THE COMPOSITION OF EXPIRED AIE AND ITS EFFECTS UPON ANIMAL LIFE.' ATLSTEACT OF A REPORT i>N THS llESULTS i)F AN" UfVESTR". ATIOS MADE FOR 1 in: SMlTflSONIAX INSTITUTION UNDER THE PROVISIONS OF THE HODCiKINS FUND. By J. 8. Billings, M. D.. S. Weie Mitchell. M. D., and r>. IL BsR.xKY, ivr. D. ALBERT R. MANN UBIRAIKY Cornell University Alumni Book Endowment FOR Mann Library Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924090242615 THE COMPOSITION OF EXPIEED AIR AND ITS EFFECTS UPON ANIMAL LIFE.i ABSTEA.CT OF A EEPOET ON" THE EESTJLTS OF AN INVESTIGATION MADE FOR THE SMITHSONIAN INSTITUTION" TTNDEK THE PEG VISIONS OF THE H0DG"B:INS FUND. By J. S. Billings, M. D., S. Weir Mitchell, M. I)., aud T). H. Bebgey, M. I). In May, 1893, a grant was made from, the Hodgkins fund to Drs. John S. Billings and S. Weir Mitchell "for the purpose of conducting an investigation into the nature of the peculiar substances of organic origin contained in the air expired by human beings, with special ref- erence to the practical application of the results obtained to problems of ventilation for inhabited rooms." For a number of years prior to 1888 the prevailing view among phy- sicians and sanitarians had been that the discomfort and dangers to health and life which had been known to exist, sometimes, at least, in unventilated rooms occupied by a number of human beings, were largely or entirely due to peculiar organic matter contained in the air expired by these persons, and that the increase in carbonic acid due to respiration had but little effect in producing these results, its chief importance being that it furnished a convenient means of determining the amount of vitiation of the air. Recently, however, several experi- menters have concluded that the organic matters in the exhaled breath are not harmful, at all events to animals, and the main object of the proposed investigation was to determine the correctness of these .conclusions. The effects produced on animals and men by an atmosphere contami- nated with their exhalations, and with particulate matters derived from their bodies or their immediate surroundings, may be divided into acute and chronic. The acute effect may be death in a few minutes or hours, as shown by the results observed in the Black Hole of Calcutta, in the steamer Londonderry, and in many of the experiments referred to iu this report, or it may be simply great discomfort, especially in those unaccustomed to such conditions. 'Tlie full report is printed in Smithsonian Contributions to Knowledge, Vol. XXIX (No. 980), 4to, pp. 81. 389 390 THE COMPOSITION OF EXPIRED AIR. The chronic effects inchide the favoring of the action of certain specific causes of disease commonly Icnown as contagious, if tliese are present, and perhaps also a general lowering of vitality. The statistical evidence collected by the Bnglisli Barrack and Hos- pital Commission (1)' as to the effects of insufficient ventilation upon the health of soldiers in barracks, published in 1861, showed that men who live for a considerable portion of their time in badly ventilated rooms have higher sickness and death rates than have those who occupy well-ventilated rooms, other conditions being the same; and this has also been found to be true with regard to monkeys aiid other animals. It is evident, however, that in a room occupied by animals or men there are many sources of impurity besides the exhaled breath, and it is still a question whether the expired air contains substances injurious to life, excluding carbonic acid. The widely divergent results obtained and conclusions reached by different investigators during the last ten years as to whether the exhaled breath of men and animals contains a peculiar volatile organic poison, have made it desirable to repeat and vary such experiments in order, if possible, to settle this important point. The chemical analyses of the air of overcrowded rooms, and the experiments upon animals with various proportions of carbonic acid, made by many investigators, indicate that the evil effects observed are probably not due to the com- paratively small proportions of carbonic acid found under such circum- stances. Claude Bernard (2), in 1857, experimented with animals confined in atmospheric air and in mixtures both richer and poorer in oxygen than atmospheric air. A small bird placed in a bell glass of a little more than 2 liters' capacity, containing a mixture of 13 per cent carbonic acid, 39 per cent oxygen, and 48 per cent of nitrogen, died in two and one-half hours. He demonstrated that carbonic acid is not poisonous when injected under the skin of animals — as much as one liter injected under the skin of a rabbit producing no ill effects. No ill eff'ects fol- lowed the injection of the gas into the jugular vein and into the carotid artery. An atmosphere of equal parts of oxygen and nitrogen had no effect upon an animal confined in it, while an atmosphere composed of equal parts of carbonic acid and of oxygen produced immediate death in the animal placed in it. He explains the poisonous eff'ects of car- bonic acid when respired to be due to the fact that it deprives the animal of oxygen. Similar results were reported by Valentin (3) and by Paul Bert (4). Eichardson, in 1860-61 (5), found that a temperature much higher or lower than 20° C. had the effect of shortening very considerably the lives of animals confined in an unventilated jar, and that these effects were more marked when the animals were confined in an atmosphere 'The mimbors in parentheses refer to the bibliographical list appended to this report. THE COMPOSITION OF EXPIRED AIK, 391 richer in oxygen than air, in which case he found that by passing electric sparks from a frictional machine through the fatal air (having previously deprived it of its carbonic acid) it was again made capable of supporting life, from which he concluded that the oxygen is " devi- talized" during respiration, and that the electric spark has the faculty of revitalizing it. Vou Pettenkofer, in 1860-1863 (6), showed that the symptoms observed in crowded, ill-ventilated places were not produced by the excess of carbonic acid, nor by a decrease in the proportion of oxygen in the air; neither of these being sufficient in our dwellings, theaters, etc., to pro- duce toxic effects. He did not believe that the impure air of dwellings was directly capable of originating specific diseases, or that it was really a poison in the ordinary sense of the term, but that it diminished the capability of withstanding the influence of disease-producing agen- cies on the part of those continually breathing such air, and laid down the rule, which has been accepted and taught by sanitarians for thirty- five years, that the proportion of carbonic acid in the atmosphere of inhabited places affords a safe indication as to the amount of the other impurities resulting from respiration and other exhalations from the bodies of the occupants. Hammond, in 1863 (7), reported experiments in which he sought to remove the carbonic acid and moisture, and to supply fresh air as fast as it is needed to take the place of the carbonic acid removed, thus leaving the "organic matter" to accumulate in the vessel. For this purpose he confined a mouse in a large jar, in which were several sponges saturated with baryta water, by which the carbonic acid was removed as fast as formed. Fresh air was supplied as fast as required by means of a tube communicating with the bell jar and closed by water in the bend of the tube, which acted as a valve. As the air in the bell glass was rarefied by respiration and absorption of the car- bonic acid, fresh air flowed in from without, while the arrangement of the tube prevented the air of the bell glass from passing out. The watery vapor exhaled by the animal was absorbed by two or three small pieces of chloride of calcium. The mouse died in forty minutes. The observation was repeated many times, and death ensued invariably in less than an hour. On causing the vitiated air to pass through a solution of permanganate of potash the presence of organic matters in large quantity was demonstrated. Eansome, in 1870 (8), reported a series of very interesting investi- gations upon " Organic matter of human breath in health and disease." By condensing the aqueous vapor of the human breath and analyzing it by the Wanklyn and Chapman method, he found that "in ordinary respiration about 0.2 gram of organic matter is given off from a healthy man's lungs in twenty-four hours," while in the air expired by persons affected with certain diseases, he found great variations in the amount of organic matter, the amount being greatest in a case of phthisis com- plicated with Bright's disease. 392 THE COMPOSITION OF EXPIRED AIR. Smith (9) employed a lead chamber in his investigatioDS upon the question whether human lungs give off' any poisonous agent other than carbonic acid. He found the pulse to fall from 73 to 57 beats per minute, and the number of respirations to rise from 15.5 to 21, as the carbonic acid in the atmosphere increased from 0.04 to 1.73 per cent during four hours. When the proportion of carbonic acid rose to 3 per cent there appeared great weakness of the circulation with slowing of the heart's action, and great difficulty in respiration. He believed that these results should be attributed to other conditions rather than to the excess of carbonic acid, because he found later that it was only when lamps became dim iu aTi atmosphere — indicating a proportion of about 10 per cent of carbonic acid j)resent — that the respiration became difficult. Seegen and Xowak, in 1879 (10), believed they had demonstrated the presence of poisonous organic matter in the expired breath by passing it over red-hot cupric oxide, but the quantity found was so small that they failed to determine its exact nature and properties. Hermans, in 1883 (11), was unable to detect any organic matter in the atmosphere of a tin cage in which several persons had been confined for a number of hours, and found that an atmosphere containing from 2 to 4 per cent of carbonic acid and 15 per cent of oxygen was not toxic. Brown-Sequard and d'Arsonval, in 1887 (12), reported that the air expired by men and dogs in a state of health has the power of produc- ing toxic ])henomena, citing three seriesof experiments on rabbits where such phenomena were observed. In the first series they injected into the vascular system of a rabbit 4 to G c. c. of water obtained by inject- ing from 15 to 25 c. c. of pure filtered water into the trachea of a dog. In a second series from 6 to 7 c. c. of a liquid obtained by condensing the moisture in the exhaled breath of a man were injected into the aorta, or into a vein, of a rabbit. In the third series from 4 to C c. c. of a liquid obtained by condensing the moisture in the exhaled breath of a trache- otomized dog were used. The condensed liquid thus obtained was fil- tered and then injected either into the jugular vein or the carotid artery. The symptoms observed were dilatation of the pupils, increase of the heart beat to 240, 280, or even 320 per minute, lasting for several days or even weeks. The temperature remained normal, the resjjiratory movements were generally slowed, and usually there was observed par- alysis of the posterior members. Choleraic diarrhea was invariably present. Death usually took place in a few days, or at the farthest in four or five weelcs. As a I'ule, it appeared that larger doses caused labored respiration, violent retching, and contracted pupils. A rapid lowering of temperature, 0.5° to 5° 0., was sometimes observed. The appearances that presented post-mortem were much like those observed in cardiac syncope. They believed they had discovered a volatile organic poison in the exhaled breath and the moisture condensed from it. This poison they THE COMPOSITION OF EXPIRED AIR. 393 believed to be of tlie nature of au organic alkaloid, or a ptomaine not unlike Brieger's ptomaine [12^]. Ill further reports, in 1888 (12''), they state that none of eleven rabbits in which the condensed pulmonary vapor had been injected into the vascular system in doses of 12 to 30 c. c. survived, but of eight rabbits receiving an injection of from 4 to 8 c. c. three were living after the lapse of from four to live weeks, but were then weak. When the fluid was injected under the skin of the thorax and in the axilla, live out of seven rabbits died rapidly. The results were much the same as when it was injected into the blood. The quantity of the condensed liquid injected in these seven was 20 c. c. in one case, 25 c. c. in three cases, 31 c. c. in one case, 40 c. c. in one case, and 44 c. c. in another case. After death, considerable congestion of the viscera was noted, espe- cially of the lungs. No appearance of embolism was noted. The brain and its membranes were congested, but without visible lesion. The condensed liquid turns concentrated sulphuric acid yellow. The poison is reduced by ammoniacal nitrate of silver solution as well as by chlo- ride of gold. After boiling in a close vessel it is still toxic, showing that the poison is not a microorganism. The boiled lung liquid poisons with more rapidity than that which has not been sterilized, and may kill a pigeon and a guinea pig as well as a rabbit ; that it may kill by being injected into the rectum or into the stomach; that a guinea pig two months old was killed within twelve hours by an injection of 3 c. c. into the peritoneal cavity. If injected into the lungs this liquid pro- duces rapid congestion followed by true inflammation and red hepati- zation. In an experiment with two dogs it was arranged that one breathed ordinary air and the second inhaled air which came from the lungs of the other. The dogs were of the same weight, 15 kilograms. The experiment continued for six hours and forty minutes. No appreciable or immediate consecutive accidents were produced. In a second experiment the pulmonary liquid was collected from dogs through a tracheotomy tube, to exclude impurities furnished by the mouth. The air inhaled was first washed to remove dust. The mois- ture in the air expired was condensed, and the liquid collected in a flask surrounded by ice. At the moment of injection this liquid was filtered, and was then injected at the temperature of the laboratory, about 12° 0. If the animal was kept immovable from twelve to sixteen hours, inflammation of the air passages was ijroduced. The liquid of the first hours came from a thoroughly sound lung, and iu the later hours from a diseased lung. The two were collected separately and tried sepa- rately. For 1 kilogram of the animal, for each hour, the mean quantity of fluid obtained was 0.38 gram, A^arying from 0.28 to 0.48 gram. It was greater in the beginning and lessened the longer the animal was kept in a; fixed position. It was injected into the marginal vein of the ear of a rabbit by means of a syringe, 75 c. c. being injected. When the 394 THE COMPOSITION OF EXPIRED AIR. iuji'ction did not exceed 40 to 50 c. c. the time occupied by the injection was from six to fifteen minutes. Experiments made by injections uj)on the dog' were negative without excej)tion. Experiments made upon the rabbit pioduced lesions, but the relation between these and the injections is uncertain. Dastre and Loye, in 1888 (13), reported that they had exposed one dog to the expired breath of another for six hours without noting any effects. They inoculated animals with the condensed moisture of respiration, as follows: Animals. Quantity of fluid. KesuUs. Cubic centi- meters. 33 to 75- 5 to. 7 30 lo 53 2 to 3 SO to 190 a 30 Negative. Do, Do. Do. Died. Do. Two guinea piga (each) T wo rabbits (each) ... a Of water. They found that 50 to 70 c. c. of the condensed fluid of respiration (20 to o5 c. c. per kilogram) could be injected into the veins of the ear of a dog without producing any of the symptoms reported by Brown- S6quard and d'Arsonval. They observed one death during the injec- tion of 190 c. c. (60 c. c. per kilogram), yet by control experiments with water they obtained a more remarkable result — a rapid deatb from the injection of 30 c. c. of distilled water (25 c. c. per kilogram). linssoGiliberti and Aiessi, in 1888 (14), reported experiments con- firming the results obtained by Dastre and Loye. Brown-Sequard and d'Arsonval, in 1889 (15), reported a new form of experiment, by means of wMch they obtained additional evidence in support of their former statements. The new form of experiment consisted in confining animals (rabbits) in a series of metallic cages con- nected by means of rubber tubing, through which a constant current of air is aspirated. The animal in the last cage of the series receives air that lias traversed the entire series of cages, and is loaded with the impurities from the lungs of the animals in the other cages. This animal succumbs, after a time, to the atmospheric conditions present. After another interval of some hours the animal in the next to the last cage also dies, the first and second animals usually remaining alive. They could not attribute the death of these animals to excess of car- bonic acid in the atmosphere of the cages, because they rarely found more than 3 per cent of this gas in the last jar with small animals, or 6 i)er cent with larger animals. On placing absorption tubes contain- ing concentrated II2SO.1 between the last two cages, the animal in the, last cage remained alive, while that in the cage before it was the first to die. They concluded from these facts tliat the death of the animals THE COMPOSITION OF EXPIRED AIR. 395 was produced by a volatile poison, wliicli poison is absorbed by tlie il2SO„ wliicli thus saves tlie life of the animal in the last cage. They stated (IG) that any alkali used to absorb carbonic acid from expired air would also cliauge the organic poison, and proposed an aj)paratus by means of wliich the organic poison should be supplied to tlie fresh air entering the jars by volatilizing it from fluid condensed from the expired air. V. Hofmann-Wellenhof, in 1888 (17), found that when he injected large quantities of the condensed fluid of respiration at 12° G., instead of at 37° C. — intravenous injection — a resemblance of the results obtained by Browu-Sequard and d'Arsonval was produced. Under such cii'cumstances he observed muscle wisakness, slowing of respira- tion, fall of temperature, and dilatation of the pupils, though the ani- mals remained alive. He injected ten rabbits with 6 to 30 c. c. of the fluid warmed to the body temperature, all the results being negative. Three other animals were injected in the jugular vein, one receiving 28 c. c. of the fluid, another 25 c. c. of distilled water, and a third 50 c. c. of distilled water. There was no difference in the symptoms noted in the animals. He noticed symptoms of depression only after injecting 50 c. c, or more, of the fluid. In a series of seventeen experiments witli inoculations of from 30 to 50 c. c. each of the fluid, in twelve there appeared h;emoglobinuria; six of these died. As the result of his experiments, he concluded that the existence of a volatile poison in the expired air of healthy human beings has not been demonstrated by his experiments, this being a direct contradiction of the results of Brown-Sequard and d'Arsonval, as were also those of Dastre and Loye. Uffelmann, in 1888 (18), found that there was a perceptible increase in organic matter in the atmosphere of a sleeping room occupied by several j)ersons for some hours, increasing in amount with the length of time the room was occupied. Lehmann and Jessen, in 1890 (19), collected 15 to 20 c. c. of the con- densed fluid per hour exhaled from the breath of a person breathing through a glass spiral laid in ice. The fluid was always clear as water, odorless, and of neutral reaction. Nessler's reagent showed the pres- ence of ammonia constantly, with good teeth but little, sometimes merely a trace, with bad teeth, more, though never more than 10 milligrams of NH4CI in 1 liter. Traces of HOI were also constantlj^ found. A small sediment remained on evaporation, ranging from 39 to 86.4 milligrams per liter of fluid. This they believed to originate from the glass ves- sel; probably calcium oxalate. They tested its reducing power upon solution of permanganate of potash, making two control determinations- The first determination showed 3.6 milligrams of O for the oxidation of 1 L. ; the second, 4.2 milligrams of O. They were unable to obtain any alkaloid reaction in the condensed fluid, or in its distillates, by means of PtOl4, AuOls, KOdI, KBil, KI, Bouchardet's reagent, KjCrOo, picric 396 THE COMPOSITION OF EXPIRED AIR. acid, metawolframic acid, or phosphowolframic acid. Only sublimate gave at times an opalescence which, like the yellow coloration of the Nessler reagent, pointed to traces of IfHj. Neither could they succeed, according to the method of Wiirtz, in obtaining a lime or oxalic acid- free filtrate. The ammoniacal silver solution, according to Brown- S^quard and d'Arsonval's method, failed to give the desired reaction, remaining clear. They confined a man, clothed in his working clothes, in a zinc cage for about one-half an hour, then allowed a boy and girl to inhale the air from the cage. No ill effects, except increase of respi- rations to 30 and 40 per minute, were noticeable. They had complete negative results from inoculations of condensed fluid into animals. Lipari and Crisafulh, in 1889-90 (20), reported results which were in accord with those of Dastre and Loye and directly opposed to those of Brown-Sequard and d'Arsonval. They could find no organic principle l)ossessing toxic properties in the expired breath of healthy persons. Margouty, in 1891 (21), reported the results of experiments similar to those of Hammond, and also of experiments in injecting fluid condensed from expired air into animals. His results did not correspond to those reported by Hammond, and there was no evidence of toxic; properties in the injected fluids. . Haldane and Smith, in 1892 (22), published an account of experiments in which an air-tight chamber, 6 feet 2 inches high, 2 feet 11 inches wide, and 3 feet 11 inches long, was employed. Samples of air for analysis were drawn off through a tube placed in the wall of the chamber, about 3 feet from the floor. When one person remained in this chamber until the vitiation was from ten to twenty times as great as in the most crowded and worst ventilated public buildings, there was no percept- ible odor or sense of oppression. Air vitiated to such an extent as to completely prevent a match from burning had no appreciable eSect upon the subject of the experiment. In other experiments hyperpnoea and other phenomena produced were apparently due to the increased proportion of carbonic acid. With rabbits weighing 1,800 grams, hematuria was produced when the amount of boiled distilled water injected passed beyond 100 c. c, and therefore 80 c. c. were taken as the maximum dose. To obtain the condensed liquid from the lungs, a man expired through a Liebig condenser, in the jacket of which was flowing a stream of ice cold water. The condensation liquid was collected in a flask, the bulb of which was buried in ice; and when the required amount (80 c. c.) liad been obtained, it was at once injected into the subcutaneous tissue of the back. Six rabbits were thus injected, each with 80 c. c. of the fluid, with no evident disturbance of health in any of them; 80 c. c. to a rabbit corresponds to a dose of about 3 liters to a man. They also repeated the experiments of Brown-S6quard and d'Arsonval in supplying to the animals air charged with organic mat- ter drawn directly from the lungs of other animals. Two large rabbits THE COMPOSITION OF EXPIRED AIR. 397 were placed in an air-tight chamber and a current of air drawn through this was supi>lied to two young rabbits under observation; no effect was produced. Merkel, in 1892 (23), reported an experiment in which four air-tight glass vessels, of IJ liters' capacity, were connected by means of glass tubes, a mouse being placed in each vessel. Between the third and fourth vessels a Geissler absorption tube containing sulphuric acid was interposed. Air was now drawn slowly through the vessels by means of an aspirator, so that the second mouse breathed the air from the first, the third from that of the second, etc. The result was, just as in the experiment of Brown-S^quard and d'Arsonval, that the mouse in the third vessel died first, after sixteen to twenty hours, while that in the fourth vessel remained alive. The conclusion is drawn that, as the fourth mouse remained alive, the death of the third can not have been due to excess of carbonic acid, or deficiency of oxygen in the air, but must have been caused by the presence of some volatile substance which is absorbed or destroyed by sulphuric acid. The symptoms presented by the mice before death were at first rest- lessness and gradually increasing acceleration of respiration, afterwards slowing of respiration, and finally spasmodic, deep respirations, becom- ing constantly less frequent until the advent of death. The proportion of carbonic acid in the air led through the glass vessels was not poison- ous; it amounted in the highest case to 1.5 per cent. He concludes that the expired breath of healthy persons contains a volatile poison in extremely small quantities, being probably a base which is poisonous in its gaseous state, but loses its toxicity after combination with acids. His belief in the toxicity of the organic matter contained in the expired breath of human beings is based solely upon the results he obtained in the Brown-Sequard and d'Arsonval experiment. Haldane and Smith, in 1893 (24), repeated the Brown-S6quard ex- periment, using five bottles, each of a capacity of 1 to IJ liters, con- nected by means of tubes. A mouse was placed in each bottle, and ventilation established through the whole system by means of a filter pump, a small gas-meter being placed between the last bottle and the pump. Specimens of air leaving the last bottle were drawn off at intervals for analysis. Full-grown mice were used. The mice in the last two bottles were exposed to the full effect of the vitiated air for fifty-three hours without detriment. In a second experiment an absorption tube containing pumice stone saturated with sulphuric acid was placed between the last two bottles. This experiment was continued for thirty hours; no serious effects were observed. The amount of ventilation furnished was from 15 to 24 liters per hour. The mice remained normal after having been in the bottle three days, and the percentage of carbonic acid in the last bottle bad varied from 2.4 to 5,2, averaging about 3 per cent. 398 THE COMPOSITION OF EXPIRED AIR. Tliey state that these experiments, like their former ones on rabbits and man, are distinctly against the theory that a volatile poison other than carbonic acid exists in the expired air. Ben, in 1893 (:i5), reported the results of experiments, made under the direction of Uffelmann, in which the condensed moisture of expired air was collected by the methods usually employed, taking the precau- tion to cleanse his apparatus with solution of KMnOi and distilled water, and likewise sterilizing the apparatus before it was brought into use. The saliva is collected in a Woulff bottle attached before the condenser. The amount of air expired, measured by a gas meter, was found to be 3,000 liters in eight hours, from which he collected 100 c. c. of fluid. A distinct ammonia reaction was obtained upon the addition of Xessler's reagent. Nitrate of silver failed to show the presence of chlorine. Its reducing power upon solution of permanganate of potash showed 50 milligrams of oxygen necessary to oxidize 1 liter of fluid, or 15 mil- ligrams in twenly-four hours, which denotes 0.0017 milligrams per liter of expired air. The alkaloid reaction with AuCls, KI, phosphomolyb- date of potash, gave negative results. He expired 500 liters through 150 c. c. of a 1 per cent solution of HCl, then evaporating to dryness on the water bath. A yellowish- brown deposit remained. This deposit, dissolved in distilled water, formed a fatty layer on the surface of the slightly yellow fluid. The whole quantity, 1.5 grams, was warmed to the body temperature and injected under the skin of the back of a white mouse without produc- ing observable symptoms. This fluid had a distinct odor not compar- able to anything. lie next confined a mouse in a sealed glass vessel, having a globe attached, with potash solution to absorb the carbonic acid; 3,200 exi)i- rations of air were conducted into the glass vessel during the three hours; no effect noticeable. In a second experiment the carbonic acid was not absorbed, the experiment lasting four hours; no effect. He repeated the Brown-Sequard experiment, using Avhite mice in four glass cages. The death of the animals, he believes, was due to chanj;es in the temperature and the accumulation of moisture in the jars. He believes the protection afforded by H2SO4 in Brown-Sequard and d'Arsonval's experiments was due to its abstraction of the mois- ture from the air. An acute poisoning through the organic matters contained in the expired air he believes to be impossible, at least as not shown by anything in his experiments. llauer, in 1.S93 (26), used white mice confined in glass vessels of about li liters' capacity, the bottom of which was covered with oats. The cork was perforated by three tubes. One of these x)assed down near the bottom of the vessel and served for the entrance of air; the second terminated just below the cork and served for the exit of air, and the third extended down to about tlie height of the animal, but THE COMPOSITION OF EXPIRED AIR. 399 was usually closed; this was only used for the removal of air for its chemical examination. In the beginning thermometers and hygrome- ters were used in the vessels, but they were found to be unimportant and were abandoned. The whole apparatus was connected with a large aspirator. In an experiment with five animals and a ventilation of 4 liters per hour, the carbonic acid was found to amount to 9.3 per cent after five hours. In another experiment with six animals and with a ventilation of 2J liters per houi', he inserted four absorption tubes, with soda lime between the last two jars, and a Geissler tube containing concentrated H2SO4 between the fourth and fifth. The sixth animal remained alive, while the fifth died earlier than the fifth animal in the first experi- ment. He concludes that there is no organic poison in expired air, death being due to the excess of carbonic acid in the atmospheres of the jars. Sanfelice, in 1893 (27), reported that he had repeated the Ham- mond experiment, using a flask of about 5 liters' capacity, the animal dying in six or seven hours. He is undecided as to the existence of a volatile expiratory poison, though he thinks that other factors, for instance, heat radiation, have an important influence upon the results. Liibbert and Peters, in 1894 (28), reported that they had repeated the Brown-Sequard experiment, placing a guinea i^ig in each of a series of four flasks. Between the third and fourth flasks they placed a combustion tube through which the air coming from the third flask was conducted, passing over red-hot cupric oxide, to remove the organic matter. Before reaching the fourth flask, the air was again cooled by conducting it through a cylinder surrounded with ice. In this manner all moisture contained in the air was condensed. From this cylinder the air passes through a series of twelve U-tubes, each made from a piece of tubing 80 centimeters in length and of 2 milliineters internal diameter. During its iDassage through these U-tubes the air assumed a tempera- ture of about I80 0. as it entered the fourth flask. The results obtained by this arrangement substantiated the conclusions they had formed from conducting the experiment in the ordinary manner, that the cause of death was traceable to the high per cent of carbonic acid. Tlie removal of the organic matter by combustion failed to save the life of the animal in the last jar when the carbonic acid had increased to 11 or 12 per cent. After the absorption of the carbonic acid by means of soda lime the last animal remained alive. They conclude, therefore, that the poison- ous expiratory poison of Brown-Sequard and d'Arsonval does not exist, but that death is produced by the excess of carbonic acid in the flasks. Brown-Sequard and d'Arsonval, in 1894 (29), reported further exper- iments, and at the same time gave fuller details as to all their experi- ments and the apparatus employed. They had inoculated over 100 animals Math the condensed fluid of respiration and believed in the trnth of their former statements as firmly as ever. They could not 400 THE COMPOSITION OF EXPIRED AIR. understand tlie failures on the part of tlie other experimenters. They emphatically reaffirm that the expired breath of man and animals con- tains a volatile organic poison producing the results reported bj' them, and that these results are not produced by excess of carbonic acid or deficiency of oxygen in the air. From the foregoing summary of the reports of different experimenteis, it will be seen that widely different results have been reported by them, but that the maiority of the later investigators agree in denying that the exhaled breath of healthy human beings or of animals contaius a poisonous organic alkaloid, or any poisonous product other than car- bonic acid, yet in any case positive results require aii explanation which shall account for the facts. DR. BEBGET'S experiments. The first experiments made by Dr. Bergey were to ascertain whether the condensed moisture of air expired by man in ordinary, quiet respi- ration contaius any particulate organic matters, such as microorgan- isms, epithelial scales, etc. The test for microorganisms was made by having an adult man expire for from twenty to thirty minutes through sterilized melted gelatin, which was then preserved as a culture for from twenty to thirty days. In the first trial, six, and in the second, two colonies of common air organisms developed; but when special care was taken to thoroughly sterilize the vessels used, the result was that in two consecutive trials the gelatin remained sterile. Eiiithelial scales and other particulate matters were sought for by condensing the vapor of the exhaled breath and examining the product with the micro- scope, with and without the use of stains. In six preparations thus examined no bacteria or epithelial cells were found. This result was to be expected, since neither bacteria nor wetted particles pass into the air from the surface of fluids, or from moist surfaces, unless the air currents are sufficiently i^owerful to take up particles of the liquid itself in the form of spray. Abbott (30), in his paper on sewer gas, reports some experiments made to determin e the possibility of conveyin g microorganisms from liquid cul- ture media by means of a current of air bubbling through such media; also by means of ordinary baker's yeast inoculated into media contain- ing from 4 to 5 i^er cent of glucose. No bacteria were carried from the culture by the exploding air bubbles ])roduced by the yeast, but a cur. rent of air equal to 3^ liters in six hours, bubbling through a li(iuid culture, carried with it some of the organisms in the culture. The determinations of ammonia in the condensed fluid of expired air, the estimation of its reducing power upon solution of permanganate of potash, and its reaction with various reagents, were made with fluids collected from a healthy man, from a man with a tracheal fistula follow- ing excision of the larynx, the expired air not coming in contact with the mouth or the pharjnrx, and from a man suffering from well-marked THE COMPOSITION OF EXPIRED AIR. 401 tuberculosis of the luugs. In eacli case the amount of ammonia and of albuminoid ammonia in the fluid was very small, the average being, in grams per liter of fluid : Healthy man ^tan with tracheal iistula Cousumptive Free am- monia. 0.019 . 00046 .003 Albuminoid ammonia. 0.081 . 00030 .0034 The oxidizabk' matter in these fluids, as shown by their reducing power on a solution of permanganate of i)otash, was determined. The average results, stated in milligrams of oxygen consumed per liter of condensed fluid, are as follows : Healthy man, 10.72 ; man with tracheal fistula, 13.49; consumptive, 19.34. The high average for the man with the tracheal fistula is dxie to a single observation, for which the figure was 24.916. Omitting this, the average for the three other observations would be 9. 08. The average for five specimens of fluid condensed from the exxnred air of a healthy man four hours after he had taken a meal was 11.98, while the average for six specimens from the breath of the same man half an hour after the meal was only 3.86. For two specimens from the same man collected three and a half and four hours after a meal, but just after the mouth had been thoroughly rinsed with warm water, the average was 2.40. These results indicate that the ammonia and oxi dizable organic matter in the condensed fluid were, to a large extent, due to i>roducts of decomposition of organic matters in the mouth. The well-known fact that the amount of oxygen absorbed and of carbonic acid given off varies according to whether the person is fasting or has recently taken a meal, may possibly be in part due to the same cause, but the results obtained by Birkholz (31) indicate that it can only be in part. Eansome (8) reports no mai'ked difference in the amount of ammonia, or of oxidizable organic matter, as determined by the per- manganate test, contained in the fluids collected from the exhaled breath soon after a meal and in that collected from a fasting person. Ben (25) found a much higher proportion of oxidizable matter in the fluid condensed from his own breath (50 milligrams of oxygen required per liter of fluid) than was found in Dr. Bergey's experiments. His results indicated the exhalation of 15 milligrams of organic matter in twenty-four hours, the corresponding figure from Eansome's results being 20 milligrams. About 12 c. c. of fluid was collected from about 335 liters of air expired per hour, being nearly equal to the results obtained by Beu (25), who condensed 100 c. c. of the fluid from 3 cubic meters of air expired in eight hours. Renk (32, p. 162) gives a table showing that in an average quantity of 9,000 liters of air expired in a day by a healthy man, the amount of moisture may be from 200 to 400 grams, depending on the temperature SM 95 26 402 THE COMPOSITION OF EXPIRED AIR. and relative hmnidity of the inspired mv. With air containing 50 per cent of moisture inspired at 25*-' C, the amount of moisture is -!!)3 grams, or about tlie result given by Beu, referred to above. Lehmann and Jessen (19) found that between 3 and 4 milligrams of oxygen were required in 1 liter of fluid to effect oxidation, and note that more ammonia was present in the fluid collected from a person with decayed teeth than in that obtained from a person whose teet'i were sound. The very considerable diff'erences in the amounts of ammonia and of oxidizablc matter found in the fluid condensed from expired air by different experimenters, and by the same experimenter, in fluids obtained from the same person at different times, are probably due to several different causes and their combinations. The amount of fluid condensed per liter of expired air varies from 0.003 to 0.004 c. c. The soundness and cleanliness of the mouth and teeth influence the amount of ammonia and oxidizable matter expired. Variations in the amount of organic matter contained in the inhaled air may possibly influence the result, but this influence must be slight. Ransoiiic's results indicate that the age, health, and vigor of the person may affect the amount of organic matter exhaled, and Dr. Bergey's experiments with the fluid obtained from the consumptive patient show that a smaller proportion of ammonia and a larger amount of oxidizable matter were present in it than in the fluid collected from a healthy man. It should be remembered, also, that it is extremely diflBcult to obtain accurate results in quantitative determinations of such very minute amounts of ammonia and oxidizable matters as are found in expired air, and a part of the differences in results obtained is no doubt due to unnoted diff'erences in the details of the experiment. The results of tests for the presence of an organic alkaloid in the condensed fluids obtained by Dr. Bergey were negative, corresponding to those reported by I^ehmann and Jessen (19) and by Beu (25). The results of attempts to condense the moisture of the air in the hospital ward were not satisfactory, and the determinations of ammonia in the fluid obtained are not comparable, except they show that the placing of a dust filter in front of the condensing apparatus causes a marked reduction in the proportion of ammonia in the condensed fluid. The evaporation equaled the condensation except on days when the external air was saturated with moisture, hence no moisture was col- lected on clear days, but on such days some dust particles may have accumulated in the apparatus which had no filter. Several series of experiments were made to determine the nature of the gaseous mixtures in which small animals die with symptoms of asphyxia. The first of these series were repetitions of the experiments reported by Hammond and described above. Mice and sparrows were used. It was found impossible, by Hammond's method, to absorb all the carbonic acid produced by an animal. At the time of death of the sparrows, the carbonic acid had increased until it formed from 12,27 THE COMPOSITION OF EXPIRED AIR. 403 to 14.08, or an average for eight experiments of 13.24 per ccut of the air, while the oxygen had diminished to from 3.25 to 5.61, or an average of 4.G7 per cent of the air. The symptoms observed were those pro- duced by iusuflficienoy of oxygen, and there was no evidence that death was due to organic matters in the air. The duration of life in the ani- mals confined was from three to six hours, being much longer than that reported by Hammond, using a slightly smaller vessel, viz, less than one hour, and corresponds to the results reported by Sanfelice (33), who found that the animals lived from six to seven hours. When the experiment was so modified that all the carbonic acid was removed from the air breathed by the animal, the animal did not die in seven hours, although the percentage of oxygen had been reduced to 18.35. These experiments, therefore, furnish no evidence of the existence of an organic poison in the expired air, but the method of absorbing carbonic acid by an alkali is said by Brown-S6quard and d'Arsonval (16) to change the organic poison which they chiiin to be i^resent, and hence these experiments are not conclusive on this point. A series of experiments was also made upon mice and sparrows to determine the time required to produce death by asphyxia when the animal is confined in a jar of known capacity, when no provision is made for removing carbonic acid and moisture, or for supplying fresh air, and also to determine the proportions of carbonic acid and of oxy- gen existing in the Inclosed air at the time of death. In connection vritli these experiments, it was also sought to determine the influence which high or low temperatures of the air would have on the result. A mouse weighing 21 grams, placed in ajar of 1,000 c. c. capacity at a temperature of 30° C, lived four hours ; in a jar of 2,000 c. c. capacity a similar mouse lived seven and a half hours; in one case when the room temperature was 25.5° 0., in another case when the room temper- ature was 5° C. In the first case, death occurred when the amount of carbonic acid was 12 and that of the oxygen 8.6 per cent of the mix- ture; in the second case, the proportions were 13.2 per cent of carbonic acid and 6.4 per cent of oxygen; and in the third case, 10 per cent of carbonic acid and 9.2 i^er cent of oxygen. There are considerable dif- ferences in susceptibility to the effects of an impure atmosphere in individual mice, but when a mouse is placed in a closed jar containing ordinary atmospheric air, the time required to produce death is usually that required to produce the proportions of carbonic acid and of oxy- gen indicated above, and, hence, is in proportion to the size of the jar. A mouse should live about twice as long in a jar of 2,000 c. c. as in one of 1,000 c. c, other conditions as to temperature, etc., being the same, and commencing with ordinary atmospheric air. The duration of life in the experiments with atmospheric air in closed vessels, making due allowance for variations in the air volume, coin- cides quite closely with the duration of life in the Hammond experi- ment. The air analyses at death of the animals in the two forms of 404 THE COMPOSITION OF EXPIKED AIR. experimeut, also gave very similar results. In comparing the results it is necessary to bear in mind the differences in the size of the jars and in the weight of the animals used iu the several experiments. As a general rule, the animal dies when the carbonic acid has increased to between 12 and 13 per cent and the oxygen has diminished to between 5 and 6 per cent. Is death due to the increase iu the carbonic acid, or to the diminution in the oxygen, or to both ? Some data for answering this