milk, with a S.G. of 1,031, gives in 20 c.c. ethereal solution 0-277 gramme of fat. In the tube there remained 6-5 c.c. of ethereal solution, making 26''^ X 0'27'7 a total of 26-5 c.c. .-• '^^-^ =0-367 gramme of fat in the 10 c.c. of milk, or 3-67 grammes in 100 c.c. But 100 c.c. of milk with a specific gravity of 1,031 weighs 103-1 grammes. Therefore 236 LABORATORY WORK there are 3-67 grammes of fat in 103-1 of milk, or 3-559 per cent. Notes upon the Process. — There floats between the brown mixture' of HCl and milk and the ethereal solution, a fluffy stratum of casein. Three-fourths of this stratum should be taken as ether in reading off the quantity of the latter. The acid and milk should not be boiled together for more than two minutes, or the ether takes up a caramel-like substance. The whole process does not take more than forty minutes ; and it is well adapted for use where the milk has decomposed. The boiling with HCl converts the albumin into soluble acid albumin, and the fat is then fully exposed to the action of the ether. Adam's Process. This process, by which an extremely thin layer of milk is spread over absorbent paper and the fat then extracted by ether, gives very exact results. 1. Shake the sample, and pipette 5 c.c. into a small beaker about 2 inches deep and 1 i inches wide, then weigh. 2. Soak up as much of the milk as possible — and in every case almost the entire 5 c.c- — by a coil of white demi-blotting-paper keeping the beaker covered during the absorption. It is, of course, imperative that the paper should be freed from fat prior to its employment. This may be done by extracting it with acid alcohol (alcohol containing 10 per cent, of acetic acid) for at least three hours, and then thoroughly drying; or the specially prepared slips of fat-free paper made by Messrs. Schleicher and Schiill, of Diiren, may be used. A hehcal coil is prepared by rolling a strip of the paper upon a glass rod of the size of a cedar pencil, care being taken not to tear the paper ; and the coil may be held together with platinum wire. 3. Remove the coil by its upper part, and place it dry end downwards upon a slip of glass, and then re-weigh the beaker with the trace of milk left behind in it. The difference in weight from the previous weighing represents the weight of milk soaked up by the coil. 4. Dry the coil in the water-oven for two hours, and then extract the fat by anhydrous ether in a Soxhlet, twelve syphonings at least being necessary {vide p. 9). 5. Receive the fat and ether in a small light flask; drive off THE ANALYSIS OF MILK 237 the ether; dry to constancy in awater-oven at about 105°^. the flask being laid in a horizontal position ; let cool, and weigh the fat. Notes upon the Process.-^By the addition of ammonia sour milk is as easily taken up as fresh. A good anhydrous ether may be prepared by placing the com- mercial article of S.G. 0720 for three weeks over qmckhme, and then distilhng. To obviate the two weighings, some analysts apply the process as follows : , . 1 j, 1 • Suspend a strip of ^at-free filter-paper over a lighted gas-ring at such a distance above it as would not be too hot for the hand to bear Distribute over this from a pipette 5 ex. of miik When dry, roll up the coil and place it in a Soxhlet's fat-extractor with some anhydrous ether. Cause the ether to syphon over at least twelve times into a weighed flask. Finally, drive off the ether and weigh. Calculate as described above. For rapid " samphng " purposes the foUowing process is useful, and the results are closely approximate to those obtamed by other processes : M. Gerber's Modification ol the Leffmann-Beam Process. A smafl test-bottle is employed with a thin, graduated neck; into this is placed 10 c.c. of sulphuric acid (specific gravity 1-82^ at 15.5° C), I c.c. of amyl alcohol (specific gravity, O'Sib at IV5° C), and 11 c.c. of the milk sample. The purity and strength of the acid are very important for securing accurate results The alcohol and milk are allowed to flow down the side of the bottle, so that the three hquids form distmct layers; the bottle is then firmly corked and smartly shaken until the curd is dissolved. It is then held upside down to aUow the acid to run down the neck, and this may be repeated two or three times to ensure that all the ingredients are well mixed. The bottle is then placed in a centrifuge, which is spun for three minutes (at about 1,000 revolutions per minute); or for nine minutes in the case of separated milk. If only one sample is being tested, another bottle should be filled with milk and placed opposite the sample bottle, in order to balance it m the centrifuge. Upon removal the test-bottle is placed neck down- wards into water for two or three minutes at a temperature of 60 to 70° C. to keep the fat hquid. The percentage weight of fat may now be read off in the graduated neck of the bottle, each of 23vS LABORATORY WORK the finer divisions on which indicate o-i per cent, bj^ weight of fat. In making this reading the bottle should be held vertically, and by slightly moving the cork at the bottom of the bottle either upward or downward, the lower level of the fat column in the graduated neck may be made to correspond to one of the longer markings which indicate whole percentages of fat, and from this base line the percentage of fat can be easily and rapidly read off. FIG. 36. — CENTRIFUGAL MACHINE, FITTED WITH TWO GEARS, THE LOW GEAR WORKING UP TO 3,000 REVOLUTIONS PER MINUTE, AND THE HIGH GEAR UP TO 10,000. Hehner and Richmond have devised a formula by means of which the fat in ordinary milk samples may be estimated: F= 0-859 T-0-2I86 G; where F=fat, T= total solids, and G=the last two units of the S.G. together with any decimal {i.e., if S.G. = 1029-5, G=29-5). The solids-not-fat may be calculated by subtracting the fat from the total solids. The Analysis of Sour Milk. — ^Thorpe has sliown that tlie l^utter- fat suffers little, if any, alteration during the souring of milk, but that the non-fatty solids are more or less affected by fermentative THE ANALYSIS OF MILK 239 changes. The principal constituent which suffers change is lactose, from a portion of which lactic acid forms at an early stage. As a rule less than half of the lactose (which averages 475 per cent, in cow's milk) is thus transformed. Concurrently with the formation of lactic acid, there are pro- duced products which are either gaseous at ordinary tempera- tures or are volatihzed during the operation of determining the non-fatty contents of the sour milk. The aggregate weight of these substances (acetic and butyric acids, ethyl alcohol, carbonic acid, and traces of ammonia) is not very large, but it is sufficient to affect any estimation of the degree of sophistication to which the milk may have been subjected. The total loss of solid matter from all causes ranges, as a rule, from only o-2 to 0-5 per cent, by weight of the milk, and nearly the whole of this is accounted for by the transformation of lactose into alcohol and volatile acid, the changes in the weight of pro- teid material being relatively insignificant. In the analysis of a sample of sour milk it is therefore recom- mended that the milk be thoroughly mixed with a wire whisk ; the lactic acid in the weighed quantity is then neutralized with -^ solution of strontia (using phenolphthalein as indicator), and from the total solid residue of the milk the weight of strontia added is deducted (each c.c. of y^ strontia= 0-00428 gramme). As regards the alcohol, 100 grammes of the milk are distilled, and the distillate redistilled after being neutralized with ^ caustic soda solution, litmus-paper being used as the indicator. The specific gravity of the distillate, made up to the original bulk, is determined in a 50-gramme pycnometer, and the quantity of alcohol corresponding to this specific gravity is deduced from a table. The percentage by weight of alcohol, multiplied by f f , gives the percentage amount of lactose which has disappeared in the production of the alcohol. The amount of volatile acid is ascertained as follows : Ten grammes of the milk, contained in a platinum capsule, are •neutralized to the extent of one-half the total acidity .(previously determined on another portion) with |^ caustic soda, and a little phenolphthalein is added. The mixture is then evaporated to dryness on a water-bath with frequent stirring, and after treat- ment with about 20 c.c. of boiling distilled water, so as to break up and thoroughly detach the milk solids from the capsule, a further addition of ^ caustic soda is made, until the neutral 240 LABORATORY WORK point is reached. The difference between the original acidity of the milk and that of the evaporated portion is regarded as acetic acid. The production of each molecule (60 parts) of this acid denotes a loss of i molecule of carbon dioxide and i of water — that is, a loss of 62 parts of the original lactose. The entire correction, which is always additive in a properly sealed sample from three to six weeks old, is fairly constant, and generally ranges between o-2 and 0-3 per cent. Microscopical Examination. — Ten c.c. of milk should be diluted with about 30 c.c. of water, then centrifugalized (at least 1,500 revolutions per minute), the supernatant fluid decanted, the sedi- ment washed with water and placed on glass slides, and the contents of the slides examined. Normal milk under the microscope consists of a collection of round highly refractile oil globules of about the same dimensions, with an occasional epithelial cell ; from three to eight days after calving colostrum corpuscles are also present in larger or smaller if.. V.O-' FIG. 37. — MILK SHOWING THE L.\RGE COLOSTRUM CORPUSCLES. (X25O.) quantities (Fig. S7)- These latter mostly consist of large yellow cells containing larger and smaller fat globules in their interior. Where, however, the animal is not in health, the following abnormal constituents may also be found: Cast of the lacteal tubes, blood-corpuscles (which closely resemble those of the human subject), pus cells and leucocytes, and various micro-organisms {e.g., fungi, such as Oidium lactis ; moulds, such as penicillium; and bacteria). Blood may be detected either by the spectroscope or microscope. When present in considerable quantities blood tinges the milk and has a tendency to settle as a brown deposit ; or after warming the milk to about 50° C, a high speed centrifuge may furnish a red deposit in the milk tube. Cow-dung shows vegetable parenchyma and vessels of a dis- tinct yellow tint. In the deposit there may be also detected yeast cells, cotton fibres, hairs, etc. All milk samples contain a certain number of cellular elements, THE ANALYSIS OF MILK 24I and in certain pathological conditions their number is enormously increased. Deductions of great value can be made from accurate determinations of their number in the milk of individual cows, but for mixed milk samples the determination is of much less value. If the diluted milk is centrifugalized, the sediment spread evenly over a cover-slip, dried, and stained by methylene blue, the number of leucocytes may be enumerated by means of the ordinary Thoma-Zeiss blood-counting apparatus. It has been suggested that the number of these cells should not exceed 500,000 per c.c, but it is probable that such a limit would at times lead to the condemnation of milk derived from healthy sources. These cellular elements have been usually regarded as leucocytes or pus cells, and hence it is that a standard has been suggested for the hygienic control of milk. There are good grounds, however, for believing that most of the cells found are young epithelial cells and not leucocytes. Savage has classi- fied these cellular elements into polymorphonuclear cells, lym- phocytes, large leucocytes, and doubtful cells. Owing to the difficulty of recognizing pus cells from other cells that may be present, it is not easy to say how frequently pus resulting from inflammatory processes in the udder gets into milk ; but in a large percentage of cases cells indistinguishable from pus ceUs are present. Lymphocytes, which stain deeply and possess nuclei which occupy nearly the whole of the cells, are not normal constituents of milk except within three weeks of parturition; but, like polymorphonuclear leucocytes, they are occasionally present for a few days in the milk of an apparently healthy cow. These bodies are found associated with staphylococci and strepto- cocci generally in diseases of the udder, and they are to be found associated with tubercle bacilli in tubercular mastitis. A high cell count accompanied by streptococci apparently indicates some udder trouble. 15 CHAPTER III THE SOPHISTICATION OF IMILK— MILK PREPARATIONS-^ I^IILK STANDARDS— BACTERIOLOGICAL NOTE Cow's milk should be the normal, clean, and fresh secretion obtained by completely milking the udder of the healthy cow, properly fed and kept. The Addition of Water. — Whether milk is naturally poor or has been made so by the addition of water, the dairyman who sells it defrauds the purchaser, for the latter demands and pays for pure milk of average quality. It is clear that the percentage amount of both fatty and non- fatty solids will be reduced by any addition of water; but the estimation of the amount of added water is always made from the non-fatty solids, because these depart less from the average than is the case with the fatty solids. The legal low limit for non-fatty solids is one of 8-5 per cent. Supposing, then, a sample yields 8 per cent, of non-fatty sohds. Then if 8-5 per cent, of non-fatty solids denotes 100 per cent, of pure milk, 8 per cent, denotes only about 94 per cent, of pure milk. Therefore there is about 94 per cent, of pitre milk in the sample, and (100-94=) 6 per cent, of water has been added. It is sometimes maintained by the milk-vendor that no water has been added and that the milk is naturally poor milk. The freezing-point test is said to be of value in settling this question, for milk as drawn from the cow freezes at - o-550°C. whatever its solids-non-fat content ; but the addition of water alters the freezing-point, and a freezing-point above - o-530°C. is said to be conclusive evidence of watering. The ash should in every case be low when the solids-non-fat arc low, or some mineral adulterant has been added. 242 THE SOPHISTICATION OF MILK 243 Cream Abstraction.- — Though the milk from the same cow may vary at times, the mixed product of many animals (" dairy samples ") varies but little. The legal low limit of 3 per cent, of fat is one which is reached by all genuine dairy samples obtained from a fairly good herd of cows, kept and fed under average conditions; though, as Bell and others have shown, the milk of individual cows may some- times fall below this limit. The percentage reduction of fat (by the removal of cream) is an easy calculation after an estimation of the fat has been made. Suppose that the fat has been found to amount to just 2*5 per cent. Then 3 per cent. -2-5 per cent. = 0-5 per cent, of fat has 2' ^ X TOO been abstracted from the milk; or =83*3 per cent, of the original fat remains and (100 -83-3= ) 167 per cent, of the total fat originally in the milk has been removed. In cases where the fat is low and the solids-non-fat are high there can be little doubt that fat has been abstracted, and that the low fat is not due to the dilution of. the milk with water. The " toning down " of good milk by the addition of separated milk is much practised, and large numbers of the samples analyzed are found to barely reach the low legal limit of 3 per cent, of fat. As milk stands, a certain proportion of the fat quickly rises to the upper layers, and a defence is sometimes set up by the dairyman that a poor sample was due to the fact that such top milk had all been sold, and the sample was some of the last of the milk in the can. This defence is, in many cases, a wcU-recognized subterfuge, for it is the duty of the vendor to mix the milk and to supply fair samples to one and all alike. Failure to draw off the " strippings " no doubt often accounts for the low figure of fat in milk. Samples collected on Sunday mornings are generally amongst the poorest, for dishonest tradesmen adiilterate on these days in order to meet the extra demand, due to the fact that more people take their meals at home on that day. In addition to water, there are other adulterants added to milk. Chalk and starch were formerly used, but they are very rarely, if ever, employed at the present day. Sodium carbonate is rarely used to preserve the milk and to neutralize it when sour. It may be tested for (E. Schmidt) by adding 10 c.c. of 244 LABORATORY WORK alcohol to 10 c.c. of milk, followed b}' a few drops of i per cent, solution of rosolic acid. Pure milk yields a brownish-yellow colour, but if sodium carbonate or borax is present a more or less marked rose-red colour appears. Boric and salicylic acids, borax, " formalin," benzoates, fluor- ides, peroxide of hydrogen, have been used as milk preservatives. Either a mixture of boric acid and borax, or " formalin," has been generally employed. Such chemical preservatives are now prohibited in anj^ form of milk by the Public Health (Milk and Cream) Regulations, igi2. Boric acid with borax is largely added to milk during the summer months, and the amount generall}' employed is about 5 grains to the pint. Experiments go to show that not less than 4 grains of a mixture of boric acid and borax are necessar}- to preserve a pint of milk for twenty-four hours in warm weather (Rideal, Foulerton). Salicylic acid is not so frequently employed, because of its lesser solubility and unpleasant taste. " Formalin " is a commercial preparation containing about 38 per cent, of formaldehyde. " Mystin " (a mixture of formic aldeh3?de and sodium nitrite) has occasionally been emploj^ed as a preservative. Annatto and turmeric, coal-tar d3'es and saffron, are j-ellow colouring agents which arc added to give the milk a rich yellow appearance. Annatto and coal-tar dyes are chiefly emplo3'ed. The tests for the antiseptic and colouring agents in milk are given in Chapter XIII., which treats of the subject of Anti- septics and Colouring Agents in Food. Cream. — Starch or gelatine is sometimes added to cream to thicken it. Gelatine may be detected by adding to 10 c.c. of the sample, 20 c.c. of cold water and 10 c.c. of a solution of acid nitrate of mercury. The whole is then well shaken, allowed to stand for live minutes, and then filtered. If much gelatine is present, a clear filtrate cannot be obtained. A portion of the filtrate is mixed with an equal quantity of a saturated aqueous solution of picric acid, when a yellow precipitate forms if gelatine is present (Stokes). Starch is detected by the bluing with iodine solution. Milk solids and other fats, and lime in cane-sugar sjTup, have THE SOPHISTICATION OF MILK 245 also been added to cream. Sucrate of lime may be detected by estimating the lime in the ash of the cream (average percentage of Ca0=22 per cent, of the ash). The same preservatives are employed as in the case of milk. Under the Pubhc Health (Milk and Cream) Regulations, 1912, no preservative may be added to cream which contains less than 35 per cent, by weight of milk fat, whereas in cream con- taining 35 per cent, or more of milk fat the only chemical pre- servatives permitted are boric acid, borax, or a mixture of these, and hydrogen peroxide; but the addition of these preservatives is subject to a system of declaration. Furthermore, no thicken- ing substance may be added to cream. In these Regulations " thickening substance " means sucrate of hme, gelatine, starch paste, or any other substance which, when added to cream, is capable of increasing its thickness. Neither cane nor beet sugar are to be regarded as a preservative or as a thickening substance. Hand-skimmed milk is sometimes made to look like good rich milk by the addition of condensed milk. An analysis of the ash and non-fatty sohds will detect the fraud, since these will both be in excess of their general proportions (more especially the sugar), and the amount of soluble albumin will be diminished (Faber) . Hand-skimmed milk is generally slightly acid, and the specific gravity is above 1032-5. The fat generally amounts to from 0-5 to 1-5 per cent. The bulk of the samples of " separated milk " contain from 0"2 to 0*3 per cent, of fat. Skimmed and separated milk must legally contain at least 8 7 per cent, of solids-non-fat. Separated milk is sometimes " enriched " — that is, the butter- fat taken out by the separator is replaced by an emulsion of some other fat. In such case a separate analysis of the fat of " the cream " must be made by the Reichert-Wollny process, as described in the analysis of butter. Condensed milk may be unsweetened or sweetened whole milk (unskimmed or non-separated) concentrated to about one-third of its original volume, and cane-sugar added; or it may be pre- pared from sweetened skimmed or separated milk. The following table, taken from Dr. Coutts's Report to the 246 LABORATORY WORK Local Government Board {191 1), indicates the percentage com- position of the chief classes of condensed milk upon the market : Full Crbam. Machine Skimmkd. Sweetened. Unsweetened. Sweetened. Lowest. Highest. Lowest. Highest. Lowest. Highest. 1 Total solids . . Protein . . Fat Lactose . . Ash Cane-sugar 68-1 7-3 8-0 II-6 1-6 36-1 83-6 II-4 13-7 I7-6 3-4 44-6 29-2 8-0 8-2 II'I 1-6 Nil 38-0 lO'O II-9 i6'0 2-5 Nil 56-9 7-6 O'l iO'9 1-6 30-4 79-1 12-3 6-5* 17-0 2-9 52-6 In the analj'sis of condensed milk 20 grammes should be taken and made up to 100 c.c. with water as a stock solution. For Total Solids. — Evaporate 5 c.c. of this as in the case of milk. For Ash. — Incinerate the above. The ash averages 1-9 to 2 per cent. For Fat. — Estimate by Adam's process. The fat averages about 10 per cent. The best method for obtaining a rapid and accurate estimate of the fat in sweetened brands is the Gottheb process (the Werner- Schmidt process being inapplicable) : Into a graduated 100 c.c. tube put 10 c.c. of above solution; add I c.c. of 40 per cent, ammonia; warm to 30° C; shake; add 10 c.c. alcohol (95 per cent.), and shake. Add 25 c.c. ether, and shake; add 25 c.c. petrol-ether, and shake; let settle; pipette off 25 c.c. of the mixed ether solution, evaporate this and weigh. Calculate as in the Schmidt process. The fat of sour milk, cream, cheese, butter, etc., may aJl be reliably estimated by this process. For Total Sugars (cane and milk). — To 10 c.c. of stock solution add 40 c.c. of methylated spirit, add one drop of acetic acid, shake (this will precipitate the curd and fat), filter. Evaporate 20 c.c. of the filtrate, and weigh the residue. Now incinerate this, and subtract this ash from the total sugar weight. Multiply the difference by 22 5, and then by 5, to give percentage of mixed sugars. Apparently one should first multiply by 25, but allow- ance has to be made for the volume occupied by the precipitated * A partly skimmed milk. CONDENSED MILK 247 curd and fat. Deduct milk-sugar (estimated Ijy Fehling's solu- tion), and the difference is cane-sugar. The milk-sugar (which averages 13 to 15 per cent.) is deter- mined by titrating a 5 per cent, dilution of the milk with Fehling's solution {vide pp. 340, 341). The cane-sugar may subsequently be estimated by boiling the stock solution with citric acid, which inverts the cane-sugar; the solution is cooled, neutralized with potassium hydroxide solution, made up to a known volume, and titrated with Fehling's solution. Many brands of condensed milk contain a very laige amount of sugar, the average being 38 to 40 per cent. ; while others are un- sweetened. When opened, the latter have inferior keeping powers. For Proteids. — Perform Kjeldahl's process on 10 c.c. of stock, and multiply the N by 6-38. The proteid matter averages 9 per cent. The " degree of condensation " may be approximately gauged by dividing the percentage of solids by 12 -6, when the condensed milk is unsweetened; or by dividing the percentage of fat by 3-6 in other cases — 12-6 per cent, and 3-6 per cent, being, respectively, the average amounts of total solids and fat in milk. Brands of condensed whole milk (not " machine skimmed ") ought to contain at least 10 per cent, of milk-fat and 25-5 per cent, of non-fatty milk solids, of which the ash should constitute about 2 per cent. Heated Milk. — Sometimes it is required to know whether milk has been steriHzed or boiled. In such a case 3 c.c. of milk may be mixed with i c.c. of a freshly prepared 10 per cent, solution of hydroquinone, and about 15 drops of hydrogen peroxide added. If the milk has not been raised to a high temperature, an imme- diate rose colour forms, but otherwise no colour is produced, as the reaction is destroyed by exposure to a high temperature. The following changes result when milk is boiled: Carbonic acid gas is expelled and the calcium and magnesium salts are therefore partially precipitated; the greater part of the phos- phates are also precipitated. There is a slight diminution in the organic phosphorus originally present ; a partial decomposi- tion of the proteins. The skin which forms on the smTace (solely when the heating is done in an open vessel) consists mainly of lactalbumin. This pellicle has approximately the foUo\nng composition: Fat, 45-5 per cent.; lactalbumin and casein, 51 per cent. ; mineral ash, 3-5 per cent. The normal emulsion of the fat 248 LABORATORY WORK globules is disturbed, so that the cream does not rise to form a layer on the surface, the lactose is partially burnt (carameliza- tion), and the milk therefore becomes slightly brownish in colour. The boiling destroys the ferments in the milk and probably also the antiscorbutic element of raw milk ; the natural germicidal power of fresh raw milk is lost, and almost all the bacteria are destroj^ed, those left consisting of sporing forms and certain highlj' resistant varieties. None of these changes takes place, appreciably, in "low tem- perature pasteurization " — namely, the heating of milk to a temperature of 60° C for thirty minutes — except the great reduction in micro-organisms. Milk Powders. — These are now very largely made, both from whole milk and from skimmed milk. The powders are cream- coloured, with a slight distinctive odour. The fat in whole- milk powder should amount to at least 25 per cent. This can be determined by the Gottlieb process or by the Soxhlet method. Otherwise the analysis follows on the general lines described. " Koumiss " consists of milk which has been skimmed of some of its cream and sugar added; it is then partially fermented by yeast or other ferments, wherebj^ much of the sugar is converted into lactic and carbonic acids. Bean or Synthetic Milk. — In the preparation of this milk soya beans are washed and soaked in water, the outer integuments being removed. The softened beans are then ground between millstones and the powder boiled with water and filtered through fine sieves. A cream-coloured liquid results, closely resembling milk, but with a distinct beany odour and taste. For use, sugar is added to suit the taste of the consumer. Bean milk on analysis contains about 2-i per cent, of fat, 37 per cent, proteins, 1-4 per cent, carbohydrates other than sugar, and 0-4 per cent, of mineral ash. Milk Standards. In addition to the legal standards at present in force, certain other standards are advocated. Only a small proportion (one- sixth to one-eighth) by weight of the cow-dung which finds its way into milk is recoverable (as dirt) from milk by centrifugaliza- tion; but despite the difficulties involved, certain standards for dirt in milk have been suggested. Most authorities agree that milk ;ynelding more than i part of recoverable dirt per 100,000 MILK STANDARDS 249 is dirty. Houston suggests that, as a working standard, (i) the deposit from a litre of milk obtained by sedimentation in a special cylindrical separating funnel after twenty-four hours should not exceed i part per 10,000 by volume; and (2) when the deposit from (i) is centrifugalized, it should not exceed half the above amount. It has been suggested that, as a general rule, the recover- able dirt should not be allowed to exceed 2 parts per 100,000 by weight, and some advocate a standard as low as i part. As a rough household standard, |- pint of milk placed in an ordinary tumbler should not throw a visible sediment in two hours. But dirt may be removed by trade filtration, which leaves behind the harmful bacteria, and therefore the only satisfactory standards are those based upon bacterial counts. Seasonal standards of total bacterial counts are serviceable. In Chicago, for instance, 1,000,000 bacteria per c.c. of milk from May I to September 30, and half that amount for the remainder of the year, is the standard of milk as it arrives in that city. Savage prefers a standard of lactose fermenters of the coli type of not more than 100 in winter and 1,000 in summer, and he suggests that initial contamination may be best judged from the number of Bacillus enteritidis sporogenes (the spores of which abound in cow-dung), as that organism shows relatively little tendency to multiply in milk. Certainly leucocytes exceeding 1,000 per c.c. along with manj- streptococci suggests the desirability of investigation. In special (certificated) milk, which is sold at an enhanced price, it is possible to impose such high standards aS' — freedom from B. Uiberculosis, a total bacterial count below 10,000 per c.c. (Class A) and 100,000 per c.c. (Class B), and delivery to the consumer at a temperature of not above 10° C A suggested standard for pasteurized milk is that the total bacterial count should not exceed 1,000,000 per c.c. prior to pasteurization, and 50,000 per c.c. when pasteurized and delivered to the consumer. Reductase test (Schmidt-MuUer) may serve as a standard for freshness. The test reagent is made by adding 5 c.c. of a satur- ated alcoholic solution of methylene blue (zinc chloride double salt) to 195 c.c. of distilled water; it should be boiled every da^^ before using. One c.c. of the reagent is mixed with 20 c.c. of milk, the surface is sealed with paraffin, and then the test-tube 250 LABORATORY WORK and its contents are placed in a water-bath at 45" C. to 50° C. Fresh milk should remain blue for twelve hours or more. The reduction of methylene blue by raw milk (in the absence of formalin) is due to bacterial contamination, and if the milk decolorizes within one hour the organisms certainly exceed 500,000 per c.c. Bacteriological Note. Milk as secreted is free from organisms, but even in the milk cistern of the udder and in the teat canals some bacterial infection takes place ; while at every stage, from the udder to the consumer, contamination with bacteria is possible, and under many of the conditions which now prevail is invited. Organisms gaining access to milk, unlike those in air and water, are usually in an environment most favourable to multiplication, and as a consequence milk as vended frequently contains one to five millions or more organisms per c.c. ; the number, as is to be expected, being considerably greater in hot weather. Park* has shown that in New York, " with only moderate cleanliness, such as can be employed by any farmer without adding appreciably to his expense, namely, clean pails, straining- cloths, cans or bottles, and hands ; a fairly clean place for milking, and a decent condition of the cow's udder and the adjacent belly; milk when first drawn will not average in hot weather over 30,000 and in cold weather not over 25,000 bacteria per c.c. Such milk, if cooled to and kept at 50° F., will not contain at the end of twenty-four hours over 100,000 bacteria per c.c. If kept at 40° F. the number of bacteria will not be over 100,000 after forty-eight hoiirs." The estimation of the number of bacteria in milk, or of some special group of bacteria such as the B. coli group, is the natural measure of the degree of contamination of milk, but the question of the numbers to allow is beset with difficulties, largely owing to the suitability of milk as a medium for the propagation of bacteria. A milk initially comparatively pure will frequently show after the lapse of twelve to twenty hours many more bacteria than one collected under much less cleanly conditions, and initially much more heavily charged with bacteria, but examined after the lapse of only three or four hours from milking. * Journal of Hygiene, 1901, vol. i., p. 391. BACTERIOLOGICAL NOTE 251 These differences in the bacterial content are not only deter- mined by the initial contamination and by the time since milking, but also by the temperature at which the milk has been kept, and the latter introduces a condition subject to great variation. The milk should be collected in sterile glass-stoppered bottles. Those used for the bacteriological examination of water may be used, or the simple and efficient apparatus described by Delcpine may be employed. It consists of a metal case containing a 7 or 8 ounce bottle and a milk-scoop. All the parts are thoroughly sterilized in the laboratory before being sent out, and the sterilized case is opened only at the time when the sample is taken. The sterilized scoop is used to remove the milk from the cans or other vessels. When obtained direct from a suspected cow, the milk may be milked FIG. 38. — delepine's milk-collecting apparatus. into the scoop. The metal cases are packed in refrigerating boxes if necessary. If the sample cannot be examined within an hour or so, it must be transmitted packed in ice. If milk from individual cows is being collected, the teats and the milkers' hands should be washed and disinfected. In some cases it is necessary to collect a separate sample from each quarter, while for a complete examination fore, middle, and end milk samples should each be collected. As a rule condensed milks are free from preservatives. In the sweetened milks the sugar is sufficient to inhibit the growth of bacteria, and in the unsweetened the milk has been sterilized at temperatures over 100° C. The processes carried out in con- densing the milk are sufficient to destroy Bacillus coli, B. tuber- 252 LABORATORY WORK culosis, and other pathogenic organisms, but spore-bearing bacilh, streptococci, sarcinae, yeasts, and other saprophytes, are often present, so that condensed milks must not be regarded as necessarily sterile. It is probable that the bulk of the organisms present have gained admission during the processes of cooling and of filling the tins. At the temperature of about lo^ C. the multiplication of lactic acid forming bacteria is checked, and those organisms are de- stroyed at a temperature of 70° C maintained for twenty minutes. The ferments hitherto detected in cow's milk (peroxidase, reduc- tase, catalase, etc), are mainly derived from bacteria; but certain of such ferments which are found to be present in uncontamin- ated milk, such as amylase, do not appear to be of value in diges- tion and nutrition. The chief enzymes are destroyed at about 70° C. in thirty minutes. At 80° C all enzymes are destroyed; but at 60° C. their activity is, if anything, slightly promoted. The discovery by Ehrlich that passive immunity could be produced by suckhng when the mother was immune, led to the investigation of the presence in milk of precipitins, agglutinins, opsonins, antitoxins, and other so-called " protective substances " in milk. These are all destroyed at about 60° C. There are also present in milk certain biological bodies (hor- mones and vitamines) produced by the direct action of living cells. As it appears that these biological bodies are not absorbed in the alimentary canal, it is not hkely that they act as antigens in the infant. The vitamine present is not destroyed at the temperature of boihng milk. CHAPTER IV BUTTER— CHEESE— LARD Butter. An average sample of fresh butter has the following com- position: Fat, 83-5 per cent. Curd (casein), i per cent. Ash, 1-5 per cent. Milk-sugar, i per cent. "Water, 13 per cent. The water may vary from 8 to 15 per cent. The butter-fat is a combination of glycerol with certain fatty acids ; and consists of — {a) The glycerides of certain volatile fatty acids, soluble in hot water— i.) parenchymatous cells (packed with starch grains) ; (/) testa; («) nucellus. >— £ parenchymatous cells of the endosperm, and to this end the whole of the complicated machinery of modern milling is con- WHEAT-FLOUR 271 trived. In practice, however, a perfect separation is never attained, and the flour always contains more or less of the other portions of the wheat grain, which are termed the "offal," con- sisting of the embryo or germ and the bran formed of the pericarp and the integuments and outermost layer of the seed. The in- clusion of this offal raises the mineral, fat, and proteid content. But the oil in the germ is very prone to develop rancidity, and it is principally for this reason that it is rejected in modern milling. " Standard " flour is described as " 80 per cent, of the wheat with all the germ and semolina," but this is unsatisfac- tory for more than one reason. In the first place, the term " semolina " does not connote any particular part of the grain; it is merely a trade name for the coarser fragments of endosperm produced in the break-roller system, and is therefore incapable of exact definition. In the second place, the requirement that the flour shall contain 80 per cent, of the wheat grain is by no means satisfactory as a " standard " of quality or composition. Wheats differ considerably from one another, and the skin or branny envelope bears a smaller ratio to the endosperm in the case of a large grain than of a small grain. The composition of ordinary baker's flour and of " whole- wheat " flour varies with different samples, but the following results would fairly represent the average: Ordinary Baker' " Whole- Wheat Flour. Flour. (100 parts by weight.) Albuminoids (N = = 6-25) . 12-3 13-8 Carbohydrates 71-2 68-1 Fat . . 1-3 1-9 Sugar . . 1-3 1-2 Fibre . . 0-4 1-7 Ash . . 0-7 1-4 (P2O5) . . .. (o-2i) (0-67) Moisture 12-8 II-9 lOO-O lOO-O The Analysis. Physical Characters of Flour. — The colour should be white, and the flour clean; a yellow hue denotes age or fermentation, and fermenting flour disarranges the digestive system, producing flatulence, dyspepsia, diarrhoea, etc. There should be no acid or mouldy odour, and no taste of acidity or mustiness. Taken up 272 LABORATORY WORK in the lingers the flour sliould be smooth and soft, with no lumpy or gritty feel; it should knit or bind together, and a little flecked on to the wall should mostly adhere; on mixing with a little water, the dough should draw out into stringy masses. There must be complete freedt)m from fungi and other parasitic growths. If flour is stored in a damp place, the number of microbes present increases rapidly, and poisonous alkaloidal products may result from prolonged storage under such con- ditions. As compared with other flours, wheat-flour is characterized by the large amount of crude gluten it contains ; and it is to this substance — or rather to one of its constituents termed " gliadin " — that the peculiar adhesiveness of the flour, which makes it so peculiarly adapted for bread-making, is due. If flour is made into a dough with water, and then the dough is thoroughly washed, it is this crude gluten which remains behind as a sticky mass, the starch and soluble substances {i.e., sugar, soluble albumin and salts) being washed away. It is of great value, therefore, both as a test of the purity and also of the quality of the flour, to estimate the amount of this substance. One means of effecting this is the following: Weigh out a quantity of flour — say 50 grammes — place it in' a small basin, and carefully mix it with lukewarm water (about 16° C.) into the condition of stiff dough ; then slowly and thoroughly work up the dough with the Angers, either under water, or while allowing a gentle stream of the warm water to fall upon it. As the dough becomes more and more washed, the water which is being constantly emptied away and renewed gets clearer and clearer, and the dough more stringy and sticky. Ultimately, the starch and all the soluble materials in the original flour are carried away, and the water escapes in a perfectly clear condition. Nothing then but crude gluten, containing generally a fraction over i per cent, of fats and salts, remains; and the entire absence of starch can be proved by treating with a little iodine. The gluten should then be spread out in a tared (weighed) flat-bottom dish, dried at 105° C, and finally weighed. A more exact method of estimating the proteid material is as follows: wheat-flour 273 The Estimation of Nitrogenous Organic Matter in Flour (Kjeldahl's Process). Special Reagents required. — (i) Decinormal sulphuric acid; (2) deci- normal soda solution; (3) methyl-orange indicator; (4) strong sodium hydrate solution (500 grammes added to 500 c.c. distilled water, and well boiled to free from ammonia) ; (5) strong sulphuric acid, free from nitrates and ammonium; (6) red oxide of mercury; (7) potassium sulphate. The Process, 1. Weigh from 0-5 to 3 grammes of the material (according to its richness in nitrogen), and transfer to a strong hard-glass boiling-flask. 2. Add 25 c.c. of strong sulphuric acid and 075 gramme of red mercuric oxide. Support the flask in a slanting position on a tripod, and by means of a bunsen burner keep the acid just below its boiling-point for half an hour. As fumes of sulphuric acid will escape from the mouth of the flask, the heating must be done in a fume-cupboard. 3. If at the end of half an hour the mixture is still black, 12 grammes of potassium sulphate (free from nitrates) are added. This raises the boiling-point, and the heating is continued for a few minutes after the liquid is clear and has no more than a faint yellow tint. 4. Let cool; add about 500 c.c. of ammonia-free water, and be careful to wash thoroughly the mouth and neck of the bottle with this water. Then more than neutralize the acid by means of the strong soda solution. Also add about 20 c.c. of a 4 per cent, solution potassium sulphide, in order to precipitate all the mercury as sulphide, and thus prevent the formation of mercur- ammonium compounds. 5. Distil over the ammonia into 50 c.c. of decinormal sul- phuric acid, using a condenser with a bulb (" anti-splasher") in the condensing-tube a little above the boiling-flask, in order to guard against the liquid spurting over. 6. When about 250 c.c. of distillate have been collected, the acidity is titrated with the decinormal soda solution, using methyl-orange as indicator. The difference between the amount of soda solution required to neutralize the 50 c.c. of decinormal acid, before and after the addition of the distillate, represents the ammonia which has come over. Each c.c. of the acid 18 274 LABORATORY WORK neutralized by the ammonia =0-0014 gramme of nitrogen, and the nitrogen multiphed by the factor 5-68 (as the protein of wheat contains an average of 17-6 per cent, of nitrogen) repre- sents the amount of albuminoid or proteid mateiial in the amount of the flour examined. Notes on the Process. — 'i he organic matter is burnt up in this process by moist combustion, and the resulting ammonia com- bines with the sulphuric acid to form ammonium sulphate. The addition of excess of soda liberates the ammonia from the acid and enables it to be distilled over. It is necessary to perform a blank experiment occasionally in order to test the reagents. The figure of the blank experiment (commonly only about 0'2 c.c. of decinormal alkali) should, of course, be deducted in arriving at the total nitrogen obtained from the material. Example. — Two grammes of flour furnished ammonia which neutralized 19-5 c.c. of the decinormal acid. .-. there are 19-5 x 0-0014= 0-0273 gramme of nitrogen in 2 grammes of flour= 1-365 grammes of nitrogen in 100 grammes of flour. .". 1-365 X 5-68 = 7-75 per cent, of proteid material, or approxi- mately 7-7 per cent, of gluten. The gluten varies from 8 to 12 per cent.; if the gluten is less than 8 per cent, the flour is not pure wheat-flour; and if it cannot be drawn out into long fine threads without breaking it is poor in quality. Rye yields a plastic gluten which cannot be separated by washing. The water of flour should not exceed 18 per cent, by weight, since more than this, besides fraudulently throwing up the weight, impairs its keeping power by favouring the development of fungi and the acetic and lactic acid fermentations, which may some- times produce gastro-intestinal disturbance. The amount of moisture is, of course, ascertained by drying a weighed quantity of flour over the water-bath (and subsequently in the hot-air oven), the loss in weight being due to moisture. The ash of wheat consists chiefly of phosphates of potassium, magnesium and calcium, and small quantities only of salts of silica, sodium, iron, etc. ; the amount should not much exceed I per cent., and as much as 2 per cent, would imply that mineral adulterants have been added. In making the estimation, cautiously incinerate a weighed quantity of dried flour in a platinum dish, until a clean white ash remains. During ignition WHEAT-FLOUR 275 a hard mass of carbon forms, and it is a good plan to moisten this with a strong solution of nitrate of ammonia, then dry and continue the ignition. It may be necessary to repeat this treat- ment before a clean ash is obtained. Adulteration. — Foreign mineral matter, which is seldom now added, may be roughly estimated by shaking up with chloro- form, when the flour floats and most of the added mineral matter settles at the bottom of the vessel. The treatment is repeated, in order that it shall be as inclusive as possible, and the sediment collected, dried and weighed; it can also be examined as to its nature. The presence of any added mineral matter (calcium phosphate, sulphate and carbonate, etc.) is also readily detected in the ash, as this is found to be exceptionally high. Phosphates and other " improvers " are employed to improve the baking quality of certain flours by improving the quality of the gluten, and thus increase the strength and water-absorbing capacity of the flour. A mixture in about equal proportions of acid potassium and magnesium phosphates and flour is said to be an effective im- prover. The use of calcium acid phosphate has greatly extended in recent years, and as this substance may contain a large pro- portion of calcium sulphate, which is valueless. Dr. Hamill, in a Report to the Local Government Board (1911), recommends that a maximum limit of 10 per cent, of calcium sulphate in calcium acid phosphate should be fixed. There is no evidence that these phosphatic improvers increase the organic phosphorous compounds present ; they are added with the object of increasing the amount of bread which can be obtained per sack of flour. Rarely it is by the addition of other flours and meals that sophistication is practised, when, of course, the cheaper varieties are selected, such as maize. To detect rice-starch in wheat 33'33 grammes of flour are made into a ball with 17 grammes of water, and worked between the fingers in a fine stream of water over a fine-meshed sieve. The starch and waste water (thus separated) are well shaken, and set aside for twelve hours in a large conical flask, when the starch separates in three well- marked layers which can be separated by decantation. The top layer contains most of the small starch granules, and the 27t) LABORATORY WORK bottom layer the largest grains, whereas the middle layer is mainly composed of the cellulose and proteid element of the flour. . When rice-starch is present, it is almost entirely deposited in this layer, and its presence can be detected in so small a pro- portion as I per cent. (E. Collin). Maize is difficult to detect; but if the flour is mixed with clove oil the hilum of maize appears under the microscope as a black star or spot. This is not the case with the other starches most commonly used to adulterate flour. Old flour is occasionally passed through the; mill with fresh flour; in these cases there is marked acidity, a reduction in the fat and the quality of the gluten. The degree of whiteness of flour depends upon the fineness of the grade, and to some extent upon the variet}' of the wheat; and the artificial means of producing a white flour is by means of bleaching with nitrogen peroxide gas (NOg). The bleaching effect appears to be due to the destruction of a yellow colouring matter dissoh-ed in a thin layer of oil which surrounds each granule of starch. Dr. Hamill, reporting upon this matter to the Local Government Board, states that the practice cannot be regarded as free from risk to the consumer, especially when regard is had to the inhibitory effect of the bleaching agent on digestive processes. This bleaching of flour is prohibited in Switzerland, United States, and certain of the Australian States. The amount of nitrite left in the flour is extremeh- minute, and in the bread made from the flour it is still further reduced; and for practical purposes it may be presumed that when over 1-5 parts of nitrites per million are present, the flour has been bleached. Small amounts of nitrate are also formed as a result of bleaching. Extremely small quantities of nitrites can be tested by the Ilosvay method, by which the colour produced with sulphanilic and a-naphthylamine hydrochloride in acetic acid solution may be compared with standards containing known amounts of sodium nitrite. Dr. MacFadden, in a Report to the Local Government Board, draws attention to the fact that the relation which may exist between apparently very minute alterations in the nature of staple food materials and the production of great and far- reaching changes in nutrition has been strikingly demonstrated in recent investigations into certain obscure disorders of metabol- ism, of which the disease known as beri-beri may be taken as an WHEAT-FLOUR 277 example, and the time has arrived for taking a wider view than has hitherto been customary of the danger to healtli wiiich may arise from the sophistication of foodstuffs. The bleaching of flour by chemical oxidizing agents has been introduced in response to the public demand for a white loaf. A simple test for bleached flour is to shake up about ^ ounce of the flour with 2 fluid ounces of petrol. If unbleached, the spirit takes up a yellow colour, but not so if the flour has been bleached. The bleaching of flour by means of nitrogen peroxide renders the gluten indigestible (Hafliburton). The seeds of the darnel grass, or Lolium temulentum, may gain access to wheat or oat flour; they are said to possess narcotic poisoning properties. Neither the starch grains nor the testa are characteristic under the micioscope, since both resemble oats very closely; but the addition of alcohol causes a greenish colour to appear, together with a pecuhar repulsive taste, if flour contains these seeds. The corn-cockle {Agrostemma githago) consists of large, dull black seeds, showing small protuberances. They are markedly poisonous. CHAPTER VI BREAD It is only with wlieaten bread that the following chapter deals. Among the means for obtaining the porosity of bread is in- cluded the use of baking-powders. These consist most generally of a mixture of sodium bicarbonate, tartaric acid and rice-flour. The rice-flour is used to keep the powder dry, and to prevent chemical action setting up until it is moistened. The tartaric acid baking-powders are by far the most common, but powders are also sold in which the acid constituent is fur- nished by acid phosphate, and in other cases by the sulphuric acid contained in some form of alum salt. It has been argued that the employment of acid phosphates is of value as replacing the phosphates lost to the bread by the removal of the bran. Certainly the use of baking-powders containing alum should be condemned. The reaction between potassium-alum and sodium bicarbonate has been shown to result in the production of alu- minium hydrate, and the hydrate of alumina is known to be dis- solved (with difficulty) by the gastro-intestinal juices. Such a baking-powder, analyzed by the writer, gave 23 per cent, sodium bicarbonate, 33 per cent, alum, and 44 per cent, ground rice, etc. Self-raising flour is flour containing the essential elements of baking-powder. In the process of cooking, some of the starch of the wheat- flour is converted into maltose. The composition of good bread (freed from moisture) is approxi- mately as follows: Starch, dextrin, etc. . . . . . . 82 "6 Nitrogenous matter Maltose Fat Salts II -4 40 0-6 i;4 loo-o 278 BREAD 279 Physical Characters.— The bread should be fairly dry, light and spongy; and not sodden, acid, or musty. It should be clean and of a good colour — nearly white, that is to say; for a yellow or dirty colour betrays age and poorness in quality. A peculiar violet tint is given to wheat containing melampyrum and other species of Scrophulariacea and trefolium (trefoil) . Other growths sometimes give the bread a dirty blue appearance {rhinanthus, etc.) ; agrosiemma (corn-cockle) furnishes a greenish tint. Oidium aurantiacum has caused poisoning in France. It is a reddish- yellow mould, giving a bitter taste and offensive odour to the bread. The estimation of water and mineral matter is performed as in flour. Fifty grammes of crumbs is a convenient amount to work with. The moisture in bread should not much exceed 40 per cent, in the crumb part and 25 per cent, in the crust. The Ash.— An increase in weight of the ash of bread over that of the original flour is due to the common salt and the baking- powder which are added in the process of baking. But any excess of ash above 3 per cent, would be due to added minerals (such as gypsum and chalk), which have been rarely added with the object of improving the colour. As chlorides and other salts may be volatilized by the pro- longed ignition necessary to furnish a white ash, the ash must be procured at as low a temperature as possible. To estimate the silica, treat the ash first with strong hydro- chloric acid, then with a little distilled water and boil; next filter through a Swedish filter-paper, wash the platinum dish by boiling more distilled water in it, and filter these washings also through the same paper. When the platinum dish is perfectly clean, well wash the material upon the filter-paper with small quantities of hot distilled water; dry in the water-oven, and then ignite in a porcelain crucible with lid; finally weigh the ash; deduct the weight of the filter-paper ash, and the difference is silica. It should not exceed 0-2 per cent. To estimate acidity soak 10 grammes of bread in about 50 c.c. of water for one hour, filter, and titrate the filtrate with a deci- normal alkaline solution, using phenolphthalein as indicator. The number of c.c. of decinormal soda used to neutralize the acidity x 6 = milligrammes of glacial acetic acid in 10 grammes of bread. Express results in terms of glacial acetic acid per cent. 200 LABORATORY WORK ^ Anything over o-i2 per cent, is rather acid, and tliis figure should, therefore, not be exceeded. Adulteration. — Mashed potatoes are looked upon as a legitimate addition in slight amount where sponginess is dependent upon fermentation, since they favour this action. It has been said that they are added to increase the weight and whiten the loaf, and since they contain between 70 and 80 per cent, of moisture, they help to keep the bread moist; but generally only the strained liquor in which the potatoes have been cooked and mashed is employed, in order to obtain a sweeter loaf. Rice when added also serves the purpose of giving a good white colour to the loaf. Dr. Alford has recorded an outbreak of lead-poisoning, affect- ing from fifteen to twenty persons, arising from the consumption of flour which had been ground by an old mill-stone in which large spaces had been filled in with lead. The starch grains of the flour used in the manufacture of bread become so altered by the process of cooking (on account of the rupture of their envelopes) as to lose most, if not all, of their microscopic characteristics. Fungi may be discovered, and notably the different forms of penioilliwn ("mildew"); these may create, if sufficiently numerous, patches of greenish, brownish, or reddish discolora- tion. Oiduim auranfiacum furnishes an orange hue. These fungi should condemn the bread off-hand, for they may give rise to considerable gastro-intestinal disturbance. FIG. 54. PENICILLIUM GLAUCUM. (x ABOUT 200.) Alumina exists normally in pure flour as the silicate of alumina; but it has been added, as alum, to inferior flour, so as to check the fermentative action whereby a large amount of sugar (glucose) is formed and a discoloured bread, unpleasant to the palate, results. Alum thus improves the taste and colour of the bread, and also to some extent its porosity. BREAD 281 The whitening of flour has also been obtained by bleacliing methods, so that the eolour is no certain indication of quality. At the present day, largely owing to the adoption of other expedients, alum is very little employed, and it is rare that alumina is detected in amounts which denote an excess over that which may be normally present. Large quantities of the salt were formerly added to flour, and the great decline in its use commenced with the passing of the Sale of Food and Drugs Act, 1875. It was generally employed in quantities of about 15 to 35 grains to a 4-pound loaf, but more than 100 grains have been separated. It is generally held in this country that no addition of alum should be countenanced, and that very small amounts in an article such as bread, of which large quantities are consumed, may prove deleterious to health, by inducing dyspepsia, constipa- tion, etc. There is, however, some conflict of opinion as to whether small quantities of alum are injurious to health; but there can be no doubt that it is an adulteration under the Bread Acts of 1822 and 1836. The best test for the presence of alum, and one which will detect as little as i grain per pound in bread which has not undergone acid fermentation, is as follows: Reagents required. — i. A strong freshly-made tincture of logwood, prepared by digesting 5 grammes of freshly-cut logwood chips in 100 c.c. of strong alcohol. 2. A solution of ammonium carbonate (15 grammes of ammonium carbonate in 100 c.c. of distilled water). About 5 c.c. of each of these reagents are added to about 30 c.c. of water, and pieces of the crumb of the bread are cut from the loaf, moistened with a little water and left to soak in this mixture for a few minutes ; the fluid is then drained off and the bread gently dried over the water-bath. The presence of alum is denoted by the appearance of a permanent lavender or violet colour, according to the amount present; while the parts of the bread which contain no alum are first stained the bright colour of the logwood solution, and afterwards change to a dirty brown tint. Wynter Blyth soaks the bread paste in gelatine, and then tests this with logwood and ammonium carbonate; a neater reaction is thereby obtained. The operator must be careful that he is not led astray b^^ magnesium salts, which are capable of creating a lavender tinge 282 LABORATORY WORK almost identical with that of alum; but the colour created by these salts is certainly not so permanent upon drying as that furnished by alum. In order to avoid the effect of acidity of old meal or of sour bread on the logwood test, the following method is recommended: From 10 to 20 grammes of the bread are triturated into a paste with water, some sodium chloride (fiee from alkali) added, and then, after the addition of 10 drops of freshly prepared logwood tincture, 5 grammes of pure potassium carbonate are gradually mixed in. After being well mixed, the whole is washed with 100 c.c. of water in a beaker and allowed to settle. In a few minutes the supernatant liquid becomes a greyish to a deep blue when alum is present, and a reddish-violet tint when it is absent. As a confirmatory test (Herz) for alum in flour, 10 grammes of the flour are mixed with water and allowed to stand for ten minutes; filter, concentrate, and precipitate the proteids with tannic acid solution. Filter and add 2 drops of tincture of cochineal. In the presence of alum a carmine red colour is obtained. One of the best methods of separating and estimating the alumina in bread is the following: Quantitative Estimation of Alum (Dupre and Wanklyn). (i) Incinerate |- pound of flour or bread to a grey or reddish ash. (2) Separate silica, etc., by treating with strong HCl and then boiling water; filter; wash filter with boiling w'ater. The filtrate contains phosphates of calcium, magnesium, iron and aluminium. (3) Add 5 c.c. of ammonia solution (when all the phosphates are precipitated) ; then 20 c.c. of strong acetic acid are added gradually (which redissolves the phosphates of calcium and magnesium); filter; wash filter with boiling w-ater, dry, ignite, and w^eigh. The residue contains the iron and aluminium phosphates and the filter-paper ash. (4) Dissolve up the residue in strong HCl and dilute to 200 c.c, and then estimate the iron colorimetrically on the lines indicated in Water Analysis. BREAD 283 {5) Convert the Fe thus estimated into ferric phosphate l)y multiplying by 2-7, and deduct this amount from tlic total weiglit found in Step (3) ; further deduct the weight of filter ash, and the difference is aluminium phosphate. Convert this into crystallized ammonium alum (the commercial " alum ") by multiplying by 37, and calculate to grains per pound. Conclusions to be Drawn from the Amount Estimated. — If the amount of alumina represents more than from 6 to 10 grains of alum per 4-pound loaf, in the vast majority of cases the latter has been fraudulently added; some pure flours, however, may undoubtedly contain a greater quantity than this, and hence it is difficult to lay down any definite quantity as a standard beyond which the proof of fraudulent addition may be certainly established. The alumina which is taken up from the soil is in the form of silicate, and if the amount of alumina considerably predominates over that of the silica, that circumstance would denote the presence of " added alum.." CHAPTER VII THE AVERAGE COMPOSITION OF OTHER FLOURS AND MEALS* —THE MICROSCOPIC CHARACTERS OF THE DIFFERENT STARCH GRANULES Oats. Arrowroc )T. Starch, dextrin, and ccllu- Starch, dextrin, and cellu- lose 64-5 lose 83-0 Nitrogenous matter I2-0 Nitrogenous matter 0-8 Fat 6-u Mineral ash . . 0'2 Mineral ash . . 3.0 Water i6'0 Sugar 2'0 Water 12-5 lOO'O Tapioca lOO'O Sago. Starch, dextrin, and cellu- Starch, dextrin, and cellu- lose S7-3 lose S6-0 Nitrogenous matter 0-6 Nitrogenous matter 0-8 Mineral ash . . O'l Mineral ash . . O'l I3-I Water Lentils I2-0 Water 1000 lOO'O CORNFLOURf (M Starch, dextrin, and lose Nitrogenous matter Fat Mineral ash . . Water aize). cellu- 68-5 13-0 3-5 1-5 13-5 Starch, dextrin, and lose Nitrogenous matter Fat Mineral ash . . Water cellu- 58-5 25*0 2'0 2-5 I2'0 IQO'O lOO'O Pea. Bean (Haricot). Starch, dextrin, and cellu- Starch, dextrin, and cellu- lose 58-5 lose 57-5 Nitrogenous matter 23-0 Nitrogenous matter 23-5 Fat 2*0 Fat 2-0 Mineral ash . . 2-5 Mineral ash . . 3.0 Water 14-0 Water 14*0 lOO'O lOO'O * Chietiy compiled from the results of analyses made by the writer. t Cornflour consists of the nearly pure starch of maize or rice. 284 THE E AVERAGE COMPOSITION OF FLOURS AND MEALS 285 Rye. Starch, dextrin, and cellu lose Nitrogenous matter Fat Mineral ash . . Sugar Water Potato. Starch, dextrin, and cellu lose Nitrogenous matter Fat Mineral ash . . Water 68-0 II'O 2*o 1-5 3-5 I4'0 lOO'O 22-0 2-0 O'l I-O 74-9 lOO'O Barley. Starch, dextrin, and cellu- lose Nitrogenous matter Fat Mineral ash . . Water yi'O "•5 1-5 I'O 15-0 lOO'O Rice. Starch, dextrin, and cellu- lose . . - . • ■ 7'^'5 Nitrogenous matter . . 6-5 Fat 0-5 Mineral ash . . . . • • 0-5 Water 14-0 A silicate of magnesia (talc) is sometimes employed to polish or " face " rice, in order to improve its appearance; oil may be em- ployed to increase translucency ; and blue pigments (ultramarine) to improve the white colour. The more expensive of these flours, or meals, are liable to be adulterated with the cheaper kinds, such as rice, tapioca, potato, and maize. It will be seen that, in comparison with wheat, barley is poor in nitrogenous matter and sugar, but rich in cellulose and mineral matter; that oats are exceptionally rich in cellulose and fat, possess a high amount of mineral matter, but are relatively poor in starch; that maize possesses a high amount of fat, but the cellulose is low; that rye is exceptionally rich in sugar, and in other respects closely approximates to wheat; and that rice is rich in starch, but poor in everything else. The amount of starch in any substance is estimated as follows : Five grammes of the dried and powdered material are mixed with 200 c.c. of 4 per cent. HCl, a reflux condenser is attached to the flask, and the liquid is boiled for five hours. The contents are then cooled, made slightly alkaline with sodic hydrate, and the dextrose estimated by Fehling's method. The dextrose xo-g^ starch. If cellulose is present, a little of this would also be con- verted into sugar by boiling with the acid, but this small quantity is often ignored. 286 LABORATORY WORK The Microscopic Characters of the Different Starch Granules. The starch, of which the foregoing foodstuffs are mainly com- posed, exists in the form of microscopic granules, which are more or less characteristic of the particular plant from which they are derived, on account of their difference in size, shape and mark- ings. These microscopic granules consist of an extremely thin envelope of cellulose enclosing the starch (granulose), and the latter appears to be generally arranged in fine superimposed strata — -which accounts for the " stride," or concentric lines, commonly discernible upon the external surface of the granule. When a sample of any flour or meal is to be examined under the microscope, very small amoimts are placed upon several clean glass slides, a drop of water is applied to each slide and a clean cover-glass is pressed firmly down over the powder and water to evenly distribute the powder. It is impossible to get too thin a layer of the substance in order that a satisfactor\' examination may be made, as otherwise granules get super- imposed and conglomerated, and their contours and markings cannot be defined. It is a good plan, therefore, to drop a small amount of the powder upon the slide, and then to gently blow it almost all away again, before applying the water and cover- glass. It is important that the reader should recognize that in the description which follows the most characteristic features are described. It must not be thought that in a sample of arrow- root, for instance, each granule will possess the characters described under that head. Such is by no means the case, for some may have the hilum in the centre, or even at the small extremity of the granule (as in potato), and yet the sample may be pure; but many of the granules will possess, in a more or less marked degree, the characters described. Where, therefore, the starch grains of different food-plants somewhat closely resemble each other it is difficult to decide as to whether there may be some slight admixture, although considerable adultera- tion admits of no questioning; but when these grains are dis- similar in appearance, the faintest possible amount of admixture is readily detected. When it is required to estimate the amount of adulteration, a rough percentage of the foreign starch grains present may be made by counting them upon the microscopic PLATE V. ARROWROOT PEA BEAN STARCH GRANULES. (X 250.) It. C. Boiis_field, photo. PLATE VI \ 1, ■• -= ) ^ ■;•' .'--' - ■■ ' ' _) M £. - ■'■; ,,/-.? ■ • yv- ■. _ ■-> - ^^ -O ^ MAIZE TAPIOCA STARCH GRANULES. (X 250.) E. C. Bousjield, photo. MICROSCOPIC CHARACTERS OF FLOURS AND MEALS 287 " field " of several mounted specimens. When the percentage amount of foreign starch has been estimated, a careful and thorough mixture is made up containing the supposed amounts of the ingredients in the composition under examination; this is then examined under the microscope and the counts compared with those of the original powder, in order to see if the estima- tion which has been made is broadly correct. If not, known quantities of the pure substance are mixed with fresh quantities of the adulterants found until a microscopic examination shows that the approximately true percentages have been arrived at. In cases where the foreign starch granules are very distinctive, the number in the specimen may be counted upon a plan very similar to that adopted in the case of counts of blood-corpuscles. It will be seen that, in many cases, the differences between the starch granules are very slight, and therefore some skiU is requisite in detecting them; such skill is only acquired from practice, and the student is recommended to fit up a small case containing samples of all the more common starches and to practise assiduously with these. Specimens mounted in glycerine are well preserved for a short time. A J-inch power should be employed, and this suffices for all practical pui poses. Mention may be made of the useful adjunct which the polari- scope may furnish to such investigations. For polariscopic examination glycerine or oil should be used instead of water. Starches, such as potato, arrowroot, bean, and maize, polarize well; while wheat, rice and oatmeal polarize feebly. I. Large round or oval granules, more or less flattened, and showing no marked concentric " strice " {or at most only a feio at the margins), together with other granules extremely small and ill- defined. May he wheat, barley, or rye. Wheat. — Relatively few " intermediary "* sizes, although the larger granules themselves vary somewhat in size. (A linear hilum and strise are visible under a very high power, and the small granules are seen to be angular.) Barley. — Similar; but the large granules are rather more irregular in shape and somewhat smaller, and " intermediary " sizes are more commonly present; lumpy forms rather more common. * A term used, in this connection, to denote a size about midway hetween that of the large and small granules. 288 LABORATORY WORK Rye. — Similar; but many show a rayed hilum, and present cracked edges; the granules are somewhat larger, and more generally circular and flattened than those of wheat or barley. Striations often distinct. Rye-flour is darker and less finely ground than wheat-flour. 2. Large pyriform or oval granules, with well-marked concentric stricB and a circular or short linear hilum. May he potato or arroicroot. Potato. — Typically, a well-marked circular or stellate hilum is at the smaller ex<-remity, and the striae aie well marked. The granules vary considerably in size. Arrowroot. — Similar, but the hilum is geneially at the larger extremity, and the granules average a trifle smaller (with the exception of the arrowroot named " tous-les-mois," in which commonly the gianules are even larger than those of potato, though they vary considerably in size) . The granules do not swell with potassic hydrate solution, as do those of potato, and the FIG. 55. BRUCHUS PISI (OF THE PEA, BEAN, ETC.). (X ABOUT 4O.) concentric rings are, generally speaking, less visible. There are many varieties of arrowroot, all of which present similar general characteristics as to their starch granules: the common variety is derived from Maranta arundinacea. 3. Oval or reniform granules, with faint concentric stricB, a central linear hilum, and very uniform in size. May be pea or bean. Pea. — Most have a central longitudinal hilum, which presents a puckered appearance. The granules are large. Bean. — Similar; but somewhat larger and more flattened {i.e., broader), and slightly more uniform in size. The hilum is much more commonly crossed by transverse hues (" puckered "). 4. Very small angular and faceted* granules, without concentric stria. May be rice, oatmeal, or maize. jlice. — The minute granules tend to collect into angular masses. * These facets are due to the close juxtaposition of the granules. MICROSCOPIC CHARACTERS OF FLOURS AND MEALS 289 Oatmeal. — The granules tend to collect into rounded masses, and are slightly larger than in rice, but still very minute. Maize. — The granules are much larger and are more irregular in shape, which tends towards the circular; they possess a visible hilum which is generally stellate. 5. Irregular in size, rounded, or partly angular with rounded edges, possessing (generally) a central hilum, and occasionally showing ill-defined concentric stricB. May be sago or tapioca. Sago. — Mostly large and irregular in shape; many elongated, with one larger end rounded and the other truncated. Hilum stellate or linear. Tapioca.- — -Similar; but much smaller, and many granules have a tendency to be truncated by one facet. Hilum generally more towards the rounded extremity. FIG. 56. SECTION OF WHEAT GRAIN (oUTER COAT) . (X50.) a, Girdle cells; b, cerealin cells. In order to get rid of starch, oleo-resin, etc., and thus bring together within a small compass much of the vessels, fibres, and parenchyma, the following steps are serviceable : Five grammes of powdered material are mixed with 50 c.c. of water to which 2 c.c. of HCl (S.G. i-i6) are added. The mixture is boiled for ten minutes, and then centrifugalized ; the solid matter is washed, partially dried, stirred with a few c.c.'s of chloral hydrate solution ; and the mixture is again centrifugalized and the deposit examined under the microscope. The flours and meals from cereals also give evidence under the microscope of the thin envelope of the grain, called the skin or testa ; and this is the case with even the finest ground and purest flours. 19 290 LABORATORY WORK In wheat the envelope is composed of three* fine membranes, the external and the middle both consisting of flattened cells, which are more or less dovetailed into each other. The long axes of the cells in the middle coat are disposed at right angles to those in the external, the latter being arranged with their long axes corresponding with that of the grain. Unicellular cells with pointed apices (" hairs ") come off in tufts from the external coat at one extremity of the grain; these " hairs " are simplj^ prolongations of the cells. The internal coat is made up of irregularly rounded, opaque- looking cells, which frequentl}^ contain one or more oil globules. The starch gi-anulcs, comprising almost the whole of the interior of the grain, are included within a thick-walled cellular network. FIG. 57. WHEAT. TISSUE FROM THE " TESTA " OF THE GRAIN, SHOWING THE APPEARANCE OF THE CELLS FORMING ITS OUTER AND INNER MEMBRANES. (X lOO.) .^m^ FIG. 58. — BARLEY. TISSUE FROM THE " TESTA " OF THE GRAIN, SHOWING THE APPEARANCE OF THE CELLS FORMING ITS OUTER AND INNER MEMBRANES, (x TOO.) In barley the envelopes are the same as those in wheat, except in the following respects : The cells forming the external coat are shorter and more uniform in size than in wheat, and their outline is serrated instead of beaded; they carry, moreover, short thick hairs. The cells of the middle coat are more elongated, and those of the inner coat are somewhat smaller. Slight as these differences are, it is to the envelopes rather than to the starch granules that one must turn in order to dis- criminate between wheat and barley. In rye the testa so closely resembles that of wheat that it is difficult to hit upon a point in which they differ, and it is for- tunate that the starch grains afford a ready means of distinguish- ing between the two. * There are probably six in all under very high powers. MICROSCOPIC CHARACTERS OF FLOURS AND MEALS 2(Jl It maybe pointed out that the unicellular hairs are somewhat shorter than in wheat. In maize (Indian corn) the envelopes arc two in number; the external consists of several superimposed layers of flattened, elongated cells, and the internal of a layer of cells of irregular size and shape, but otherwise resembling the internal layer of wheat. A very characteristic circumstance about maize is that the cellular network which holds the starch granules in this plant forms an irregular mosaic, most often pentagonal but occasionally polygonal in design. Spoilt maize taken as food may be responsible for pellagra, which is probably a food intoxication induced by some toxi- cogenic saprophyte. Many varieties of parasite are found on maize, including certain moulds, the spores of which are not destroyed by cooking. FIG. 59. RYE. TISSUE FROM THE " TESTA " OF THE GRAIN, SHOW- ING THE APPEARANCE OF THE CELLS WHICH FORM ITS OUTER AND INNER MEMBRANES, (x lOO.) FIG. 60. OATS. TISSUE FROM THE " TESTA " OF THE GRAIN, SHOW- ING THE APPEARANCE OF THE CELLS WHICH FORM ITS OUTER AND INNER MEMBRANES, (x 100.) In oats the envelopes consist of an external one of long narrow cells with evenly serrated contours (not wavy or beaded), and carr3nng sharp spinous "hairs"; a middle, somewhat similar coat, but indistinct and poorly seen; and an inner layer of cells resembling the internal one of wheat, but larger. In rice the external coat of the husk, consisting of long narrow cells, is characterized by the number of fine silicious particles it contains, which are collected together into ridges crossing each other at right angles. The spinous hairs are long and numerous, and the other coats, of which there are several, consist also of elongated, narrow, flattened cells, variously arranged. The " polishings " from white rice appear to contain a sub- stance the absence of which may lead to beri-beri when white rice is the staple food. 292 LABORATORY WORK The polishing of rice results in the removal of the pericarp and of the whole of the greater part of the sub-pericarpal layers of the rice grain. C. Funk has pointed out that a satisfactory measure of the degree of polishing to which rice has been sub- jected is the estimation of its total phosphorus, and that a rice which jdelds less than 0-4 per cent, of P2O5 cannot safely be permitted to form the staple diet of man. CHAPTER VIII MEAT— PARASITES OF FLESH— POISONING BY FOOD- MEAT PREPARATIONS It is sometimes necessary to make a laboratory examination of meat, in order to decide whether it is fit for human consumption. The Characters of Good Meat, It should have a marbled appearance, due to little streaks of fat between the muscular fasciculi; the whole surface should have a glossy appearance, and the colour should be of a bright florid hue and not too dark, or the meat is that of an old or diseased animal. The colour of veal, mutton and pork is always paler than that of beef, and this fact depends to some extent upon natural causes (the flesh of all young animals is naturally paler than that of older ones), but mostly upon the fact that calves, sheep and pigs are bled more at the time of killing. In old animals the flesh is darker and tougher, and the fat more yellow and soft. The connective tissues should glisten when exposed, and the muscular fasciculi should not be too large and coarse. To the touch the meat should be firm and slightl}^ elastic, which implies that the meat is fresh and has set well (rigor mortis) ; it should, moreover, be so dry upon the surface that the finger is only slightly moistened by being passed over it; such moisture should be of a clear red colour and of an acid reaction. In taking the reaction the litmus-paper should first be dipped in water, as otherwise the serum glazes the paper and obscures the reaction. On cutting through the flesh, the whole thickness should present a uniform colour, or the interior must be but very slightly paler than the more external flesh. The odour of meat is best obtained either by drenching it 293 294 LABORATORY WORK (when finely minced) with very hot water, or by phmging a clean odourless knife or new wooden skewer deep down into its sub- stance — prefcrabl}^ in the direction of bone — and then with- dra\ving and smelling the knife. The peculiar odour of good fresh meat is familiar to all, both in the raw and cooked state, and any departure from this would create suspicion. The fat should have a firm and greasy feel; the normal faint yellow colour must not be excessive, although the fat of animals fed upon some oil-cakes acquires a very marked yellow hue. The fat deepens in colour with age. It should present no hemorrhagic points. Any lymphatic glands attached should be firm, smooth, slightly moist, and of a pale, greyish-brown appearance on section. The pleura and peritoneum should be smooth, glistening and transparent. The marrow of the bones should be light red; that from the bones of the hind-quarters sets firmly within twenty-four hours, but that from the fore-quarters remains diffluent for a longer period. The ash of the meat is alkaline, and consists almost entirely of phosphates and chlorides. The Characters of Bad Meat. Bad flesh is frequently moist, sodden, flabby and dropsical, and may be infected with parasites. It must be remembered, however, that the flesh of young animals is always pale and moist. Some parts of the meat may feel softer than others — that is to say, there is not a uniform resistance to pressure, and occasionally there may be emphysematous crackling. The flesh of \eal and lamb may be blown out artificially and the surface then smeared with fat, and thus an artificial plumpness is given to poor meat. Dishonest butchers may also rub melted fat over the flesh of diseased animals to give it a healthy and glossy appearance. The fat is generally soft and flabby, or gelatinous ; frequently highly coloured, or exhibiting small hasmorrhagic points. Any attached lymphatic glands may be enlarged, softened, hypencmic, ecchymosed, caseated, calcified or suppurated. The marrow of the bones is discoloured (brownish) and sets badly. MEAT 295 A deep purple or dark tint suggests that the ;i,ninial has not been killed and bled, but has died with the blood in it, and probably of some acute feverish condition or pulmonary com- plaint. A yellow or mahogany hue denotes bile-stained flesh. Should the meat be very pale (" white flesh "), and the animal an adult one, fatty infiltration or degeneration, or fibroid infiltra- tion, may be the cause. A magenta hue of the flesh points to some acute specific condition being present at the time of death. Well-defined and dark-coloured areas full of blood are due to hypostatic congestion or post-mortem staining. Pus may be seen lying between the muscle fibres, and boils or small abscesses may be present (as in anthrax, etc.). There is frequently too great a proportion of bone to flesh, the animal having been greatly emaciated. The reaction of the juice (which may be dark or discoloured) may be alkaline or neutral. The odour may be that of putrefaction or of a faint and sickly nature. The pleura and peritoneum may be wet or roughened, opaque, congested, or blood-stained. Sometimes there is an odour of physic, as when, previous to death, odorous and volatile drugs (such as camphor, prussic acid, turpentine, creosote, chloroform, etc.) have been administered; or the animal may have fed upon odorous plants; or, subse- quent to death, the carcass may have been hung in an atmo- sphere which is odorous from any cause (tobacco, carbolic acid, etc.). It has been shown that no dangers to the meat would arise from the administration during life of medicaments such as arsenic, antimony (tartar emetic), or strychnine. The animal may in some cases take in poison by its food, by feeding upon such herbs as bryony, meadow-saffron, rhus toxicodendron, etc. There are few changes which are so easy to detect as commencing putrefaction in fish ; this is fortunate, inasmuch as decomposition sets in rapidly and appears to be more generally productive of poisonous symptoms than decomposing meat. The bright gills, the prominent eyes, the elastic resistance of the firmly adherent flesh, and the absence of any but the characteristic odour, are all evidence of freshness. The soft inelastic feel, the readiness with which the flesh can be detached from the bone, and the stale and unpleasant odour, furnish the chief clues — and the most reliable — of commencing decomposition. It has been 296 LABORATORY WORK found possible to revive the greyish gills by artificial colouring agents, and to keep the eyes prominent by a small piece of stick, fixed trans\ersely in the head, so that it presses the eye outwards on either side. Greenness, iridescence, and sometimes luminosity, may be seen upon the surface of the flesh of decomposing fish. Stale fish float, while fresh fish sink in water. In putrefaction of meat the flesh softens, and tears readily; it becomes paler ; the elastic resistance gradually diminishes and becomes less uniform — i.e., some parts are softer than others; the characteristic odour is developed; the marrow softens and turns brownish; and the juices become alkaline in reaction, due to ammonia and substituted ammonias being formed by the action of schizomycetes. Later, the meat becomes of a greenish hue, and a glance then suffices to detect the presence of putre- faction. Occasionally meat becomes luminous, chiefly from the presence of Bacillus phosphorescefis ; but putrefaction eventually disperses this condition. Putrid meat may grow dull, dark moulds upon its surface. As a test of putrefaction Eber recommends the use of a reagent composed of i part of hydrochloric acid, 3 parts of alcohol and I part of ether. A few c.c. of this reagent are placed in a cylinder, which is then shaken so that the reagent applies itself to the inner surface of the cylinder. If a fragment of meat in the state of incipient putrefaction is introduced on the end of a wire, greyish or whitish fumes of ammonium chloride will generally appear. The moulds which grow on the surface of meat are of numerous varieties — penicillium, mucor, phycomyces, verticillimii, oospora, etc. Red growths of Bacillus prodigiosus may also be associated with mould. There is little or no evidence that these growths or their products are injurious to health, although the meat is rendered unsightly and often unsaleable. Mould contain ination is especially liable to occur when meat has been improperly handled or stored. Certain diseases may cause characteristic appearances in the meat. When any such suspicion attaches itself to the sample of meat under examination, it is a great advantage to obtain a glance at the offal of the animal, and more especially to carefully inspect the liver, lungs and l3miphatics. The term " offal " includes the head, the feet, the skin, and all internal organs except the kidney; the remainder of the animal is termed the " carcass." PARASITES OF FLESH 297 It is when there is evidence of parasitic attack that the flesh presents the most characteristic appearances. Of those organisms, which are commonly classified as " animal parasites of flesh," some only are capable of infecting human beings when the flesh is eaten. Harmless Animal Parasites of Flesh. Coccidia oviformes (Leuckart) infest most animals (rarely man), and are chiefly found in the livers of rabbits, where they appear as small white nodules, which under the microscope are seen to contain clear ovoid bodies with either granular contents or egg- like structures known as sporoblasts. Ccenufus cerebralis forms hydatids, varying in size from a pea to a small walnut, in the brain and spinal cord of the ox and sheep. It is the cystic worm FIG. 61. COCCIDIUM OVIFORME, SHOWING DEVELOPMENT OF SPOROBLASTS. of TcBfiia ccenurus of the dog. Cysticercus fisiformis is found in the abdominal cavity and liver of the rabbit and hare; it is occasionally found in man, and the cysts are about the size of a pea. C. tenuicollis is found in the abdominal cavity of animals generally; it is the hydatid of the tape-worm, T. marginata, which inhabits the intestine of the dog; the cysts vary in size from a pea to a small orange, and do not invade the organs; the long thin neck is characteristic of the parasite. C. serialis is the immature form of a tape-worm affecting dogs; the cysts, varying in size from a hazel-nut to a pigeon's egg, are found Under the skin and between the muscles. Strongylus filaria in the bronchial tubes of sheep, 5. micrurus in the lungs of cattle, and S. paradoxicus in the lungs of the pig, are nematodes. There is another parasite found in the lungs of sheep, known as Strongy- lus riifescens, of which the eggs or embryos are deposited in the lung substance, forming little nodules which are usually of a greyish-yellow colour. This condition is found in adult animals. 298 LABORATORY WORK and is \-ory common; it is called by many "pseudo-tubercu- losis " of sheep. The lungs of at least 60 per cent, of all sheep slaughtered are affected by this parasite. Certain nodular masses in frozen quarters of meat arriving from Australia have been found to contain the parasitic worm Onchocerca Gibsoni, which gives rise to a condition known as onchocerciasis. Dr. Leiper found no evidence of vitality in the worm or in its embryo in the Australian beef arriving in this country, and it seems to be impossible for the parasite to develop in man from eating the affected meat. Apart, however, from this danger, the meat itself should be classed as unsound. The condition is most marked in the fore-quarters, where it is more or less con- fined to the region of the flank and brisket. Dangerous Anim.^l Parasites of Flesh. Cysticerci. — The cysticerci, or " bladder- worms," cause the condition known as " measles " in the pig, ox and sheep. Cysti- cerci celhdoscB are the bladder-worms which form a stage in the development of Tcenia solium. In the flesh of the pig, and rarely in that of dogs, monkeys, or man, a number of small oval or round cysts are seen, occupying a position between the muscle fibres, and commonly of the size of a small pea — though they have been found as small as -^-^ inch, and as large as ^ inch, in length. They are surrounded by a pale, milky-looking fluid, and the cyst wall shows a white spot (generally central) upon its surface. The affected flesh is pale, soft, unduly moist and flabby, and has a smooth, slippery feel. The flesh does not set well, and quickly decomposes. Sometimes there is some degree of calci- fication of the capsule, and the result is that when sections are cut a grating sensation is experienced. The bladders should be incised with a sharp knife, and the worm examined by a powerful hand-lens, when at one extremity will be found the blunt square head provided with a sucker at each " angle," and a fringe of about twenty-eight booklets placed more centrally. These latter are very characteristic, and must always be found before a definite diagnosis is ventured upon. Those cysts that are dried up and indistinct can be made visible by soaking in weak acetic acid. Ostertag attaches great PARASITES OF FLESH 299 diagnostic importance to the rounded or oval calcareous cor- puscles which are so generally embedded in the tissue of the head, but which disappear on the addition of acetic acid. The liver and the muscles of the shoulders, intercostals and loins, are chiefly affected. A staining test will generally suffice to determine if the cysti- cerci are alive or dead. If dead, the whole bladder- worm readily stains with carmine ; if alive, the head at least resists the stain. Cysticercus bovis, or " beef -measles," chiefly affects the calf, and is never found in man. It is somewhat smaller than C. cellu- loses, and possesses a flat head with no booklets, but merely suckers, around which there is frequently a considerable deposit of pigment; and on the surface of the head there is a pit-like FIG. 62. HEAD OF T^NIA SOLIUM. (OBJ. I INCH.) FIG. 63. HEAD OF TJENIA MEDIO- CANELLATA. (OBJ. -|- INCH.) depression (" frontal suction cup "). It develops into the adult tape-worm called Tcsnia mediocanellata, or T. saginata, which is longer than T. solium. Fish are subject to parasitic attack, and notably is this the case with the cod, in which many parasites have been found. Cooking effectually destroys them, for in the case of fish the flesh is not palatable unless the cooking is thorough. Bothriocephalus latus, a tape -worm which is almost limited to certain parts of the continent of Europe, is even larger than Tcania mediocanellata, and has a club-shaped head, not armed with rostellum or booklets, but possessing two deeply grooved longitudinal suckers, one on each side. The eggs are oval and comparatively large, with a characteristic operculum. Man is infected through eating imperfectly cooked fish, especially the pike, perch, and several members of the salmon family. There is no cysticercus form. 300 LABORATORY WORK T. cchinococcHs is the small tape-worm, of three or four seg- ments, which is commonly found in the dog. The encysted stage (" hydatids") is most generally found in the lungs and liver, of oxen, sheep and swine, but also (more especially in Iceland) in man. The hydatids consist of thin pale vesicles, floating in a clear liquid, and the whole is encysted in a tough capsule. The inner lining of the capsule consists of ciliated epithelium, and inside of the cyst wall there are generally many so-called " brood capsules" (Fig. 64). The cysts vary in size from a pin's head to that of a large orange. They may exist in such numbers in the liver that they replace the greater part of the entire tissue of that organ. The condition is diagnosed with certainty by the microscope, either by the discovery of the characteristic heads or detached FIG. 64. BROOD CAPSULE OF AN ECHINOCOCCUS. hooklets in the clear liquid of the cyst; valuable corroborative evidence being furnished by the fact that the liquid is quite free from albumin, and, in consequence, does not coagulate on boiling. Tcsnia nana is the smallest human tape-worm (12 to 20 milli- metres in length), and is not uncommon in Italy. The head con- tains four suckers, and a rostellum carrying twenty-two to twenty-four hooklets. It differs from T. solium in being very much smaller, and the rostellum of the latter carries a double row of hooklets, twenty-eight in number. T. cucumerina is a little larger than T. nana ; it occurs in man, especially in Norway and Sweden, but it is most common in the dog. The head contains four suckers, and three or four rows of hooklets (sixty in all) are disposed round the rostellum. Among the Cestoda monstrosities may sometimes be observed, with abnormalities as to the number of suckers and hooklets. PARASITES OF FLESH ^,01 Trichina spiralis. — This parasite has been fonnd in tJic flesh of many different animals (pigs, pigeons, eels, etc.), but most commonly by far in that of pigs; oxen and sheep do not suffer from attack by these nematodes. The disease is often seen in Germany, but rarely in England. The shape of the minute worms is nearly that of a typical nematode — i.e., a slender rounded body tapers gradually at either end; the extremity which constitutes the head goes to a long slender point, which presents a small central orifice, the mouth. The other extremity, the tail, ends more bluntly. The worm possesses a distinct alimentary canal, and even rudimentary sexual organs are present. In the female a uterus is discernil)]e, which will frequently be seen to be full of minute free embryos FIG. 65. TRICHINA SPIRALIS, ENCYSTED IN MUSCLE, (X ABOUT 50 DIAMETERS.) curved upon themselves; these latter have been observed to become extruded from the vagina, and subsequently to move sluggishly about the field of the microscope. The male worm is much smaller than the female, and is only about yV ™ch long when mature; the length of the latter reaches | inch. The long slender head and blunt tail are two characteristics which serve to distinguish these worms from parasites which otherwise resemble them, such as Drac^mctdus and Filaria sanguinis hominis. The small worms are mostly coiled up in cysts, so disposed that their longest diameter is in a line with the muscular fibres, and a drop of acid will stimulate them to transient movements if they are alive. These cysts lie between the muscle fibrillse, and their walls are sometimes partially or completely calcified. so as to give a grating sensation when the finger is passed over 302 LABORATORY WORK a section of the flesh. This calcareous deposit serves to shield the parasites from the destructive consequences of salting, and maybe even of cooking. There may be from one to three trichinae in a cyst. Frequently 25 per cent, of these parasites are thus encysted in the diaphragm, and therefore, when possible, a piece of this muscle should be procured ; the back muscles, on the other hand, are the least attacked. Either a section may be made of the muscle, or it may be teased out with needles ; and in the case of a long muscle, a point near its insertion should be selected, since this is a favourite site for encystment. The affected muscle is seen to be pale and oedema- tous, and if the worms are encapsuled, small, rounded (or more truly, lemon-shaped), whitish specks, averaging about the size of a very small pin's head, are visible to the naked eye. These can be made very distinct by means of a hand-lens; but a low power of the microscope should be employed in every case, when the most characteristic appearance will be got by making a thin longitudinal section of the affected muscle and immersing this in potassic hydrate solution of medium strength, which serves to make the muscle fibres transparent and leaves the worm exposed in its coiled condition within the capsule. The soaking should not be prolonged beyond a minute or two, or the worm itself will also be cleared up. Glycerine is a good mounting medium when a permanent specimen is desired. Sometimes a view of the worm is obscured owing to considerable calcareous deposit in and around the walls of the capsule; in these cases a drop of dilute hydrochloric acid, run under the cover-glass, will dissolve up the deposit; or an oil globule, or several, may partially obscure the worm, when a drop of ether, applied in a similar manner to the acid, will clear away the fat. There are generally oil globules at the poles of the capsule. The parts which are most likely to be affected will easily be remembered if it is borne in mind that the worms migrate to their settlements from the gastro-intestinal tract, and chiefly from the commencement of the small intestine. The diaphragm, the liver, the intercostal and abdominal muscles, are necessarily the first encountered, and therefore suffer most; but in later stages of the infection there is scarcely a muscle which may not be affected. Ihere are small, semi-transparent bodies called " psoro- spermia," or " Rainey's capsules," which to the naked eye PARASITES OF FLESH 303 resemble trichinae ; but they consist of small dark oval or elliptical bodies, of greater length than encysted trichinae. They are, moreover, made up of a thick membrane formed by small hair- like fibres arranged in lines, which encloses small, kidney-shaped, granular cells closely adherent together; and the whole lies embedded in the muscle substance itself— ^'.g., the sarcolemma. They are extremely common, and may exist in the flesh of most of the animals used for human consumption, and apparently when eaten they produce no ill-effects. Several other obscure bodies, the nature and significance of which we are still more ignorant of, may exist in flesh, such as bodies somewhat resembling pus cells, and others forming minute concretions or tiny hard nodules. Interesting as these are FIG. 66. ONE OF RAINEY'S CAPSULES. (X4O.) pathologically, they are rare ; and when present, even in numbers, do not appear to effect the wholesomeness of the meat to any degree. Actinomyces. — The " ray fungus " (actinomyces), one of the " fission fungi," is now recognized as a parasite of commoner occurrence in the ox than was once suspected; the difficulties which stood in the way of an earlier appreciation of this fact arose from the circumstance that both the ante- and post-mortem appearances of actinomycosis so closely simulate those of tuber- culosis. It has not yet been proved that the disease can be communi- cated by the flesh of animals (bovines) suffering from attack, and the vitality of the fungus when exposed to heat is very low. The parasite almost entirely affects the tongue, the jaws (especially the lower one), the muscles of the cheek and the lungs, where they may be detected by the naked eye as small dirty white specks, commonly about the size of a barley grain, but varying from the tiniest speck to | inch in diameter. On section, the centre of the nodule is seen to be softer and of a 304 LABORATORY WORK greenish-yellow colour, or, less frequently, the nodule is firm and fibrous throughout. The condition is generally associated with considerable fibrous proliferation of affected parts. The parasites assume, when encysted, a peculiar synmietrical appearance, due to the fact that they consist of small linear elements, thicker at one extremity than at the other, and are so arranged that their smaller extremities are all directed towards a central point ; the stellate or rayed appearance thus created is sometimes remarkably regular and uniform. The tongue when affected is hard and swollen, and presents the flattened nodules chiefly upon its dorsal aspect. The size of these nodules may vary from I inch to 2 inches, and the glands at the root of the tongue are also commonly infected. Distoma Iwpaticmn. — To examine for these parasitic trema- todes, the liver should be taken and the bile ducts carefully FIG. 67. DISTOMA HEPATICUM. (NATURAL SIZE.) exposed. In shape like little soles, they are of a pale-brown or slaty colour, and are provided at their broad extremity with two suckers, one at the anterior end and the other a little above the junction of the anterior and middle thirds of the median line. Their surfaces are beset with many little warty points, and they average in size from i to i| inches in length and about \ inch in width. The bile ducts of the affected liver stand out on the surface (" pipey " liver). They are sometimes found encysted in the lungs of both sheep and cattle, when the cyst wall is usually calcareous and contains a chocolate-coloured fluid. Frozen Meat. — Meat which has been frozen may be detected by expressing a drop of the meat-juice on to a glass slide, covering with a cover-glass, and examining by the microscope. The blood- corpuscles will be found to be much distorted in form, to have lost their pigment, and to be floating in a highly coloured serum; whereas the juice of fresh meat will show corpuscles of normal shape and colour, floating in a practically colourless serum. POISONING BY FOOD 30^ Compared with fresh meat frozen meat has usually a darker and more diffused red colour when thawed, owing to the haemo- globin permeating the tissues; it is also somewhat softer. Poisoning by Food. There is no doubt that flesh in a very early stage of decompo- sition disagrees with many persons; and abundant evidence is ' not lacking that when an advanced state of putrefaction has been reached, violent gastro-intestinal irritation, followed by diarrhoea, vomiting and toxic symptoms, may be induced. Recorded cases of grave and fatal food poisoning have been very numerous, and minor disturbances of the gastro-intestinal tract are probably often due to small doses of poisons produced by bacteria. Often the offending substance appears, to all physical tests, to be quite good and wholesome ; but the meat or jelly formed is sometimes observed to be softer and moister, and a slight peculiar odour has been noted. The poisonous food has most commonly been brawn, sausages, ham, pork, veal pie, rabbit pie, potted shrimps, tinned salmon, mackerel, mussels and oysters ; but many other varieties of food have been incriminated, such as cheese, ice-cream, canned goods, potatoes, etc. There is a strange absence of recorded instances where the flesh of sheep has been the offending food ; but this is seldom used in the preparation of made foods. In food which has become poisonous, a living micro-organism has produced an organic chemical poison, which may be a ptomaine, albumose or toxine. The substance is the immediate cause of the morbid symptoms, and is probably produced by the action of the micro-organisms on the albuminous constituents of food. Both the products of specific micro-organisms in an in- fected food and those basic substances resulting from putre- factive micro-organisms (" ptomaines ") may be fleeting, as regard their existence, since the micro-organism may be killed by its own products or by heat, or the chemical poison (from its unstable nature) may undergo decomposition ; so that an infected food which may be poisonous at one time may fail to be poisonous at another. The micro-organisms may produce their peculiar chemical poisons from material affording them nourishment, which may be either outside the body of man or within it. Food 3o6 LaBoratorV Work poisoning outbreaks are far more prevalent in the summer months. In the majority of cases the symptoms of poisoning occur within twelve hours, when they are due to the ingestion of already formed poisons (" intoxication ") ; but in other cases the symp- toms may be delayed for twelve to forty -eight hours, when they are probably due to a food " infection " by organisms which produce poisons after the food is taken into the human body. Generally there is a mixture of bacilli and toxines and therefore variable incubation periods. In most cases the sjmiptoms include con- siderable abdominal pain and tenderness, vomiting and diarrhoea, with tenesmus, headache, and marked depression or collapse. Other frequent symptoms include dilated pupils, rashes, and albuminuria. The temperature may be raised, but in most cases it is subnormal. Although the acute symptoms disappear after two or three days, marked prostration may persist for a much longer period. The mortality among those affected by meat poisoning is often from 2 to 5 per cent., and except in cases where rapidly fatal results follow, a post-mortem examination usually discloses gastro- enteritis, and sometimes ulcers in the small and large intestines, with enlarged spleen and congested liver and kidneys. The most frequent fonn of meat poisoning results from con- suming the flesh of diseased animals, and is associated with the presence of Bacillus enieritidis (Gaertner) and the Paratyphoid Bacillus ; but healthy carcases may be infected in the slaughter- house by the knives used for the purpose of dressing many animals, and cooked or uncooked meat may be infected by rats, mice, flies, dust, ice, or the soiled hands of human "carriers." Ptomaine poisoning from eating putrefied meat, and usually associated with the presence of B. proteus and B. coli, is less common; and " botulismus," which results from the toxine of B. hotulinus, is of extreme rarity in this country. Ptomaine poisoning has most frequently resulted from eating meat in a cut-up condition (sausages, minced meat, etc.), and game; and especially meat which has been insufficiently cooked (for the thorough cooking of meat destroys B. proteus and its toxines). Ptomaines are not the cause of extensive outbreaks of meat poisoning. The infection of food by the bacilli of the Gaertner group is the common cause of these. B. suipestifer, foimd in POISONING BY FOOD 307 the pig's intestine, is another member of the Gaertner group, and is the bacillus of swine fever. This organism has been shown to be responsible for some outbreaks of poisoning follow- ing the consumption of pig's flesh. The power of certain shellfish to create violent gastro-intestinal disturbance and urticaria is well recognized; mussels, cockles and oysters collected from near sewage outfalls have had virulent poisonous properties ascribed to them, and they may convey the infection of enteric fever. From mussels a poisonous ptomaine, mytilotoxine, has been isolated by Brieger. This exists chiefly in the hepatic organ of the mussel. Acute gastro-intestinal and profound nervous symptoms may follow the consumption of either the raw or cooked mussel. Perch, sturgeon, turbot, pike, crabs, shrimps, salmon and sardines have all given rise to poison- ing, either from the development of putrefaction toxines, or from bacterial infection or intoxication. Some fish normally contain a toxine poisonous to man, others develop such a poison only in the spawning season. The consumption of canned fish may give rise to symptoms akin to " botulismus." Some ptomaines are highly poisonous, while many are inert. The majority of the known ptomaines contain only C, H and N, and represent simple ammonia substitution compounds. The kind of ptomaine formed will depend upon the organism present, the nature of the food substance, the temperature, the stage of putrefaction, etc. A large number of toxic bodies have been isolated and de- scribed, some of the more important being — Methylguanidine Dihydrocollidine Neurine Choline Muscarine Poisonous. Poisonous. Poisonous. Poisonous. Poisonous. Obtained from poison- ous mushrooms and fish. Gadinene But little poisonous. Obtained from putrid fish. Mytilotoxine . . . . Poisonous. Obtained from mussels, taken from positions liable to sewage pollution. Tyrotoxicon .. .. Poisonous. Obtained from cheese, cream, and ice-creams. 308 LABORATORY WORK The B. botulinus is an anaerobic organism, and botulism generally results from the eating of sausages in thick skins (German sausages) and other food which has been hermetically stored or embedded in fat. In botulismus the main symptoms are not gastro-intestinal, but nervous, such as dryness of mouth and throat, difficulty in swallowing, ptosis, double vision, convulsive muscular tremors, vertigo, constipation, retention of urine, disturbances of heart's action and of respiration; consciousness is unimpaired, and the combined symptoms, which generally appear in from twenty- four to thirty-six hours, though sometimes earlier, are akin to those of atropine poisoning (Dieudonne). Vanilla flavouring in sauces or ice-cream has often given rise, within two hours, to vomiting, tenesmus, diarrhoea, and signs of collapse. The explanation appears to be that the vanillin, by its reducing action, favom^s the growth of anaerobic bacteria ; the vanilla itself being harmless. Potatoes may give rise to similar symptoms of poisoning, which are said to be due in some cases to a considerable increase of the trace of the poisonous substance " solanin " to be found in normal potatoes, but generally to bacterial decomposition by proteus bacilli (Dieudonne). It has been recommended that in order to guard against solanin poisoning, the peel and any sprouts should always be thoroughly removed. Cheese, milk, cream, butter, and ice-creams, etc., may all rarely contain tyrotoxicon, which develops under those circum- stances most conducive to fermentative changes generally — viz., warmth and moisture, impure and confined air, and deficient light. It is a diazo-benzene-butyrate, and is found to occur under conditions of improper storage in the various food articles above mentioned; and when the ingestion of any of these articles has given rise to serious consequences, then a search should be instituted for this most powerful poison. The symptoms it creates commonly pass off within a few hours, but occasionally serious consequences have arisen, such as the development of symptoms akin to atropine poisoning, which may be followed by fatal collapse. The physical characters of the article are not necessarily altered in any way, but acidity is always marked, and where this is normally present (as in cheese) it is invariably increased. The method of examination in the case of milk is as follows : POISONING BY FOOD 3^9 1. The filtrate from the milk is first n^ndercMl distinctly alkaline by means of sodimTi carbonate; then an equal bulk of pure ether is added, and the whole well shaken up in a separator. 2. The mixture is next allowed to stand until all the ether has separated into a layer upon the surface. 3. This ethereal layer is then decanted on to a saucer, where it is left until the ether has spontaneously evaporated and a comparatively dry residue remains. 4. The residue is carefully dissolved in a little pure water and then filtered to free it from fat. 5. The filtrate is next well shaken with an equal bulk of pure ether, and the ethereal layer, having separated, is removed and allowed to again evaporate spontaneously. 6. The residue left upon the saucer will then contain any tyrotoxicon which may have been originally present, sufficiently pure to respond to the following test : on the addition of a few drops of a mixture of equal parts of pure carbolic and sulphuric acids, if the poison be present, a reddish colour appears. In the case of cheese and butter, these are first thoroughly worked up (triturated) with water, and the filtered extract is then treated in the manner indicated above. It is probable that cheese poisoning is usually due to toxines produced by bacteria. In order to investigate cases of alleged food poisoning a con- siderable amount of the suspected material is necessary. Perish- able articles should be placed in an ice-box. The amount generally submitted for examination is often too small to admit of a thorough bacteriological and chemical examination. The bacteriological evidence will be furnished by the morphological characters of the organisms found in the food, or the sufferer's vomit, fgeces, etc., their pathogenicity, and serum reactions. With respect to the chemical examination, that will be concerned more particularly with the nature and amount (if present) of chemical preservative, metallic contamination, or colouring agent; little is to be gained by testing for the presence and attempting to define the nature of " ptomaines." Rats and mice should be fed with portions of the sample, while other animals — e.g., guinea-pigs— should be inoculated subcu- taneously and intraperitoneally from cultures and also from the broth emulsions of the food. If any of the animals die, a com- plete post-mortem examination should be made. It is also advisable to cut sections and otherwise microscopi- 310 LABORATORY WORK callv examine the meat to see if the bacilli are chiefly on the surface, and also if the meat-fibres arc from apparently healthy animals. An autopsy should be made in all cases of death from food poisoning of this nature, and cultures made from the different organs (spleen, mesenteric glands, intestines, etc.). As already mentioned, the agglutinative properties developed in the blood of suspected cases (sufferers or carriers) should be made use of for diagnostic purposes. The alkaloidal products of putrefaction are separable by their relative solubility in alcohol, ether, chloroform, etc. The following method (after Stas) may be adopted : The finely mixed material is drenched with 90 per cent, alcohol, tartaric acid is next added until the liquid is definitely acid (if the fluid is already acid this is not necessary), and the mixture is then allowed to digest for several hours at 70° to 75° C. After cooling, the alcoholic liquid is removed (the last portions by pressure) and filtered. This operation is repeated several times, and the united filtrates are evaporated in vacuo at 35° C. to a small bulk. The liquid is then filtered through a wet filter to remove fatty matter, powdered glass is now added to the filtrate, and the mixture evaporated nearly to dryness over sulphuric acid in a vacuum. The residuum is next digested in pure alcohol for twenty-four hours, and again evaporated (at 35° C. in vacuo) to dryness. This residue is dissolved in a little water made alkaline with sodium bicarbonate, and the solution is then well shaken with 4 volumes of pure ether. Finally, the ethereal solution is decanted, evaporated to dryness, and the alkaloid is left behind. The extract which may thus be obtained may be dissolved and tested with various reagents in order to ascertain its par- ticular nature; or it may be given to a lower animal and its effect noted, if it is only necessary to learn whether the original material was harmful or not. The best animals for such ex- perimental purposes are mice. A quantitative and qualitative bacteriological examination of oysters and other shellfish is often very desirable in the interests of public health, to see how far they are free from excretal and sewage pollution bacteria. The most thorough method (Houston) for the bacteriological examination of oysters is briefly as follows: POISONING BY FOOD 3^^^ I The outsides of the oyster-shells are well scrubbed with soap and water, and cleaned as thoroughly as possible in running tap-water, finally with sterile water. _ o Hands of the investigator are thoroughly cleaned, washed in I in 1,000 corrosive sublimate solution, and finally with sterile 3 Oysters opened by a sterile knife held in position by a sterile cloth and with concave shell underneath. Great care must be taken to avoid any loss of the liquor. The liquor m the shell is poured into a sterile i,ooo c.c. cylinder, and the oyster and oyster liquor are added after the oyster has been cut into small pieces by sterile scissors. 4. Ten oysters are to be treated as above m each experi- ment. T _CC 1 11 5 The volume of oyster + oyster liquor is read off, and usually varies between 8o and 120 c.c. For qualitative work 100 c.c. may therefore be taken as a fair average of the total shell contents of ten oysters. ^ .u Sterile water is then poured into the cyhnder up to the I 000 c c. mark, and the whole well stirred with a sterile rod. Each 100 c.c. of this liquor may be considered to contain the bacteria in one oyster. 6 Various amounts and fractions of this liquor are used for the examination for B. coli, B. enieritidis sforogenes, and for streptococci. tt + a For the B. enieritidis sporogenes examinations Houston used 10 I o-i o-oi, o-ooi c.c, and for B. coli these dilutions, and m addition 100 c.c. and ooooi c.c. He made these primary cultures in triplicate, and required at least two out of the three to be positive for a preliminary numerical diagnosis. Cockles and mussels may be examined by a very similar procedure. Ice-Cream.— There is a considerable probability, and some evidence, that ice-cream may be a means of spreadmg disease, particularly typhoid fever. In investigating such a possibility the bacteriological examination of ice-cream is of value. The ice-cream should be collected in a sterile vessel— ^.g., a wide-mouth sterile bottle with glass stopper— and packed m ice if it cannot be examined at once. To examine, melt the ice-cream by placing for fifteen to twenty minutes in the 22° C. incubator, then treat as a milk sample. 3T2 LABORATORY WORK Edible and Poisonous Fungi, -With a viow to enable residents in tlie coiHitry to distinguish between tlie poisonous and edible kinds of fungi, and to utilize to a greater extent those varieties which are useful as food, the Board of Agriculture and Fisheries has published a small book of illustrations of those species which are more commonly found in Great Britain, together with brief descriptions of them. It is pointed out in this publication that, contrary to popular belief, the poisonous kinds of fungi are com- paratively few in number, while there are, on the other hand, some fifty species of edible fungi which may safely be eaten. In order to recognize with certainty these different kinds, it is necessary to know those special features possessed by each species which separate it from all others. The rule-of-thumb signs for discriminating between edible and poisonous fungi are valueless, and no reliance should be placed on the presence of a skin that is readily peeled off as an indication of an edible fungus, or on the statements that a silver spoon placed in contact with poisonous kinds becomes tarnished, and that all fungi growing on wood are poisonous. In England and France most of the cases of fungus poisoning which have been reported are due to Amanita phalloides. Mus- carin is the toxine to which mushroom poisoning has been com- monly ascribed, but the poison of this fungus is probably a toxalbumin. Amanita -phalloides is white beneath the cap, with a yellowish-white or greenish-white, shining top; and the stem, which is white and smooth, is bulbous, and is clothed at its upper part with an expanded pendant ring. It usually occurs in woods, and it rarely, if ever, grows far away from trees, having a preference to the proximity of the oak variety of tree. The fungus peels almost as well as the common mushroom. Horse-Flesh. In horse-fiesh the meat is darker and more brownish; it is coarser (the muscular fasciculi being broader) than in ox-flesh; the odour of the fresh meat is different, and after the lapse of a day or two, as the flesh dries, it develops a peculiar faint odour, a bluish sheen, and imparts a soapy feeling to the fingers. The fat is more yellow and soft, and possesses a sickly taste; and, in consequence, it is sometimes removed and replaced by ox-fat, which is skewered on to the meat. If the bones have not been HORSE-FLESH 31 3 removed, they wi]] afford an additional clue, inasmucli as they are larger and their extremities (tuberosities, etc., for the attachment of muscles and ligaments) are larger and more marked; there are, in addition, some anatomical differences in the construction of the horse's skeleton. In cattle the breast-bone is broad and flattened, while in the horse the front portion is keel-shaped. The ribs in cattle are flatter and broader in the middle and lower thirds than in the horse. Horse-flesh is richer in glycogen than ordinary meat, and as this remains unchanged for a long time it is taken advantage of in the following test : An infusion of the suspected meat is prepared by boiling 100 grammes of it, in a finely minced condition, with | litre of water, for nearly one hour; the broth is then concentrated to about 100 c.c, and is mixed, when cold, with dilute nitric acid in the proportion of 5 c.c. to loo c.c. of the broth; it is then filtered and tested with hot (freshly prepared) saturated solution of iodine, gently let down upon it so as not to mix the liquids. If the substance is horse-flesh, a marked reddish or brownish-violet ring appears at the junction of the two liquids. If a deep violet colour results, starch is present, as may be the case in sausage- meat. The iodine solution should consist of iodine, i gramme; iodide of potassium, 2 grammes; water, lOO c.c; and it must be carefully added, since a slight excess changes the colour to reddish-brown . Glycogen is present in the livers of cattle and in meat extract, and the dextrines derived from starch in sausages give a similar reaction; therefore it is only when the amount found is considerable that the glycogen reaction can be taken as certain evidence of horse-flesh in sausages. Moreover, in old sausages containing horse-flesh the glycogen is liable to undergo decomposition, and then no reaction is obtained. In smoked horse-flesh the glycogen is destroyed. Certain organs of the horse are occasionally sold as the corre- sponding organs of the ox. The tongue of the horse is broad and rounded at its free end, instead of pointed, as in the ox; and it is smooth at the base, where the tongue of the ox is rough with prominent papillae. If the hyoid bone is attached, it is found to be made up of five parts, whereas that of the ox consists of nine. The epiglottis is, moreover, smaller and more pointed in the horse. The liver, whether of the ox or sheep, consists of one very large lobe, and another relatively small one; in the 314 LABORATORY WORK horse tliero arc three large and distinct lolios, and a fourth which is very small, and there is no gall-bladder. The kidney of the horse is more heart-shaped, and cannot be mistaken for the long lobulated kidney of the ox. The heart of the horse differs from that of the ox in being less conical ; it is also darker, softer, and has less fat at its base. There is a bone in tlie heart of the ox (os cordis). Sausages. These are made of the chopped flesh or internal organs of various animals, mixed with condiments, flour, bread, or potato meal, and filled into clean gut or parchment; the sausages are then generally boiled, smoked, or scalded. Saltpetre is some- times added to furnish a good red colour to the meat, and often colouring agents (carmine, cochineal, or aniline) and preservatives are added. The colouring matter can generally be extracted by warming for several hours with a mixture of equal parts of glycerine and water. Boric acid is often used as a preservative. It is certain that since boric acid prevents objective decom- position, such as is detectable by odour, it permits of the use of stale meat and meat in the early stages of decomposition for the making of sausages. Dr. C. A. MacFadden, in a Report to the Local Government Board (1908) expresses the view that if boron preparations are permitted, 0-25 per cent. (17-5 grains per pound) of boric acid should not be exceeded, and that even then such addition should be made known to the purchaser. The use of antiseptics in sausages is to be discouraged. Even the boric acid, which is most frequently employed, may be injurious to those with weak stomachs or kidneys. Meat sausages (pork and otherwise) can be made, sold, and consumed without any such addition, if good fresh material is employed, for people do not purchase sausages to store in the house ; and the dried " German " sausages can be sufficiently preserved by other means. While the amounts of boric acid usually employed will not enable the use of meat which has reached a stage of marked putrefaction, they permit of the use of stale material in a state of incipient decomposition, and while they may reduce the danger from putrefactive toxines, they do not materially reduce the risks of poisoning from organisms of the Gaertner group or from SAUSAGES 315 Bacillus hoiulimis. Owing to its cheapness (beef sausages arc generally sold at about 5d. a pound), the poorer people consume a considerable quantity of sausage, f pound of which could be eaten at a meal by an adult. That the use of chemical antiseptics is unnecessary is demonstrated by the large number of makers who never employ them. The writer has satisfied himself, by experiments, that if the meat is sound and fresh at the time it is put into a sausage it will keep in hot weather for forty-eight hours, and that as little as 12 grains of boric acid to the pound will enable the sausage to be kept in such weather for four days ; so that even if antiseptics are permitted, and it is necessary to keep fresh sausages for four days, there is no case for the employment of the 20 to 30 grains of boric acid to the pound which is sometimes found to be present. It is not a difficult matter to detect early decomposition in sausages ; the alteration in the odour will sometimes suffice ; for if a little of the sausage is boiled with water and some freshly prepared lime-water is added, good meat yields only a faint ammoniacal odour, whereas bad meat will give off a peculiarly offensive ammoniacal odour. Putrefaction generally commences in the middle of the sausage, when a dirty greyish-green colour is often to be noted. The skins of sausages have been known to contain mineral poisons, but this is very rare. Eggs. The best tests for bad and stale eggs are the following: 1. Fresh eggs are most translucent towards their centres if held vertically against the light ; stale eggs are most translucent at their upper extremities. 2. If 2 ounces of salt are dissolved in a pint of water, fresh eggs when placed in the solution sink, and stale ones float. Opaque spots are generally due to moulds that have gained access through a crack or cracks in the shell, or to embryos. Eggs generall.y contain bacteria and sometimes parasitic worms, but there are no recorded instances of human infection. Toxic poison has been separated from decomposing eggs. Meat Preparations. Many meat extracts are now upon the market, the tendency being for the public to over-estimate their food value. They consist of the extractives of meat, and not of the meat itself; 3t6 laboratory work and they act as stimulants and regulators of digestion rather than as true foods capable of providing the necessary amount of nitrogenous material for tlie needs of the body. A meat extract should consist of a golden-brown sticky sub- stance, with a pleasant meaty odour. It should never be hard, and should attract moisture strongly from the air. The reaction should be slightly acid. The usual method of preparation consists in heating raw meat, which has been finely divided, with a little water under pressure. The extract thus made is filtered and evaporated in vacuo. It is essential that a tem- perature below 75° C. be employed if all gelatine is to be excluded (Beveridge). The extract thus made contains the flesh bases or extractives and mineral matters of the meat, and is free from albumin, meat fibre, gelatine, and fat; but in some of the meat extracts on the market these substances and also vegetables are subsequently added in order to give the extract a certain food value. Meat juices are prepared in the cold by subjecting finely divided meat to strong pressure, and ultimately concentrating by evaporation in vacuo. They contain the soluble proteins of meat. Protein matter constitutes the bulk of meat extract, but it varies greatly in amount in different samples. The total nitrogen, calculated as protein, should not be less than 55 per cent., of which not less than 25 per cent, must be insoluble in alcohol. Protein soluble in alcohol is presumed to consist of half meat bases and half meat extract, but from a practical point of view these distinctions are unnecessary in routine examina- tions. Water should not exceed 25 per cent, and the ash 20 per cent.; the former varies from 14 to 25 per cent., and the latter from 14 to 33 per cent., in samples of the usual brands of meat extracts. Meat Essences. — ^A meat essence is a more liquid extract, containing more water, but has the same colour, odour, and reaction. Protein matter should not be less than 9 percent., of which not less than 7 per cent, should be insoluble in alcohol. Water should not exceed 90 per cent., nor the mineral ash 1-5 per cent. The analysis of a meat essence is similar to that of a meat extract, about ten times as much of the substance being taken for each examination. On account of the large amount of water present, frothing of MEAT PREPARATIONS 317 the contents of the Kjeldahl flask is apt to be troublesome, and unless the first stage of the process is carefully watched the fliusk contents may be ejected. The gelatine in table jellies usually amounts to from 13 t(j i'] per cent., so that what nutrient value they possess is derived mainly from the sugar, which amounts to from 50 to 80 per cent. W. W. O. Beveridge gives the following method for obtaining a closely approximate estimation of gelatine : Twenty-five grammes of the material containing gelatine are dissolved in hot water, and filtered if necessary to remove any insoluble matter. The solution is evaporated to a thick syrup on the water-bath in a platinum capsule, then removed and cooled ; 5 c.c. of a 10 per cent, solution of formaldehyde is then added, which renders the gelatine insoluble. Other proteins must not be present. The soluble matters, such as sugar, etc., are dissolved out by means of boiling water, when the gelatine remaining behind can be dried and weighed. CHAPTER IX ALCOHOLIC BEVERAGES The law allows a margin of 2 per cent, of proof-spirit before regarding a beverage as an alcoholic fluid . The estimation of alcohol in alcoholic beverages may be made as follows : Three hundred c.c. of the beverage (generally diluted) is placed in a retort and boiled gently until about 200 c.c. have distilled over into a flask. The distillate is next made up to the original bulk of 300 c.c. with distilled water, and the specific gravity is taken in a S.G. bottle, when the temperature has cooled to about 15° C. If this is 1,000 the fluid is free from alcohol, and the amount of alcohol which has distilled over will be great in pro- portion to the extent to which the S.G. falls below 1,000, since pure absolute alcohol at 15° C. has a S.G. of 793-8. To find the percentage amount of alcohol from the S.G. of the distillate, the alcohol table on pp. 320 and 321, in which these data are arranged side by side, may be consulted. In estimating alcohol in spirit, take 100 c.c. of the spirit, and dilute with 200 c.c. of distilled water. Before commencing the distillation of beer, it should be well shaken to expel as much carbonic acid as possible, passed through a coarse filter paper, diluted with an equal volume of water, and rendered alkaline with caustic soda; and as a further precaution against the beer frothing over, a small flame only should be applied to the flask and a little tannin powder may be added. An estimation may be made without distillation in the fol- lowing manner: Ascertain the specific gravity at 15° C. of the original liquid, then take 300 c.c. and boil down to about 100 c.c, thus driving off the alcohol; make up to the original bulk with distilled water, and again take the S.G. at 15° C. The difference between the first and second gravities deducted from 1,000 318 SPIRITS 319 (the gravity of water) gives the S.G. of the alcohol evaporated. If, for instance, the first S.G. of a sample of beer is 1,009 '"^^^^ the second S.G. is 1,018, then 1,018-1,009 = 9, '"^^^^ 1,000 -(j (/ji. A reference to the table on p. 320 will sliow that a liquid with a S.G. of 991 (or 0-991, as there expressed) contains 6-55 per cent, of alcohol by volume. Spirits. Spirits constitute the distillates from various liquids containing alcohol derived from grain. The expressions " over proof," " proof," and " under proof," are commonly employed to denote the amount of alcoliol in spirits. The above terms had their origin in a former practice of pouring the spirit over gunpowder, and applying a light to it. If the spirit burned without igniting the powder, owing to the large admixture oi water, it was " under proof " ; and the weakest spirit capable of firing the powder was called "proof." Such a spirit was stronger than the present " proof -spirit," for by " proof -spirit " is now implied a mixture of 57*06 per cent, by volume, or 49-24 per cent, by weight, of pure absolute alcohol in water, with a S.G. of 0-9198 at 15° C; and solutions weaker or stronger than this are " under " or " over " proof. The proportions of alcohol in alcoholic fluids may be stated as either a percentage of alcohol by weight, a percentage of alcohol by volume, or as a percentage of proof -spirit. The percentage of alcohol by weight may be obtained by multiplying the percentage by volume by 0-7938 and dividing the product by the specific gravity; and, conversely, the per- centage of alcohol by volume may be obtained by multiplying the percentage by weight by the specific gravity, and dividing the product by 0-7938. The percentage of proof -spirit may be obtained by multiplying the percentage of alcohol by volume by 175. A percentage of proof -spirit is always expressed by volume. The Sale of Food and Drugs Acts fix the following low limits of alcohol in " spirits." Brandy, whisky, and rum may be 25° "under proof," corresponding to 75 per cent, of proof -spirit — i.e., the S.G. may be as high as 0-9474 — and there may be only 42-8 per cent., by volume, of absolute alcohol. Gin may be 35° " under proof "■ — i.e., may only contain 37-1 per cent, by volume of absolute alcohol; and the S.G. may 320 LABORATORY WORK be as high as 09563, corresponding to 65 per cent, of proot- spirit. Suppose a sample of whisky is 44"" under proof; it therefore contains 100 - 44= 56 per cent, of proof-spirit. What percentage of spirit of 25° under proof does it contain ? A spirit of 25° under proof contains 75 per cent, of proof-spirit. Therefore the whisky contains -^-^^^^=74*6 per cent, of spirit of required 75 strength and 25 4 per cent, of added water. The percentage amount by weight of absolute alcohol generally present in spirits=32 to 50; wines=8 to 18 (about 10 per cent, in clarets); strong ales and porter=5 to 8; small beer=2 to 3. Lager beer contains more dextrine than English beer, and usually less alcohol. Short Alcohol Table. Specific Gravity at 15° C. >. it 3 rt >> Per Cent, under Proof. Specific Gravity at 15° c. Per Cent, of Alcohol by Volume. Per Cent, under Proof, I '000 0-00 lOO-OO 0-972 24-08 57-80 0-944 44-79 21-50 0-999 0-66 98-84 0-971 25-07 56-06 0-943 45-41 20-43 0-998 1-34 97-60 0-970 26-04 54-37 0-942 46-02 19-36 0-997 2-12 96-29 0-969 26-95 52-77 0-941 46-59 i8-:i6 0-996 2-86 95-00 o-96b 27-86 51-18 0-940 47-13 17-40 0-995 3-55 93-78 0-967 2S-77 49-60 0-939 47-67 16-46 0-994 4-27 92-50 0-960 29-67 48-00 0-938 48-21 15-50 0-993 5-00 91-23 0-965 30-57 46-44 0-937 48-75 14-57 0-992 578 89-87 0-964 31-40 44-97 0-936 49-29 13-63 0-991 6-55 88-50 0-963 32-19 43-60 0-935 49-81 12-70 0-990 7-32 87-16 0-962 32-98 42-20 0-934 50-31 11-82 0-989 8-i8 85-65 0-961 33-81 40-74 0-933 50-82 10-94 0-988 9.04 S4-15 0-960 34-54 39-47 0-932 51-32 10-05 0-987 9-86 82-70 0-959 35-28 38-18 0-931 51-82 9-20 0-986 IO-73 81-20 0-958 36-04 36-83 0-930 52-29 8-36 0-985 11-61 79-65 0-957 36-70 35-68 0-929 52-77 7-52 0-984 12-49 78-10 0-950 37-34 34-57 0-928 53-24 6-70 0-983 13-43 76-46 0-955 38-04 33-32 0-927 53-72 5-86 0-982 M-37 74-82 0-954 38-75 32-08 0-926 54-19 5-03 0-981 15-30 73-18 0-953 39-47 30-84 0-925 54-66 4-20 0-980 16-24 71-54 0-952 40-14 29-66 0-924 55-13 3-38 0-979 17-17 69-90 0-951 40-79 28-52 0-923 55-60 2-56 0-978 18-25 68-00 0-950 41-32 27-60 0-922 56-07 1-74 0-977 19-28 66-20 0-949 41-84 26-67 0-921 56-54 0-92 0-976 20-24 64-53 0-948 42-40 25-70 0-920 56-98 0-14 0-975 21-19 62-87 0-947 42-95 24-74 0-9198 57-06 -Proof 0-974 22-18 61-13 0-946 43-56 23-66 0-973 23-10 59-52 0-945 44-18 22-58 SPIRITS 321 Short Alcohol Table — continued. "o >> w "0 >■ "o >> rt ■ XI .: c's ^ .-^ »i tio « •■° V - u (L) a> ^% 0-68 Specifi Gravity 15° c. Per Cent Alcohol Volume C V Theobromine, 348-349 Thorpe's method of estimating COo in water, 102-103 Thread-worm, the, 72, 74 Thresh's test for dissolved oxj'gen in water, 104-108 Tidy's process in water analysis, 86- 90 Tilletia caries, 269; T. IcBvis, 269 Tin in water, 46, 49; in food, 361- 362 Tinned provisions, 358-365 Total solids in water, 60-03; volatile and non-volatile, 61-62 Trichina spiralis, 301-302 Trichocephalus dispar, 72-74 Turmeric, 244 Tyrotoxicon, 308-309 Ulva latissima, 131, 154-155 Upland surface water, iio-iii Uredofcelida, 269; U. segetum, 269 Uroglena in water, 29 Valenta test for butter, 260 Vanillin, 308 Vibriones in wheat, 266-267 Vieth's ratio, 226 Vinegar, 332-334 Vogel's test for CO in air, 198-199 Volatile disinfectants, test of, 411 Wanklyn, Chapman, and Hall pro- cess, 76-85 Water, collection of samples of, 19- 21 ; information required as to samples, 21-22; report on, 22-23; opinion on, 11 3-1 29; physical char- acters, 25-30; clearness, 25; colour. 25-27; taste, 27; odour, 27-29; aeration, 29; reaction, 29-30; tem- perature, 30; chlorine, 31-36; hard- ness, 37-43; poisonous metals, 44- 53; non-poisonous metals, 54-56; sulphates, 57-58; phosphates, 58- 59; solid residue, 60-63; suspended and deposited matter, 30, 64-74; orgixnic matter, 75-76; Wanklyn's process, 76-S5; oxidizable organic matter, 86-90; Frankland's pro- cess, 90-91 ; oxidized nitrogen, 92- 100; gases in, 101-109; water from various sources, 110-113; opinion on water samples, 11 3- 129; water standards, 114-116, 123-129; "bac- teriological evidence, 123-129; scheme for water analysis, 140-141 Water-bath, 5-6 Weevil, 266 Weighing, operation of, 3-5 Weights and measures, 16-17 Well-water, deep, 111-113, liS, 120 Werncr-Schmidt process, 234-236 Westphal balance, 8-9 Wheat-flour, composition of, 270-271 ; analysis, 271-275; adulteration, 275-277; microscopical characters, 287, 290 Whip-worm, the, 72, 74 Whisky, 319. 322 Wiley's experiments on chemical antiseptics, 368-369 Wills' combined water-bath and drying -oven, 2 Wine, 325-329 Winkler's process for dissolved oxy- gen in water, 108-109 Wood acid, 332 Wynter Blyth's tube for sediments, 64-65 Zinc in water, 45-46, 49, 51-52 H. K. LEWIS, 136, GOWF.R STREET, 1.0NU0N ^ i ■' '..in; •.'.•'■< mm tlMii_Bc'aU\\ Ud\ioru'(or\i % Kv \'^\^ J^^ \ox\K