ay a A SHA Y +t ey is Ph Hes ai Rory | BY } taal {stat vt aif iy ; ae SRT DS tetett et Ste oti eas abr on # ae bani ieee a ue ane he vi Aah Baie ely ee ue see a MSE RA RH He San ay MG ] tah SOLO Maa TILA Heathaaa ne arth iy t Ve e yee f He wal we uy Wi, tt U eee Pata iy i Ae 7 ie ae Phe ae eH > 8t% So ne ere * sae Se re a fet potas a j ag nk oe Hie i ma i is Bays eg Poe : : ae ae) oy 1st oe ae i ae Hate ‘yet ne Hive i: Heiy ie aes Wy i a Ree ye i SO Tait { Rea ee dy = RO eee ~~ ty ene sales ae artes Fan Wa ane optinel . i Hara it} sei i? i lle se y hel higel angie Bil if tate prelencts ote sth i ey a an An Bid Di iichialey, ay “i ies , =, n : 7) By a, j a i 7 ; ti h ( 4' 1 Jn _ : ; ' : at M : ide 1 j , : Mm Wy i : Thai nt i's | it : i va ; AN 7 V4 i : at ‘a a : Ee | as yy) we . si aD 1 | iG Pw ik ; a, VR eo EL ./ hy a THE AMERICAN JOURNAL ? OF PHY SIOLOGY. EDITED FOR The American Physiological Society BY Py H. P. BOWDITCH, M.D., BOSTON FREDERIC S. LEE, Pu.D., NEW YORK R. H. CHITTENDEN, Px.D., NEW HAVEN W. P. LOMBARD, M.D., ANN ARBOR W. H. HOWELL, M.D., BALTIMORE G. N. STEWART, M.D., CHICAGO W. T. PORTER, M.D., BOSTON ELE AMERICAN JOURNAL bl Yeast OlnO.G ¥ VOLUME XIII. BOSTON, U.S.A. GINN AND COMPANY 1905 me AS vila Se Coppright, 1905 By GINN AND COMPANY Gniversity WBress Joun Witson anv Son, CampBripcE, U.S.A. CON TEINS. No. I, FEBRUARY I, 1905. THE REVERSAL OF CILIARY MOVEMENT IN METAzoaNnsS. By G. H. OPES PS EE a ee ee reese 5 EXPERIMENTAL STUDIES ON THE PHYSIOLOGY OF THE MOLLUSCS. — First PAPER. By Lafayette B. Mendel and Harold C. Bradley SOME NOTABLE CONSTITUENTS OF THE URINE OF THE CoyoTe. By erie SUA 6 ws, a) eee THE CHEMISTRY OF THE PROTEIN-BODIES OF THE WHEAT KERNEL. ParRT I.— THE PROTEIN SOLUBLE IN ALCOHOL AND ITS GLUTA- MINIC ACID CONTENT. By Thomas B. Osborne and Isaac F. Harris APPROXIMATELY COMPLETE ANALYSES OF THIRTY “NORMAL” URINES. By Otto Folin ° . . . . . . . LAWS GOVERNING THE CHEMICAL COMPOSITION OF URINE. By Ofto ec dee es PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL SOCIETY... . No. II, Marcu 1, 1905. A THEORY OF PROTEIN METABOLISM. Sy Otto Folin .... . MEASUREMENT OF ELECTRICAL CONDUCTIVITY FOR CLINICAL PURPOSES. 0 RS GIN GT, I a ae a a . THE EFFECT OF SALT SOLUTIONS ON CILIARY AcTiviTy. Sy S. S. LTE i a No:, LIT, Aprit oF, 1905- FURTHER OBSERVATIONS ON THE CATALYTIC DECOMPOSITION OF Hy- DROGEN PEROXIDE. By A. S. Loevenhart PAGE 30 SE) 45 66 xi 117 139 154 171 vl Contents. TWITCHINGS OF SKELETAL MUSCLES PRODUCED BY SALT SOLUTIONS WITH SPECIAL REFERENCE TO TWITCHINGS OF MAMMALIAN MUS- CLES.. \By Walter E. Garrey. 9. 2 = a THE ROLE OF CERTAIN IONS IN RHYTHMIC HEART AcTiviTy. By LAY FE BENEAUE «2 ees es) THE EFFECT OF PIGMENT-MIGRATION ON THE PHOTOTROPISM OF GAM- MARUS ANNULATUS S. I. SMITH. By Grant Smith . . ... - THE NATURE OF CARDIAC INHIBITION WITH SPECIAL REFERENCE TO THE HEART OF LimuLus. Sy A. /. Carlson “= | 3) ce PERFUSION EXPERIMENTS ON EXCISED KIDNEYS. By Torald Sollmann. I. General Methods. .. . re 3 ee II. The Effects of Changes in the Anteral Veron: and Tei Pressure {il.. Anisotonic Solutions: (2 72 ":) 37s Bee en ee IV. Solutions of Non-Electrolytes. -. < s..9. 2 3 = 2 ee V-, The Effect of Viscidity <_<) /- re eee SSC; VI. The Action of Blood on the duh oo, ee pie 0a On Errors OF ECCENTRICITY IN THE HuMAN Eye. Sy Cherles S. FLGSTIR GS fe ee Is Ye nme, Nee Ss ee THE DETERMINATION OF WATER IN FOODS AND PHYSIOLOGICAL PREP- ARATIONS. Ly Francis G. Benedict and Charlotte R. Manning . . No. IV, May 1, 1905. PHARMACOLOGY OF ETHYL SALICYLATE. By E. M. Houghton . . . THE CHEMISTRY OF MALIGNANT GrRowrTHs. — III. NUCLEO-HISTON AS A CONSTITUENT OF TUMORS. Sy S. P. Beche . => = ee No. V, JUNE 1, 1905. FURTHER EVIDENCE OF THE FLUIDITY OF THE CONDUCTING SUBSTANCE IN NERVE. By A> J. Carlson: ie 2 ae THE PHYSIOLOGICAL BEHAVIOR OF METHYLENE BLUE AND METHYLENE AZURE: A CONTRIBUTION TO THE STUDY OF THE OXIDATION AND REDUCTION PROCESSES IN THE ANIMAL ORGANISM. By Frank P. Underhill and Oliver FE. Closson. =). 2 ee eee ON THE UNION OF A SPINAL NERVE WITH THE VAGUS NERVE. Sy SOSCp-EVIGKLET . k n m COMPARATIVE PHYSIOLOGY OF THE INVERTEBRATE HEART. —II. THE FUNCTION OF THE CARDIAC NERVES IN Motuuscs. By A. /. Garson: Se ek ee ee THE INFLUENCE OF ALKALOIDS AND ALKALOIDAL SALTS UPON CATAL- ysis. By Orville H. Brown and C. Hugh Neilson .... . PAGE 186 192 205 2FF 241 253 278 286 289 291 304 309 331 341 351 358 372 396 427 Contents. vil PAGE THE PRECIPITATION LIMITS WITH AMMONIUM SULPHATE OF SOME VEGETABLE PROTEINS. Sy Thomas B. Osborne and Isaac F. Harris 436 CONTRIBUTION TO OUR KNOWLEDGE OF THE ACTION OF PEPSIN, WITH SPECIAL REFERENCE TO ITS QUANTITATIVE ESTIMATION. By Percy re aT el Mk Vk Re Pee ede Nc) yp ak (sae lg Sie os oe te, te AAS NOTE ON THE PREPARATION OF Nucleic Acip. By Ylenry B. Slade . 464 Rr Bags Fost IL Be eS Ow ac we) Sead ss se SAO PeOcCkEDINGS OF THE AMERICAN PHYSIO— EOGICALS SOCIE ALY, SEVENTEENTH ANNUAL MEETING. PHILADELPHIA, PA., DECEMBER 27 and 28, 1904. . q PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL SOCIETY: A CONTRIBUTION TO CELL-CHEMISTRY. By VICTOR C. VAUGHAN. AFTER the bacteria have reached maturity, they are removed from the agar in which they have grown, and washed with dilute alcohol, then extracted with absolute alcohol, and finally with ether. After this they are finely powdered and heated for from one to two hours in a flask with reflex-condenser with twenty-five times their weight of sodium alcoholate (2 per cent of sodium hydrate in absolute alcohol). This splits the cell-substance into two portions, which I will desig- nate as A and BZ. Portion A constitutes in round numbers about one-third of the cell- substance. It is freely soluble in water and absolute alcohol, insolu- ble in ether, petroleum ether, and chloroform. Its aqueous solutions are decidedly acid to litmus, and decompose sodium bicarbonate. It gives all the proteid color-reactions, and can be divided into sub- portions (a and a’) by precipitation with platinum chloride. Portion a constitutes from 10 to 15 per cent of 4, is freely soluble in water and absolute alcohol, gives all the color-reactions for proteids, and kills animals (guinea-pigs, rabbits, and goats), in amounts varying with the animal and method of administration, in less than an hour. Its poisonous effects are identical with, but much more rapid than those of the living bacteria and the dead germ-substance. It lowers the temperature, and kills by arresting respiration. Animals treated with gradually increased doses of a acquire an active immunity to the living germ, and furnish a serum containing an antitoxin. Therefore, portion a is the intracellular toxin of the bacillus. Portion & is soluble in water, partly soluble in dilute alcohol, in- soluble in absolute alcohol, and consists of several sub-portions xi xii Proceedings of the American Physiological Society. (6, 6’, 6", etc.). When portion B is dissolved in water, and acidified with hydrochloric acid, portion 4 is precipitated ; after removing this precipitate by filtration, and on heating the filtrate, portion 0’ is thrown down, while 4” remains in solution. Portion & is soluble in water, insoluble in absolute alcohol, thermostabile, and hzmolytic. In neutral or feebly acid solution, it dissolves red blood-cells at body- temperature after a period of incubation. Portion 6” splits heemo- globin with the formation of hematin. Animals treated with portion B acquire immunity to the living germ, and furnish a bacteriolytic, but not an antitoxic, serum. The essential part of the germ-substance is a definite chemical compound, containing, along with other groups, toxic, haemolytic, and hemoglobin splitting groups. Toxins have been extracted so far in my laboratory from colon, typhoid, and anthrax cells, and immunity has been secured with the isolated toxins of the first two. THE HYDROLYTIC CLEAVAGE OF PROTOALBUMOSE. By P. A, LEVENE. Tue observation described in this report has been made in the course of a comparative study of the composition of various albumoses. In the fraction containing the hexon bases of the protoalbumose a substance differing in its properties from the three known bases was observed. It did not readily form compounds with picric acid and platinic chloride. Towards silver solution it behaved similarly to histidin. The copper and the silver salts were analyzed. The analysis made it probable that the substance had the composition Ci.H.N,O;. (Substances with the composition of C,,H,,N,O; and C,,H,,N,O,; have been described by Skraup and by Fischer respectively.) For Cy2.HeyN4O,Cu CALCULATED. FOUND. C’ 39.67 per ‘cent 40.14 per cent H 551 “ 549° | ee: 2 16.40 aS O 22.04, 04 Cs A al Cienigr oe i He 0 ad Proceedings of the American Physiological Society. xii The analysis of the silver salt gave the following results: CALCULATED. FOUND. Ne—) to:s9 pen cent 10.32 per cent ie AROS ae mo 41.11 ss Further analysis of the substance is in progress. CILIARY REVERSAL IN METAZOANS. By G. H. PARKER. THE reversal of the effective stroke of cilia, though common among protozoans, has been recorded only in a few scattered cases among metazoans (sponges, sea-anemones, planarians, larvae of marine worms, gills of mussels, and possibly frog embryos). In a common sea-anemone, Metridium, the labial cilia ordinarily strike outward and filter-paper pellets, etc., when placed on the lips, are discharged from the mouth. A piece of crab-meat placed on the lip causes the cilia to reverse, and it is thus swept inward. The cilia reverse toa filter-paper pellet soaked in crab-meat juice, but not to a piece of crab-meat deprived of most of its extractives. The crab-meat juice contains the chlorides of sodium, potassium, calcium, and magnesium. Sea-water containing about 2.5 per cent of potassium chloride will reverse the cilia; the other chlorides will not. The labial cilia will live several hours in 37 NaCl solution. They reverse in a 3m NaCl + 4 KCI solution which has an osmotic pressure of about 36.9 atmospheres, but not in a § # NaCl + 4 NaCl solution which has an osmotic pressure of about 36.0 atmospheres. Hence, reversal is not produced by osmotic pressure nor by Cl ions. They reverse in 8m NaCl + 4m KNOg, but not in $2 NaCl + 4 NaNO,. Hence, reversal is due to K ions. Meat-juice containing much less potas- sium than the weakest artificial mixture producing reversal will still cause a vigorous reversal. Hence the reversal caused by meat-juice is probably not simply dependent upon the contained potassium. Temporary reversals of this kind may occur in the ciliated organs of many of the higher animals, for instance, the palps of lamellibranchs, and the female genital ducts of vertebrates. xiv Proceedings of the American Physiological Society. A STUDY OF THE CONDITIONS FOLLOWING THE ESTAB— LISHMENT OF THE ECK FISTULA IN DOGS. BYP? Ba HAWK THE operations were successfully performed by Dr. J. E. Sweet. The liver function was greatly impaired. Dog No. 2. Body-weight, 9.64 kg. Fed mixed diet eleven days, and on the following four days beef meal (pre-digested powdered beef) and milk. On the afternoon of the fourth day of this diet there were pronounced ataxia, entire loss of sight and hearing, com- plete anzsthesia and catalepsy. In about four hours the symp- toms began to abate, and on the next day had entirely disappeared. The animal fasted for twenty hours, and was then placed on a diet of fresh lean beef. On the morning of the fifth day of this diet symptoms were observed similar to those occurring after the beef- meal diet. These disappeared in about five hours, leaving the dog very weak. The animal had now lost 28 per cent body-weight. Placed on a diet of milk and bread the animal died on the fifty-ninth day of the experiment. A poor appetite and a large loss in body- weight (42 per cent) were noted. The autopsy revealed about 800 c.c. of ascitic fluid in the body-cavity, and numerous whitish nod- ules the size of a small pea in the omentum, mesentery, and liver, smears from which showed presence of numerous tubercle bacilli. The fistula was about 2 cm. long. There were no anastomoses. Dog No. 4. Body-weight, 13.08 kg. Beef fed thirty days with negative result. On thirty-first day injected sodium carbamate. No effect. Beef fed twenty-nine days, and no symptoms resulted. Liebig’s extract was added to the diet, and on the morning of the tenth day of this diet, and seventieth of the experiment, symptoms were noted similar to those observed in the case of Dog No. 2, in this instance being supplemented by tetanus. There was a loss of 15 per cent in body-weight. The autopsy showed no anastomoses and no tubercular nodules. The fistula was 2 cm. long. The feeding of sodium carbamate, and the injection of this salt into the blood-stream of normal dogs, were productive of none of the symptoms noted above. Proceedings of the American Physiological Society. xv ON JENSEN’S THEORY OF GEOTROPISM IN PARAMCECIUM. By E. P. LYON. JENSEN! attempts to explain the orientation of paramcecium in response to gravity by supposing that the difference in hydrostatic pressure between the upper and lower surfaces of the animal is the effective stimulus. But this difference is extremely small. Assum- ing that the organism is at a depth of ten centimetres in water, that the pressure of the atmosphere is 1030 grams per square centimetre, and that the thickness of the animal is 0.001 cm., I calculate that, according to Jensen’s theory, the protozoan would be capable of re- sponding to a pressure-stimulus of only ;57}p 5, Of the total pressure on one side, or a still smaller fraction of the total pressure on the entire cell.? Moreover, when one increases the total pressure, the difference in pressure between the upper and lower surfaces being due to the dif- ference in depth must remain constant. If Weber’s law holds good, one should be able to reach a total pressure of such magnitude that the organism would no longer respond to the very small difference in pressure between its upper and lower sides. I have increased the pressure up to three atmospheres, and do not find a diminished re- sponse. On the other hand, if one removes the whole or part of the atmospheric pressure, thus reducing the total pressure to a small fraction of its former intensity, one would expect the stimulus (which is now a much larger fraction of the total pressure), to produce quicker and more accurate orientation. Such is not the case. Using the centrifugal machine in the ordinary way, ¢. e., with the tubes open towards the axis of rotation, one can increase the accel- eration (equivalent to gravity), with but very slight or no increase of total pressure. If the centrifugal force be greater than gravity (within limits) the response of the animals is quickened, even though the pressure be unchanged. Finally, if one uses on the centrifuge tubes open at the outer end and closed at the axis, the condition is such that the pressure in- 1 JENSEN: Archiv fiir die gesammte Physiologie, 1893, liii, p. 428. ? When this was written I had not seen JENNINGS’ last paper in which practi- cally the same calculation is made, and JENSEN’s theory rendered doubtful by quite different observations from my own. See JENNINGS: Journal comparative neurology and psychology, 1904, xiv, p- 442. xvi Proceedings of the American Physiological Society. creases from the periphery toward the axis. Nevertheless, the ani- mals go against the centrifugal force, and from low to high pressure. Therefore Jensen’s theory is incorrect. ON PARA-LACTIC ACID." By ARTHUR R. MANDEL (READ BY INVITATION). EXPERIMENTS by Ray, McDermott, and Lusk? showed that after producing phlorhizin diabetes in a fasting dog poisoned with phos- phorus, the urine changed from an ammoniacal to an acid reaction. It seemed possible that lactic acid, which is produced in phosphorus- poisoning, and probably is the cause of the excessive quantity of am- monia in the urine, might be derived as a cleavage product of sugar which originates from proteid. If this be true, no lactic acid should be formed in diabetes even though phosphorus-poisoning be present. Experiments showed that such was the case. The blood and urine of a fasting dog, poisoned with phosphorus, contained lactic acid, but this disappeared when diabetes was induced by phlorhizin. In another case, diabetes was induced, and then for three days phos- phorus and phlorhizin were administered. At the end of the period there was no lactic acid in the blood or in the urine. It seems, therefore, probable that the lactic acid in phosphorus- poisoning is derived from the cleavage of sugar. Preliminary experiments after ingesting fermentation lactic acid in diabetes apparently show a slight reduction of proteid metabolism, and therefore of sugar output, and also indicate the synthesis of a small quantity of lactic acid into sugar. FURTHER RESULTS IN SUPRARENAL GRAFTING. By F. C. BUSCH anp C. VAN BERGEN. Five cases of suprarenal grafting in rabbits’ kidneys, in addition to the one reported at the last meeting of the society, have shown physiological activity of the grafts. In all of these cases, living grafts, composed in whole or in part of * Read by Dr. Graham Lusk. * Ray, W. E., T.S. MCDERmort, and G. Lusk: This journal: 1899, iii, p. 139. : Proceedings of the American Physiological Soctety. xvii the medullary portion of the suprarenal, have been demonstrated microscopically, from thirty-one to two hundred and forty-seven days after transplantation. A METHOD FOR THE QUANTITATIVE ESTIMATION OF CARBAMATES IN ANIMAL FLUIDS. By J. J. R. MACLEOD anp H. D. HASKINS. ALTHOUGH the method described by us in this journal (Vol. XII, p. 444) for the estimation of carbamates in blood gives very accurate results for added carbamate, we do not consider it, as yet, sufficiently reliable for the detection of minute quantities of carbamate in blood. Thus we have, on several occasions, obtained a slight positive read- ing in the unheated tube (B), which certainly did not come from carbamate. When the blood is in the slightest laked, this positive reading is invariably obtained. Whenever possible, we use as the control, instead of the heated specimen, some normal blood-serum treated exactly like that under examination. We have made a number of estimations of the rate of decomposi- tion of carbamate of ammonia in blood and in water, and have found that at room-temperature blood-serum retards the decomposition to a slight extent. Thus, carbamate of ammonia, yielding 0.950 c.c. of gas, was dissolved in blood-serum; after half an hour 8 per cent of this gas was recovered. In water in which this amount of car- bamate was dissolved only the minutest trace of gas due to carba- mate could be obtained. This preserving property would appear to be due to the inorganic salts, for artificial plasma yields a similar result. SOME REMARKS ON THE CHEMISTRY OF CARBAMATES. By J. J. R. MACLEOD anp H. D. HASKINS. WHEN ammonia water is added to a solution of sodium carbonate or bicarbonate no carbamate is formed unless there be a large excess of ammonia present. Thus, when solutions of ammonia and of sodium carbonate were mixed in such quantities as to satisfy the equation 2 NH,OH + Na,CO, =(NH,),CO, + 2 NaOH, and allowed to stand in a well-stoppered vessel over night, no trace of carbamate could be xvi Proceedings of the American Phystological Society. detected by the method described by us in this journal (Vol. XII, p. 444), or by Drechsel’s method. In the presence of a large excess of ammonia however (1 c.c. 7 Na,CO, + 8 c.c. 4 NH,HO) carba- mate was formed in small amount (Drechsel’s method).! By mixing solutions of ammonium chloride and sodium carbonate in various proportions, large amounts of carbamate were invariably obtained. These results conclusively demonstrate that carbamates are not readily formed when preformed ammonium hydroxide is brought in contact with soluble carbonates ; whereas ammonia (NH,) in the free state (“im Enstehungszustande”’) readily forms it. Drechsel’s con- clusion that carbamate can result without the ammonia being free, would therefore not hold unless when a large excess of ammonia is present. It appears to us that this result satisfactorily explains why no carbamate should be formed when a mixture of ammonium chloride and milk of lime is treated with sodium carbonate solution, whereas it is formed when the sodium carbonate is added before the milk of lime.? When we first noted this result, we thought that the presence of the sodium hydroxide formed by the chemical reaction might be the cause of the non-formation of the carbamate. That such is not the case we have shown by the fact that carbamate readily forms in a mixture of 5 c.c. 4 Na,COg, 5 c.c. 4 NH,Cl, and 10 c.c.4 NaOH. In the oxidation at body-temperature of glycocoll, aspartic acid, and tartaric acid, with ammonium permanganate in the presence of 2.5 per cent ammonia water (2.¢., 0.7 per cent NH,), a very large amount of carbamate is produced, especially at an early stage in the process. In our experiments, total CO, and carbamate CO, were estimated at varying intervals during the oxidation, the permanganate being added in~ small quantities at a time (0.05 gm.). The following were the maximal values, expressed as per cent of carbamate to total CO,: glycin 25 per cent, aspartic acid 43 per cent, and tartaric acid 29 per cent. With oxalic acid at first, acetamid and ethyl alcohol oxidation to CO, was very much slower, and only in the case of oxalic acid was any measurable amount of carbamate CO, obtained.® 1 DRECHSEL: Journal fiir praktische Chemie, 1877, xvi, p. 180. ? DRECHSEL’s first two experiments, Loc. cét. * HOFMEISTER: Archiy fiir experimentelle Pathologie und Pharmakologie, 1896, xxxvii, p. 426. Proceedings of the American Phystological Society. xx ON THE ORIGIN OF KREATININ. By WALDEMAR KOCH. In the course of an investigation of the food-value and intermediary metabolism of lecithin, the interesting relation of its methyl groups to the methyl group of kreatinin suggested the possibility of measur- ing the amount of lecithin metabolism by the amount of kreatinin excretion. Feeding experiments with kreatin-free diets varying in lecithin from 0.5 to 7 gm. per day indicate that the amount of kreatinin can be varied, but only to a limited extent, as the capacity of the animal body for forming kreatinin seems to be limited. The excess of lecithin is probably stored, as it does not appear in the faeces. The figures which will be published later further indicate that besides the lecithin another factor (proteid) is involved in sup- plying most or all of the nitrogen of the kreatinin molecule. The possibility here suggested of two factors entering into the formation of kreatinin may lead to an explanation of the hitherto conflicting statements as to the source of this interesting substance. ON THE ELIMINATION OF CREATININE. By LAFAYETTE B. MENDEL ann OLIVER E. CLOSSON. Tue available data regarding the origin and elimination of creatinine in man are relatively meagre and conflicting. Thus the presence of creatinine in the urine of vegetarians has recently been denied by Caspari and Glessner, while other observers have reported figures scarcely smaller than those noted in persons living on a mixed diet. The relation of creatinine excretion to various types of diet is rather uncertain ; and its variations under different conditions of nutrition are practically undetermined. The importance of an investigation of this subject is apparent in view of the fact that creatinine, next to urea, is one of the most prominent nitrogenous constituents of the urine, and its connection with muscular metabolism or proteid trans- formations may be significant. The writers have determined the total creatinine output in a con- siderable number of individuals under varying conditions of diet, in health and disease. The collected statistics will be published later ; xx Proceedings of the American Physiological Soctety. attention is here directed to the noteworthy excretion of creatinine in vegetarians, as well as in individuals living on a low proteid diet for long periods, and to the rate of excretion during brief intervals in comparison with the simultaneous total nitrogen and uric acid output. In general, on a creatin-free diet, there is a tendency towards paral- lelism between the total nitrogen output and the elimination of the creatinine group. A CASE OF TUMOR OF THE FLOOR OF THE FOURRGE VENTRICLE WITH CEREBELLAR SYMPTOMS, EN PAR CAGE By FRANK P. KNOWLTON (READ By INVITATION). A cat brought to the laboratory presented the following interesting condition : There was a marked weakness of the muscles of the left side, as compared with those of the right. This weakness affected the hind limbs more than the fore limbs. Spasm and rigidity of the muscles of the right side were especially noticeable, and there was a tendency toward extensor rigidity of the fore limbs. The animal’s body was curved to the left. The occiput was drawn to the left and the face to the right. The head and tail were bent to the left. ) Tremor was present, increased by movement and by excitation. Movement was in circles toward the left, and there was a tendency to roll around the long axis, and always to the left. The animal was readily fatigued by slight exertion. Examination of the eyes showed rotary mystagmus. Hearing and cutaneous sensation seemed unimpaired, and there was no evidence of pain. Post-mortem examination showed a tumor arising from the left half of the floor of the fourth ventricle and extending from the median line to the restiform body, at the level of the ventral nucleus of the cochlear nerve. Microscopical examination of the specimen by Dr. H. S. Steensland showed that the tumor was a glioma. Sections of the medulla and pons, stained by the Marchi method, Proceedings of the American Physiological Society. xxi showed marked degeneration of the pons fibres and the eighth nerve on the left side, with compression and some degeneration in other tracts. ON THE PRESENCE OF SOAPS IN THE ORGANISM IN CERTAIN PATHOLOGICAL CONDITIONS (A PRELIMINARY COMMUNICATION). By OSKAR KLOTZ (By INVITATION). Tuts I can quite confirm, that the centre of the calcareous deposits consists of calcium salts laid down in a homogeneous matrix (Litten), but this is not the case at the periphery, where the active process of calcification is taking place. On the contrary, here we find sub- stances which chemically react as soaps, and can be isolated as such. Such soaps are here present as the potassium, sodium, or even the ammonium compounds of the fatty acids, being rendered insoluble by the combination with albumins of the degenerating tissues. Such soap-albumin compounds, in the presence of the calcium of the blood, form insoluble calcium curds or double calcium soaps. These insoluble calcium substances are then decomposed into the phos- phate and carbonate of lime, which, as has been said, remain as the permanent deposits of calcium in a homogeneous matrix. It is further shown that soaps may be demonstrated in pus, particularly that of chronic abscesses, and Langerhaus has shown that a calcium soap is formed in fat necrosis. In the production of an experimental calcareous degeneration in the kidneys of rabbits, it is shown by means of staining that the fatty changes taking place in the cells of the tubules is in the same location as the calcareous deposit, and further that in these kidneys a soap may be demonstrated by proper chemical analysis. The following conclusions are arrived at: 1. The earliest change in cells, which later undergo calcareous de- generation, is one of cloudy swelling or coagulation necrosis. 2. Following this, fatty changes are noticeable in the cells, and now, by means of proper reagents, soaps with potassium, sodium, and presumably ammonium bases, can be detected. . 3. Such soaps and albumins form a combination which is insoluble in water or salt solution. xxii Proceedings of the American Physiological Society. 4. Soaps and fatty acids have an affinity for the calcium salts in solution in the body-fluids, and form with them an insoluble compound. 5. Later, judging from the fact that phosphate and carbonate of lime are formed, and the deposits give no reaction for fats, the fatty acid moiety of the calcium soap is replaced by the more powerful carbonic and phosphoric acids. OBSERVATIONS ON THE ALIMENTARY CANAL AFTER SPLANCHNIC AND VAGUS SECTION. By W. B. CANNON. IN some animals the greater and lesser splanchnic nerves were cut on both sides; in other animals the right vagus nerve was cut below the recurrent laryngeal branch, and later the left vagus was divided in the neck. X-ray observations were made on the movements of food mixed with bismuth subnitrate. The prolonged stasis of food in the thoracic cesophagus after vagus section (noted first by Reid in 1839) is in marked contrast to the almost normal advance of the food after it has entered the stomach. The movements of the stomach and the intestines (peristalsis and rhythmic segmentation) were seen after vagus or splanchnic section. The stomach movements were inhibited by distress in both conditions. The difference in the rate of discharge of carbohydrates and proteids was also preserved in both conditions, but the discharge of both foodstuffs from the stomach was slower than normal when the vagi were cut. There was also a slight delay in the passage of the food through the small intestine after Vagus sectlon. THE EFFECT OF CEREBRAL INJURIES ON THE BULBAR VASOMOTOR CENTRE. By W. T. PORTER anv T, A. STOREY. THE condition of the vasomotor centre was determined by measur- ing, with a membrane manometer, the fall in carotid blood-pressure following stimulation of the depressor nerve before and after injuries to the brain. Rabbits were used. They were anzsthetized, when Proceedings of the American Physiological Society. xxiii this was necessary to prevent pain. The experimental injuries con- sisted of removal of the cerebral hemispheres and cerebellum, fractures with intracranial hemorrhage following blows on the cranium, and compression of the brain by means of thin bags inserted through a trephine opening and distended with water. The most extensive injuries did not inhibit the vasomotor centres. Indeed, the fall in blood-pressure following stimulation of the de- pressor nerve is commonly greater after the removal of the cerebral hemispheres than in the normal animal. We are making similar experiments upon the cat and dog. As the depressor is not a separate nerve in the dog, and is frequently not separate in the cat, it is necessary to use in its place some other nerve afferent to the vasomotor centre. Professor Porter is deter- mining to what extent the sciatic and other nerves thus replace the depressor nerve. Observations made simultaneously on the de- pressor and sciatic nerves, during shock produced by painting the intestines with nitric acid (in anesthesia), show that the sciatic nerve may be safely employed to measure the working power of the vasomotor cells, although the necessary use of curare introduces some difficulties. THE CURVE OF LESSENING CONDUCTIVITY DURING INCREASING TONUS OF THE HEART. By W. T. PORTER anp F. H. LAMB. IN 1893 it was suggested! that “ fibrillar contractions of the heart may be due to an interruption of the contraction-wave. The con- traction-wave would thus be prevented from running its usual course, and the normal co-ordinated action of the ventricular cells would give place to the confusion conspicuous in fibrillary contractions.” In 1899 it was pointed out 2 that ‘‘the tonus of the ventricle is greatly increased at the onset of fibrillation, and that co-ordinated beats do not return unless the extreme, apparently tonic contraction is con- siderably lessened. The continued shortening observed may be due to an abnormal, long-continued, tonic contraction of a number of the cardiac fibres. The appearance of such a spasm at several points 1 PorTER, W. T.: Journal of physiology, 1893, xv, p. 134. 2 PorTER, W. T.: This journal, 1899, ii, p. 129. xxiv Proceedings of the American Physiological Soczety. might break up the normal conduction of the contraction wave by interposing here and there regions, the intense spasmodic contrac- tions of which would block the passage of the contraction-wave, just as it is blocked experimentally by squeezing the muscle-fibres to- gether with a clamp, and would leave the remaining muscle-fibres dissociated and in confusion.” In 1902,! this hypothesis was brought formally to the attention of the Society. In 1903,? it was experi- mentally demonstrated that as the tonus increases the conductivity of the heart diminishes. The present observations attempt to fix quantitatively the relation between the rise of tonus and the consequent fall in conductivity. This relation is expressed by a curve in which the millimetres of tonic shortening of the muscle are plotted as abscissz, and the hundredth seconds of increase in latent period as ordinates. Sucha curve steadily rises, but the rise is broken by deviations statistically . accidental. By increasing the number of observations, these acci- dental errors may be balanced against each other, and thus eliminated. We have obtained in this way a curve which rises apparently in a straight line, as does the curve of developing energy in muscle. We have also noted instances of extreme tonic spasm. These observations support the hypothesis outlined above. THE VOLUME-CURVE OF THE MAMMALIAN VENTRICLE (ILLUSTRATED). By YANDELL HENDERSON anp MARVIN McCRAE SCARBOROUGH. THE principal method employed was to enclose the ventricles (of dogs) in a plethysmograph, and to record the volume-changes by means of air-transmission and a tambour. Above the volume-curve was recorded the intraventricular or aortic pressure-curve. The volume-curve of the ventricle, both in the heart-model shown at the last annual meeting of this society and in the living animal, closely resembles a series of inverted isotonic muscle-curves, and indi- cates that in the instant after systole is at an end, and the tension of the muscular walls is relaxed, the ventricles are refilled to their full * PorTER, W. T.: This journal, 1902, vi, p. xxiv. * Porter, W. T.: This journal, 1903, viii, p. xxvi. Proceedings of the American Phystological Society. xxv capacity by the rush of blood from the large veins and auricles. The ventricles are therefore full, and, as will be shown in another paper, the auriculo-ventricular valves are closed several hundredths of a second before the succeeding auricular systole. Under normal conditions, the blood flowing toward the heart is sufficient in volume to distend the thoracic veins and the elastic walls of the auricles, which are merely enlargements of the veins. It is this elasticity of the distended walls of the cardiac reservoir, largely assisted in the left heart by the suction of the empty ventricle, and not the contrac- tion of the muscle-fibres in the walls of the auricles, the auricular systole, which fills the ventricles. When a man lies upon a plank suspended from long wires, each movement of blood feetward into the ventricles or headward into the arteries causes an equal mass-movement of his body and the plank in the opposite direction. By recording these movements (magnified by levers fifty to one hundred times) the quantitative volume-curve of the human heart is obtained. It is similar to those above described. THE EVENTS WITHIN THE HEART. By YANDELL HENDERSON. SINCE the volume and pressure-curves of my heart-model are (as shown by the tracings presented with the previous paper) essentially similar to those obtained from animals and from man, it may fairly be assumed that the events within this model afford a true picture of the corresponding events within the living heart. Through the courtesy of Professor C. H. Judd, it was’ made possible by means of a kinetoscope! to record the working of the heart-model photograph- ically. Twelve successive pictures, in a total period of six-tenths of a second, show all the events of a cardiac cycle, and also the move- ments of the stylus of a Hiirthle manometer recording intraventricu- lar pressure on a drum. These photographs show, therefore, the pressure-changes, the instant of opening and closing of the valves (especially noteworthy being the closure of the mitral in mid-diastole), and the periods of the filling and discharge of the ventricle. 1 The kinetoscope was lent to the Yale Psychological Laboratory by the Edison Manufacturing Company. —— xxvl Proceedings of the American Phystological Soctety. Further investigations on the model (by the ordinary graphic method) show that the diastolic wave in the intraventricular pressure- curve is due to the inertia of the column of blood which, after pour- ing through the auricular orifice during the earlier portion of diastole, is brought to a standstill by the full ventricle resisting further disten- tion. The narrower the auricular orifice, and the longer the tube or funnel formed by the flaps of the mitral valve when open, the greater is the pressure which this column of blood exerts in coming to a stand- still. This increased pressure intensifies the diastolic wave. It also causes a more sudden and forcible closure (diastolic in time) of the mitral or tricuspid valve. Thus are produced the diastolic sounds incident to mitral stenosis; thus also arise the ‘‘ weak systoles,” alter- nating with the strong, often noted in mitral stenosis, — the weak and abortive ‘‘ systole” being truly no systole, but merely an exaggerated diastolic wave. AN INSTANCE OF COMPLETE “ HEART-BLOCK” TINS MAR: By JOSEPH ERLANGER. ANALYsIS of tracings of the cardiac-impulse, jugular pulse, and brachial pulse obtained from a case of Adams-Stokes disease demon- strates that the primary circulatory disturbances are consequent to a diminution in conductivity at the auriculo-ventricular junction. At most times this block is complete, the ventricular rhythm being totally independent of the rhythm of the venous end of the heart. In one determination the rate of ventricular beats was 27.6 per minute, of auricular beats 98 per minute. It seems probable that the main ventricular systole is followed, after an interval of 0.27 of a second, by an extra-systole, although the evidence for this is not perfectly clear. The absence of ventricular beats in some parts of the tracings permits of the dissociation.of the activities of the auricles from those of the ventricles. In each auricular cycle there appear in the jugular vein one negative and two positive waves. The higher positive - wave is undoubtedly produced by the contraction of the right auricle. The deep negative wave that follows it probably results from active dilatation of the auricle. The small but distinct positive wave that precedes the wave produced by auricular systole may owe its origin Proceedings of the American Physiological Soczety. xxvii to contraction of the great veins. This wave is most distinct when the venous pressure is high. Each contraction of the auricles sends into the peripheral arteries a small positive wave followed by a distinct negative wave. The results obtained in an experiment testing the effect of posture upon the rates of the auricles and ventricles indicate that extrinsic nervous influences are brought to bear on the venous end of the heart almost exclusively. For when the patient’s posture was changed from recumbent to erect the ventricular rate was increased, but 0.6 ofa beat per minute, whereas the auricular rate was increased 15.5 beats per minute. One tracing obtained after the patient had been in the recumbent posture for some time shows a condition of partial ‘ block” in which the rhythm of the auricle is to the rhythm of the ventricle as 3 is to 1. This relation was maintained when the patient stood, although the heart-rate was increased. While the subject was holding his breath, the condition of partial block gave way to one of complete block. It is possible that this change in conductivity was produced by a venous condition of the blood. A METHOD OF MEASURING AND RECORDING THE MAXI- MUM AS WELL AS THE MINIMUM BLOOD-PRESSURE IN MAN. By GAYLORD P. CLARK. To determine the minimum blood-pressure, a wide cuff is applied to the arm with suitable arrangements for increasing the atmospheric pressure within the cuff, which is connected with a straight-tube mercury manometer and a recording mechanism similar to the one reported by Erlanger.! To determine the disappearance and the reappearance of the pulse below the arm-band, and therefore the maximum blood-pressure, a wide cuff is also applied to the forearm, and the entire apparatus employed in connection with the arm, except the manometer, is duplicated for the forearm. The pressure is applied to each cuff independently, in the following manner: a bottle that can be elevated at will, filled with water, is 1 ERLANGER: This journal, 1go1, vi, p. xxii. xxviii Proceedings of the American Phystological Soczety. connected through an outlet in the bottom by means of a large rubber tube with the bottom of another bottle stationary upon the table. The air-space in the upper part of the latter bottle is connected with the interior of the cuff. In order to secure uniform increase of pres- sure in the arm-cuff, the pressure-bottle connected with that cuff is raised by means of a small windlass. To determine the minimum and maximum pressure of the blood, the pressure-bottle connected with the forearm-cuff is quickly ele- vated to a height approximating the average minimum blood-pressure (75 mm. Hg), thus producing a good excursion of the lever connected with its recording apparatus, then by means of the windlass the other pressure-bottle is elevated to a point above that at which the pulsa- tions of the forearm-cuff cease, after which it is lowered. The amount of the pressure is recorded by pinching a rubber tube connected with a tambour as the column of mercury passes each 5 mm. mark on the scale, both in rising and in falling. The disappearance and reappearance of the pulse in the upper trac- ing indicates the point of maximum blood-pressure, and the greatest amplitude of the excursions in the lower tracing indicates the point of minimum blood-pressure. SOME OF THE PHYSICAL PHENOMENA OF MUSCLE-FATIGUE. By FREDERIC S. LEE. THE investigation of the subject has been continued by the employ- ment of a method in which the isotonic curves of all the contractions of an excised, non-curarized muscle, stimulated at regular intervals, are superimposed upon a recording surface. The differences which were previously? pointed out in the mode of fatigue of the muscles of the frog, the turtle, and a mammal, have been confirmed. Lohmann’s work, in which the frog’s gastrocnemius, on being heated to a mam- malian temperature, shows a course of fatigue similar to that of mammalian muscle, has been repeated and found in general correct. But the coracoradialis profundus of the turtle, similarly heated, con- tinues to give its characteristic curve of fatigue. Hence Lohmann’s conclusion that there is no physiological difference, in the matter of fatigue, between cold-blooded and warm-blooded muscle, does not seem justified. 1 LEE: This journal, 1898, ii, p. xi. Proceedings of the American Phystological Soctety. xxix Kaiser's method for determining the point on the isotonic curve where the contractile stress terminates, has been employed for the frog’s gastrocnemius, and it has been found that as the height of the curve diminishes in the course of fatigue, the contractile stress terminates at progressively lower and lower points. The lowering of the latter does not, however, seem to keep pace with the lowering of the summit of the curve. Hence the two points seem to approach one another. GELATINE AS A SUBSTITUTE FOR PROTEID IN THE FOOD. By J. R. MURLIN (READ By INVITATION.) Docs were placed in nitrogen equilibrium, at the starvation level, with diets containing proteid, fat, and carbohydrate. Different frac- tions of the proteid-nitrogen in the food were then replaced by gela- tine-nitrogen, and the effects on the equilibrium, as expressed in the percentage of the body-proteid spared, were determined. The calo- rific requirements of each dog, estimated from Rubner’s tables, were supplied in full in all diets’ The experiments lasted from two weeks to two months for each dog, the same diet being ingested in periods of from three to five days, separated sometimes by fasting periods of two or three days. Results agree in showing that it is a matter of indifference how much of the proteid-nitrogen is replaced by gelatine-nitrogen, up to one-half of the starvation requirement. With two feedings per day, the percentage of body-proteid spared does not decrease, even if two- thirds of the proteid-nitrogen are replaced by the gelatine-nitrogen ; and perfect balance may be maintained with this fraction, provided the carbohydrates in the diet amount to one-half to two-thirds of the calorific requirements. In an experiment on a man of 70 kg.,a fasting period of three days was followed by a period of the same length during which nitroge- nous equilibrium was maintained on a diet containing the quantity of nitrogen eliminated during starvation. The proteid was supplied by beefsteak, oatmeal, and eggs, which, with cream and cane-sugar, fur- nished a total of 3,000 calories. Two-thirds of the total proteid- nitrogen were replaced by gelatine-nitrogen for two days, with the result that on the second day there was a gain of 0.06 gm. N. The xxx Proceedings of the American Physiological Soczety. energy available in this last period was raised by addition of carbo- hydrate to 3,400 calories. Moderate muscular exercise was taken throughout the experiment. The conclusion is that a mixed diet, more than covering the heat- requirement of the organism, and containing two-thirds of the starva- tion minimum of nitrogen in gelatine, and one-third in proteid, will maintain nitrogenous equilibrium for a few days at least. In other words, the proteid requirement under the combined sparing action of gelatine, fat, and carbohydrate falls to one-third the starvation- requirement. THE EFFECTS OF ISOTONIC SOLUTIONS ON THE KIDNEY? By TORALD SOLLMANN. PeRFuSION of excised kidneys with various solutions of the same freezing point, as I per cent sodium chloride, produces the following effects : Non-Electrolytes. — Cane-sugar causes very slight changes, due to its viscidity and specific gravity. Glucose produces very little effect. Alcohol and urea cause marked diminution of the vein and ureter flow, and of the volume of the kidney. The effect is that of hypo- isotonic solutions; these substances penetrate the cells. Cations. — Potassium and ammonium produce very little change. Magnesium increases the vein and ureter flow. Barium, calcium, and hydrogen diminish the vein and ureter flow, and the volume. Anions. — The sulphate, and especially the citrate, increase the vein and ureter flow and the volume. The hydrate, carbonate, and bicar- bonate diminish them. The effects of any of these solutions can be removed by subsequent perfusion with sodium chloride. Mixtures of ions, as far as investi- gated, produce a combination of the effects. Locke’s solution (omit- ting the dextrose) causes a slight increase of ureter-flow, without change of the vein-flow or areometer. As far as tried, the effects of any of these solutions can be produced several days after the excision of the kidneys. They are probably osmotic actions, solutions of the same freezing point being anisotonic to the kidney-cells, on account of differences in permeability. Proceedings of the American Physiological Society. Xxx1 THE EFFECT OF BLOOD ON THE BLOOD-VESSELS OF THE KIDNEY. By TORALD SOLLMANN. THE perfusion of excised kidneys with viscid solutions (egg-albumen or acacia) results in a marked decrease of the vein and ureter flow, and of the volume of the kidney. The same result occurs on perfus- ing kidneys with diluted defibrinated blood, several days after excision: in these, the vein-flow is almost stopped. A very different result is seen when recently excised kidneys are perfused: in these, blood causes generally a considerable increase of vein-flow; at the same time, the ureter-flow and renal volume are diminished. The vein-flow may be two or three times greater than with saline solution. This indicates a very pronounced active dilation of the blood-vessels, involv- ing probably the afferent, and especially the efferent arteries. This dilator effect is also produced by blood saturated with carbon mon- oxide, by blood laked at 63 C., and by serum. It is not produced by Locke’s solution, nor by blood coagulated in the steam-bath. It is therefore not produced by the corpuscles, by oxyhemoglobin, or by the serum salts ; but by some substance (proteid?) which is either destroyed or precipitated on heating to coagulation (not at 63° C.). The dilator action is somewhat different, quantitatively, in different samples of blood. The response of different kidneys is also variable; so that in some cases the viscidity effect predominates. The dilator reaction diminishes slowly with the time elapsing after excision, dis- appearing in eighteen to forty-eight hours, at about the same time as the power to reduce haemoglobin. The dilator effect varies with dilution of the blood; but for slight dilution the change is less than that of the viscidity; so that a moderately diluted blood gives a greater vein-flow than an undiluted blood; while a greatly diluted blood may give a lesser vein-flow. Hydrocyanic acid has also a very marked dilator effect on the renal vessels; but this is confined mainly to the afferent vessels. Adrenalin may also cause a dilation under certain conditions, whilst its ordinary effect is a pronounced constriction. xxxil_ Proceedings of the American Physiological Soctety. FURTHER STUDIES ON RICIN. By THOMAS B. OSBORNE ann LAFAYETTE B. MENDEL. In continuation of the experiments reported at the last meeting,! the castor-bean proteids have been separated further by fractional pre- cipitation with neutral salts into portions of relatively great, and relatively slight toxicity, and the ricin prepared in still purer form. The toxic preparations alone possess the power of sedimenting the colored corpuscles of mammalian blood; this property is not lost - until the proteid solutions are heated to the temperature of coagula- tion. The pure ricin preparations suffer no deterioration in their physiological activities after being kept for many months. Even after administration of quantities many times larger than the fatal dose, a “latent period” of at least fifteen hours intervenes before the toxic symptoms become manifest, as has been noted by other observers. The increase in the dosage is important mainly in determining the rapidity with which the subsequent pathological conditions arise. This fact throws some light on the probable nature of the toxicological action produced by ricin. The toxicity of the pure ricin is enormously diminished when it is administered through the alimentary canal, instead of subcutaneously. ON THE RATE OF ABSORPTION FROM INTRA- MUSCULAR TISSUE. By S. J. MELTZER anp JOHN AUER. WE came across the observation that absorption from the muscles is incomparably more rapid and efficient than from the subcutaneous tissue, and we tested the fact with several substances. With supra- renal extract we tested it in three ways: (1) By the effect upon blood-pressure: a dose of 0.6 c.c. or less of adrenalin per kilo (rabbit) exerts in subcutaneous injection no effect, and the unstable effects of larger doses consist in a rise of pressure varying between 10 mm. and 20 mm. mercury, which sets in late and develops slowly; while an intramuscular injection of 0.5 c.c. or 0.4 c.c., or even less, per kilo * OsBornE, T. B., and L. B. MENDEL: This journal, 1904, x, p. xxxvi. Proceedings of the American Phystological Soctety. xxxiil causes invariably a considerable rise of pressure which sets in after a very short latent period, and reaches its maximum in a few minutes. The obtained curve looks very similar to that of an intravenous in- jection. The rise can be as high as 50 mm. or 60 mm. of mercury and even higher, and the course is frequently interrupted by ‘ vagus- pulses.” (2) It was further tested by the effect upon the pupil on the side where the superior cervical ganglion had previously been re- moved: a dose of 0.5 c.c. or 0.4 c.c. of adrenalin per kilo, given in- tramuscularly, will cause the dilatation of the pupil in less than a minute, while such a dose, in subcutaneous application, has rarely any effect, and the effect of a larger dose sets in only after ten or fifteen minutes. (3) It was finally tested by the prostrating effect: a dose of 0.5 c.c. per kilo will prostrate a rabbit in a minute or two after intramuscular injection, while in subcutaneous application the effect will not come before twenty or thirty minutes, and then only after much larger doses. We tested the fact further by means of curare: a dose can be found which will have no apparent effect in subcutaneous injection, but will cause, in a few minutes, paralysis of the voluntary muscles after an intramuscular application. We have established the striking difference between the effects of the two modes of application also for morphine and fluorescein. A STUDY OF THE COLORING MATTERS IN THE PURPLE PITCHER PLANT (SARRACENIA PURPUREA). By GUSTAVE M. MEYER anv WILLIAM J. GIES. Two years ago, while studying the digestive powers of Sarracenia purpurea, Gies detected a sensitive pigment in aqueous extracts of the pitchers, to which the name a/kaverdin was given (Journal of the New York Botanical Garden, 1903, iv, p. 37). A few preliminary experiments showed that alkaverdin bore superficial resemblance to the characteristic pigments in red cabbage, elderberry, dahlia, etc. The earlier experiments have lately been extended, with the following general results: Alcohol extracts three coloring matters from the macerated purple pitchers, — chlorophyl (green), alkaverdin (purplish red), and a brownish-black material. Water extracts only the last two coloring xxxiv Proceedings of the American Physiological Soctety. matters. The brownish-black material may be separated from the alkaverdin in an aqueous extract by evaporating the latter 2 vacuo nearly to dryness. On pouring the residue into absolute alcohol, the brownish-black substance is precipitated. The alcoholic solution contains all of the alkaverdin, and, when evaporated z7 vacuo nearly to dryness, yields a residue resembling dark molasses, both in con- sistency and color. This syrup is insoluble in ether and chloroform, but is soluble in water and alcohol. The color of the brownish-black material is unaffected by acid oralkali. The substance fails to give a reaction with ferric chloride, and is without reducing power. The syrup containing the alkaverdin has a bitter taste and sugary odor. The characteristic odor of the macerated pitchers is also present. Halogens, nitrogen, and sulphur are absent. The syrup contains a large proportion of fermentable and dextrorotary carbo- hydrate, which has a strong reducing action on Fehling’s solution, even in the cold, and yields phenyl dextrosazone-like crystals. Sugar can be removed by fermentation without apparent detriment to the pigment. Aqueous solutions of the syrup are reddish in color. When very dilute, the solutions are colorless. Such colorless solutions be- come green when treated with a drop of % alkali (“alkaverdin ”) and are rendered colorless by a drop of % acid. A drop or two of acid in excess causes the production of a pzzk color quite unlike the red of the syrup itself. On standing, the pink color is somewhat inten- sified. In less dilute solutions these tinctorial changes are very strik- ing. Delicate test papers have been made with alkaverdin. Although no crystalline pigment has yet been separated from the syrup, it is certain that alkaverdin has no effect on the spectrum. On warming its aqueous solutions for some time, on the water bath, alkaverdin is transformed into a brownish product which no longer yields a pink color with acid, but which is still colored green by alkali. After hydration with dilute sulphuric acid (2 per cent), alkaverdin entirely loses its tinctorial properties. Whether alkaverdin is closely related to chlorophy] or not, has not yet been determined. Proceedings of the American Physiological Society. xxxv A FURTHER STUDY OF PROTAGON. By E. R. POSNER anp WILLIAM J. GIES. RECENT observations by Cramer and by Thierfelder indicate that protagon, as made by Liebreich’s or Gamgee’s method, contains an admixture of cerebron (“pseudocerebrin”). The latter cannot be entirely removed from the mixture by the usual cooling process. In all probability every product heretofore called protagon has been a mixture of substances, as Lesem and Gies and others have sug- gested, and it is doubtful whether the name protagon signifies anything definite. This view is confirmed by our recent experiments. Cramer pre- pared protagon by a new method, which involved preliminary “ co- agulation” of the protagon by heating the brain-tissue on the water bath (at 100° C.) with sodium sulphate solution. Protagon was then prepared by Gamgee’s method from the tissue which had been treated in this way. In our own experiments, thus far, protagon has been prepared by the Gamgee method and purified by recrys- tallization from 85 per cent alcohol. The purified products were then heated with sodium sulphate solution, as in the Cramer process. The protagons prepared by the latter method dissolved readily in 85 per cent alcohol at 75° C., and were subjected to the fractionation process (at 45° C.) of Lesem and Gies, with results simi- lar in detail to those published by them. Thus, the phosphorus-con- tent of each successive fraction of the protagon decreased, while the phosphorus-content of the substance in each successive protagon filtrate (at o° C.) rose far above that of the corresponding protagon, or of any protagon-fraction. The phosphorus-content of the insoluble matter was much lower than that of the original protagon. Protagons containing as much as 1.7 per cent of phosphorus may be obtained by changing somewhat the method of preparation. Ad- ditional facts in this connection will be given at the conclusion of the experiment now in progress. xxxvi Proceedings of the American Phystological Society. CERTAIN ASPECTS OF EXPERIMENTAL DIABETES. FRANK P. UNDERHILL. Wirth the purpose of extending our knowledge of carbohydrate metabolism, particularly under abnormal conditions, an explanation has been sought for the hyperglycemia and accompanying glycosuria which can be called forth by experimental methods, — painting of drugs upon pancreas, etc. When piperidin is painted upon the pan- creas of dogs, hyperglyceemia is a constant symptom, and it bears a striking similarity to that provoked by adrenalin. The application of the substances directly to the pancreatic cells is not essential, since similar results may be obtained by painting the spleen, by intraperi- toneal injection, or by direct introduction into the blood. It seems likely, therefore, that the influence of piperidin, etc., is exerted through the intervention of the circulation rather than directly upon the gland- cells to which the substances have been applied. The experimental diabetes is not due to an irritant action upon, or an “insult” to, the pancreas. The hyperglycemia may be induced independently of the pancreas. The experiments indicate that the action of piperidin, etc., is not specific; nor is the pharmacological action of the drugs neces- sarily related to their chemical structure. The tentative explanation offered for the hyperglycemia and glycosuria provoked by such sub- stances as piperidin, potassium cyanide, ether, chloroform, morphine, carbon monoxide, strychnine, pyrogallol, pyrrol, pyridin, coniin, nico- tin, curare, etc., and possibly adrenalin, is not that they have any specific action (such as deprivation of oxygen) upon any particular gland like the pancreas, but that these substances have more or less effect upon the respiratory centre in producing dyspncea. Dyspnoea calls forth a marked hyperglycemia and glycosuria without the inter- vention of any drugs. The typical rise in the sugar-content of the blood produced by piperidin fails to appear when oxygen is adminis- tered. It seems probable, therefore, that this experimental diabetes is due to diminished oxidation of carbohydrate material, with the con- sequent accumulation of the latter in the blood, and its elimination by the kidneys. ao Ti Proceedings of the American Physiological Society. xxxvii SOME OBSERVATIONS ON THE CATALYTIC DECOMPOSITION OF HyDROGEN PEROXIDE BY ORGAN Extracts. By A. S. LOEVENHART. ON THE THEORY OF GEOTROPISM IN PaRAMccIUM. By E. P. Lyon. THE NITROGEN OF URINE; ITs DISTRIBUTION AMONG THE Four IMPORTANT ConsTITUENTS — UREA, Ammonia, Uric AcID, Krratinin. By O. Forin, (By invitation). On Psycuic SECRETION. By W. Kocu. THE PHARMACOLOGY OF ETHYL SaLicyLaATe. By E. M. HouGuron. DEMONSTRATION OF A NEw WaTER Manometer. By F. S. Lee (for H. EMERSON ). AN ADJUSTABLE TRACHEAL CANNULA FOR ARTIFICIAL RESPIRATION. By Go P. CLARK. AN EXHIBITION OF PHYSIOLOGICAL APPARATUS. By W. T. PorTER. DEMONSTRATION OF APPARATUS. By E. P. Lyon. DEMONSTRATION OF APPARATUS. By A. P. BRUBAKER. The following communication was read by title : A SECOND COAGULATION OF THE BLOOD DUE TO A SUBSTANCE NOT IDENTI- CAL WITH FIBRINOGEN, AND COAGULABLE BY SATURATION WITH NEUTRAL OxaLaTE. By E. T. REICHERT. THE American Journal of Physiology. VOL. XIII. FEBRUARY 1, 1905. N@r Ir CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E. L. MARK, Drirecror. No. 159. THE REVERSAL OF CILIARY MOVEMENT IN METAZOANS. Bw (Ge Ee eACR Sen Re I. INTRODUCTION. HE reversal of ciliary movement, though one of the commonest phenomena among protozoans, seems to be of much rarer occurrence among the higher animals. In a recent review of the problems of ciliary movement, Piitter (:03, p. 35) devotes only two paragraphs to ciliary reversal in metazoans, nor does the literature on the subject warrant a more extended treatment. Apparently Purkinje and Valentin (35, p. 67), who observed spon- taneous ciliary reversal on the accessory gills (Nebenkiemen) of the mussel, were the first to record this condition in metazoans. Their statements were repeated by Valentin (42, p. 513) in his article on cilia in Wagner’s Handworterbuch der Physiologie, and were confirmed later by Engelmann (68, p. 476; ’79, p.386; 98, p. 788). Accord- - ing to Miklucho-Maclay (’68, p. 232), some sponges can temporarily reverse the direction of their water-currents, an observation sub- scribed to by Haeckel (’72, p. 26), and indicating a probable reversal in the action of their flagella. Minot (’77, p. 407) observed that the cilia at the head ends of certain fresh-water planarians strike in vari- ous directions, while those on other parts ofthe body always strike backward, and Iijima (’84, p. 436) noticed that certain bands of cilia in Dendroccelum, Planaria, and Polycelis strike now in this direction, I 2 G. H. Parker. now in that. Probably the cilia of many of the smaller water-inhabit- ing metazoans reverse, as recorded by Kleinenberg (’86, p. 22), for the larva of Lopadorhynchus, and by Schulze (91, p. 13), for Trichoplax. Finally the labial cilia of certain actinians reverse dur- ing feeding, as shown by me (Parker, ’g6, p. 114) in Metridium mar- ginatum, by Vignon (:o1, pp. 638, 656), in Sagartia parasitica, and by Torrey (:04, p. 212), in Sagartia davisi. Engelmann (’79, p. 386) stated that he never observed ciliary reversal in any of the verte- brates; but Roux (’95, p. 106) recorded an instance of the reversal of the direction of rotation in a frog embryo that may have been due to this cause. Thus only a few scattered cases of ciliary reversal in metazoans have been put on record, and these only in the briefest way. In the present investigation I have attempted to work out, in more detail than has been done heretofore, the conditions of reversal, and for this purpose I .have experimented with the labial cilia of the actinian Metridium marginatum. II. OBSERVATIONS. In Metridium marginatum the oral disk consists of three concentric areas which from the periphery inwards are as follows: the ten- tacular zone (see figure), bearing the numerous tentacles; the inter- mediate zone, without special organs; and the labial zone, represented by the swollen lips and siphonoglyphs, which together surround the mouth. The tentacles, lips, and siphonoglyphs are well ciliated; otherwise the oral disk is without cilia or at most sparsely provided with them. In an expanded resting Metridium the mouth is usually somewhat open, the lips are protruded, and the tentacles are recumbent and point away from the centre of the disk. Under such circumstances the effective stroke of the cilia can be demonstrated, by means of car- mine particles, to be inward on the siphonoglyphs, outward on the lips, and outward on the tentacles, z.¢., toward their tips. Thus cur- rents pass into the animal at the siphonoglyphs, and out of it over the lips and tentacles. In all the trials that I have made on the direc- tions of currents in quiescent animals, those just stated have invari- ably been met with, and I, therefore, have no grounds for supposing that any of these ciliated parts of Metridium show spontaneous reversals such as have been claimed to occur on the gills of mussels. The Reversal of Crliary Movement in Metazoans. 3 So far as my experience extends, the application of various stimuli to the tentacles has never resulted in a reversal of the effective stroke of their cilia, and the same is true of the siphonoglyphs. On the lips, however, as I have elsewhere shown (Parker, ’96, p. 113), ciliary reversal may be accomplished by the application of appropriate stimuli. The normal outward stroke of the labial cilia is not reversed by carmine, India ink, sand, or pellets of filter-paper moistened with sea-water ; or with solutions of sugar, quinine, or picric acid in sea- water; but it is reversed when pieces of crab meat are applied to the lips. That this reversal is dependent upon the dissolved substances in the meat is seen from the fact that a pellet of filter-paper, which when moistened with sea-water is carried outward, is swept inward when soaked with meat juice. Moreover, a piece of crab meat that has been so_ thoroughly washed in sea-water as to have lost almost all its extractives will usually not induce reversal. It therefore appears to be certain that the reversal caused by crab meat is dependent upon certain dissolved substances contained in Oral view of an expanded specimen of Me- tridium marginatum, showing the ten- the meat. tacular zone (outermost), the intermediate Another plece of very conclu- zone, and the labial zone (surrounding the sive evidence in favor of this mouth). The specimen, which was di- eee Ge obtained from mi glyphic, was prepared by Semper’s method croscopic preparations. When a piece of the lip is placed in sea-water under a microscope, the normal direction of the ciliary stroke can be easily observed. If now such a preparation is flooded with sea-water containing meat juice, an imme- diate reversal can be seen to take place, and on washing the prepa- ration subsequently with pure sea-water, a return to the normal direction occurs. Such changes can be repeated many times without impairing the piece of tissue. Hence there can be no doubt that the dissolved substances in the meat cause ciliary reversal. The juice obtained from the meat of the common green crab, Carcinus mznas, which was used in these experiments, was in large part blood, and hence it seemed not improbable that the and is reproduced natural size. 4 G. Hl. Parker. blood of this animal contained some substance or substances that caused the reversal. Whether these were organic or inorganic could not be foretold; but because of the greater ease with which pure inorganic substances could be obtained, I resolved to test these first. According to Griffiths (’92, p. 293), the blood of Carcinus contains, besides traces of iron and copper, sodium, potassium, calcium, and magnesium in the form of chlorides, phosphates, and sulphates. As a preliminary test, solutions in sea-water of the more soluble phos- phates and sulphates and of the chlorides of all four metals were prepared and tried on the labial cilia under the microscope. The strengths used varied from 10 per cent to I per cent. No clear evidence of reversal was obtained from any of the solu- tions tested except from that of potassic chloride. The following records from one series of experiments with the solutions of this salt will suffice to make clear its effects and range. 10 per cent potassic chloride in sea-water. When applied, much mucus was discharged and many nettle capsules were exploded. 15 min. 33 sec. after application, ciliary movement ceased. ‘The tissue was then rinsed in pure sea-water and immersed in it. I5 min. 27 sec. after immersion, there having been no signs of recovery and the cells having been apparently killed, the experiment was discon- tinued. 5 per cent potassic chloride in sea-water. When applied, some mucus was discharged and many nettle capsules were exploded. 5 min. 30 sec. after application, the first irregular ciliary movements were observed. 4 min. 30 sec. later, ciliary reversal was well established, though the mucus and exploded nettle capsules impeded the cilia. 5 min. later, the material was transferred to pure sea-water in which ciliary movement continued, but so obscured by mucus that the direction could not be determined. 4 per cent potassic chloride in sea-water. When applied, some mucus and nettle capsules were discharged and the cilia ceased moving. I min. 45 sec. after application, ciliary movement was renewed in irregular form. 3 min. later, ciliary reversal was well established, though the action of the cilia was encumbered by mucus. ro min. later, pure sea-water was applied and ciliary action ceased. The Reversal of Ciliary Movement tn Metazoans. 5 6 min. rs sec. later, the first irregular ciliary movements began. g min. 15 sec. later, normal ciliary movement was re-established, but not as vigorously as at first. 3 per cent potassic chloride in sea-water. When applied, nettle capsules were discharged. 1 min. later, irregular ciliary movements were observed. 2 min. 45 sec. after the first irregular movements, complete reversal was established. Io min. 15 sec. after this, the tissue was placed in pure sea-water. Ciliary action ceased. 3 min. 45 sec. later, irregular local movements began. 9 min. 45 sec. later, normal movement was vigorously re-established. 2 per cent potassic chloride in sea-water. 30 sec. after application, reversed ciliary movement was well established. Some mucus and nettle capsules had been discharged. I2 min. 4osec. later, the tissue was flooded with pure sea-water. I min. 50 sec. later, irregular movements began. 4 min. 20 sec. later, vigorous normal movement was re-established. L per cent potassic chloride in sea-water. I5 min. 45 sec. after application, normal movement still continued. Some mucus and nettle capsules had been discharged. 2 min. 15sec. later, the potassic chloride solution was renewed. 30 min. 30sec. after the first application, the normal movement was still unchanged and the experiment was discontinued. From these records it can be seen that potassic chloride is a stimu- lant for the production of mucus and the discharge of nettle capsules, and that these two results are at times so excessive as to hinder or at least obscure the action of this salt on the cilia. A r per cent solu- tion is apparently not strong enough to cause ciliary reversal, and a IO per cent solution soon kills the ciliated cells. The most success- ful concentration of those just described, whereby reversal could be effected, was 2 per cent, for by this solution reversal was not only quickly accomplished, but the production of mucus, etc., was minimized. Subsequent experiments demonstrated that 2.5 per cent was still better than 2 per cent, and this strength was therefore used in much of the later work. When care was taken to prevent mucus from accumulating on the pieces of tissue, the reversal by means of a 2.5 per cent potassic chloride solution could be brought about as quickly and with as much certainty as with crab-meat juice. Moreover, the artificial fluid would, without great damage to the tissue, induce repeated reversal 6 G. H!. Parker. of the ciliary stroke as the natural fluid did. Thus, in an experiment lasting somewhat over four hours,a single preparation was reversed five times by 2.5 per cent potassic chloride and brought back to its normal state by pure sea-water without apparently suffering any serious injury, for at the end of this experiment the preparation was still vigorous. The five intervals between the moment of applying the salt solution and the occurrence of reversal, and those between the application of pure sea-water and the return to the normal stroke were, as the following table shows, all brief, and in this respect resembled the changes as produced by the natural fluids. TABLE. TIMES IN MINUTES AND SECONDS FOR THE REVERSAL (IN 2.5 PER CENT KCL) AND: THE RETURN TO NORMAL (IN PURE SEA-WATER) OF THE CILIARY STROKE. No. OF EXPERIMENT. In 2.5 per cent potassic chloride. In pure sea-water. Although it must be admitted that potassic chloride is in some way the cause of the reversal of the ciliary stroke, a further analysis of this question requires a consideration of somewhat more artificial conditions. In accordance with the dissociation-hypothesis an aque- ous solution of potassic chloride consists of potassium ions, chlorine ions, anda certain amount of undissociated potassic chloride. It is possible that any one of these three bodies, or some combination of them, may cause the reversal of the cilia. To ascertain upon which of these the reversal depends it was found necessary to devise some chemically simple solution in which the cilia would continue to live, and which might therefore be used as a substitute for the highly complex mixture, sea-water. After some few trials, it was found that the cilia would survive several hours in a 3 7z solution of sodic chloride. When to this was added enough potassic chloride to make a 4 m solution (= circa 2.5 per cent) of that salt, a relatively simple chemi- cal mixture was obtained, by means of which ciliary reversal could be invariably accomplished, though at a slower rate than by potassic chloride in sea-water. Thus the ciliary stroke ofa fragment of a lip was reversed by an immersion of somewhat less than fifteen minutes in a solution containing % # NaCl and 4 m KCl, though a second frag- The Reversal of Ciliary Movement in Metazoans. 7 ment from the same lip used in a check experiment retained its normal direction of strokes for two and a half hours in a 3 mm NaCl solution. Ina second experiment the reversal was accomplished in eight minutes, and the reversed stroke was maintained for over thirty- five minutes, till the preparation was clogged with mucus. By using a solution of 3 NaCl in place of sea-water, and by adding other salts to this solution, it was now possible to proceed to experiments for the elucidation of certain other questions. Since the cilia will continue their normal aes for some time in a solution of 3 NaCl to which has been added 41m NaCl, but reverse it in one to which has been added 4m KCl, it follows that this reversal cannot be due to the osmotic erences of these two solutions, for they are practically isotonic, the pure sodic chloride solution having an osmotic pressure equal to 36.0 atmospheres, and the mixture having one equal to 36.9 atmospheres. It is also improbable that the anions are concerned with the ciliary reversal, for the degree of dissociation of the two solutions, 67.9 per cent in the pure sodic chloride, and 74.0 per cent in the mixture, shows that the state of the chlorine is nearly the same in each. Moreover, no reversals were obtained with §m NaCl + 4LiCl, or with §m NaCl+ im NH,Cl, in both of which the chlorine content is not very unlike that of the reversing fluid. These fluids had very different actions on the cilia; that containing lithium soon brought them to a standstill, while the one with ammonium in it had little or no effect on their movement. If the osmotic pressure and the chlorine ions are not the cause of ciliary reversal, this must be sought for in the potassium ions or the molecules of potassic chloride. Which of these was the probable cause was ascertained by subjecting the cilia to the action of solu- tions containing other potassium salts than the chloride and noting whether reversal occurred or not. A reversal similar to that ob- served in the potassic chloride solution was obtained in six minutes from a solution composed of 8 NaCl +4m KNO;. Since no re- versal was produced by a solution containing 37 NaCl+4 NaNO,, it is clear that the reversal is caused by the potassium ions. Hence in this instance the cation is the stimulating component, though Mathews (:04, p. 456) has recently endeavored to show that these ions are depressing in their action and anions are stimulating. If the potassium ions must be admitted to be the means of revers- ing the effective stroke of the cilia in the experiments just described, 8 G. Hl. Parker. it might be assumed that, since they are also in the blood of the crab, they are the active agents in the reversal produced by that material. This conclusion, however, is probably not.entirely true, for the action of the potassic chloride solution is at least in two respects different from that of the meat juice. The potassic chloride solutions always induce a considerable secretion of mucus and a discharge of many nettle capsules, reactions not especially noticeable with the meat juice. Further, a 1 per cent solution of potassic chloride will not reverse the ciliary stroke though a dilute solution of meat juice, which must contain very much less than 1 per cent of potassic chloride, is very effective in this respect. Because of these differences it seems improbable that the reversal by meat juice is - due simply to the contained potassium. Possibly potassium, in com- bination with organic materials, is the really effective agent in bring- ing about the reversal, but on this point I have no conclusive evidence to offer. Torrey (:04, p. 212) observed that in Sagartia davisi such chemi- cally inert substances as paraffine, glass, paper, etc., were swallowed, and he (:04, p. 210) believed that these substances cause a ciliary reversal by purely mechanical ‘means. In my own experiments on Metridium, though I have looked with care, I have never observed a ciliary reversal due to such stimulation. Perfectly insoluble sub- stances like pellets of filter-paper, etc., are always swept outward on the lips. In some of my earlier experiments I noticed occasionally that a filter-paper pellet supposed to be moistened with sea-water only, would take a somewhat ambiguous course, moving for a short time inward and then outward. By further experiments I convinced myself that these apparent exceptions were due to organic contamina- tion from my fingers, for when care was taken that the pellets should consist of nothing but clean filter-paper wet with sea-water no such reversals occurred. Torrey (:04, p. 210) has suggested that possibly the inward movement of pieces @f India rubber observed by me on the lips of Metridium might be evidence of the reversal of the ciliary stroke in that sea-anemone by mechanical means. I have, therefore, repeated my experiments with small fragments of white India rubber such as is used for laboratory tubing, and I find, as I previously stated (Parker, ’96, p. 115), that these pieces are occasionally swallowed, and are almost always carried inward some distance before they are finally discharged. Filter-paper pellets, however, that have been soaked in a filtrate from a mixture of sea-water and many small pieces The Reversal of Crlary Movement in Metazoans. 9 of rubber, call forth the same reactions, and I therefore conclude that this response is really dependent on dissolved products from the rubber. Thus while the labial cilia of Sagartia davisi may be re- versed mechanically, I know of no conclusive evidence in favor of this means in Metridium. It has long been known to students of ciliary activity that the direction of the effective stroke of a cilium, 7. ¢., the direction in which the adjacent fluid is moved, is independent of the direction of what may be called the propagation wave, 7. ¢., that disturbance which determines the seguwence of stroke of the individual cilia in a general field. Thus the direction of the effective stroke of the cilia on the frog’s palate is inward, as is also the direction of their propa- gation wave. On the other hand, the direction of the effective stroke of the swimming plates of ctenophores is toward the aboral pole, that of their propagation wave toward the oral pole. Thus the two phenomena are independent so far as direction is concerned, since, in the first instance, the directions agree, in the second they are opposed. ; In the normal beat of the labial cilia of Metridium, the effective stroke and the propagation wave agree in direction in that both are toward the exterior. From what has been said, it is conceivable that when the effective stroke is reversed, the propagation wave may or may not be reversed. To ascertain what happened under these cir- cumstances, I made the following experiment. A piece of Metridium lip was placed in sea-water under the microscope, and the normal direction of the stroke was noted. A small piece of crab meat was then attached to the end of a glass rod and brought in contact with the ciliated lip. In a very short time the cilia immediately under the meat could be seen to have reversed their effective stroke, though those in front of and behind the meat were beating in the normal direction. In the region of the reversed cilia not only had the direc- tion of the effective stroke changed, but the propagation wave had also reversed. The lines of separation between the reversed area and the normal areas were rather sharp and located just at the edges of the applied meat, showing that the reversal of both the effective stroke and the propagation wave was a strictly local phenomenon and de- pendent upon local stimulation. Since the reversal includes not only the direction of the effective stroke, but also that of the propa- gation wave, it is probable that the potassium ions or other reversing materials stimulate not only the cilia, but also the deeper lying cyto- 10 G. Hl. Parker. plasm of the ciliated cells through which the propagation wave is transmitted. The strictly local character of the response, as indicated in the pre- ceding experiment, agrees with my previous observations (Parker, ’96, p- 114) on the whole lip of Metridium, and leads to a conclusion, expressed in my former paper, that these experiments yield no evidence in favor of the view that the labial cilia are under nervous control. I see no reason for assuming, as Vignon (:O1, p. 659) has done, that the reversal is dependent upon nervous operations in the nature of central reflexes. Perhaps one of the most noteworthy results of the present investi- gation is the fact that though the ciliary stroke on the lips of Metridium is quickly reversed by weak crab-meat juice, that of the siphonoglyphs and tentacles is unaltered in direction by this material. These conditions point to a high degree of differentiation in the ciliated surfaces of Metridium and emphasize the general truth that ciliated surfaces are closely adapted to their special environments. That such surfaces have much individuality is clearly seen from the fact that the labial cilia of Metridium will live several hours in a pure 2m sodic chloride solution and reverse in one to which 4 potassic chloride has been added, though in both these solutions, according to Lillie (:o1, pp. 62, 65; :02, p. 41), the cilia of Arenicolalagvassane almost immediately destroyed, and those of Polygordius larvze are soon broken down. On the other hand, lithium chloride seems to act uniformly on the cilia of all these animals, and quickly checks their action; whereas ammonium chloride has little or no serious effect. (Compare the statements made by Lillie, :04, p. 422, and on page 7 of this paper.) These observations show the necessity of wide ex- perimental acquaintance with the reactions of cilia before general conclusions of much value can be arrived at. Whether ciliary reversal is a phenomenon of wide or restricted occurrence among metazoans is a question still to be answered. The part that it plays in the feeding habits of actinians suggests that it may be a factor of like import in the feeding and respiratory activities of many of the lower animals, such as the lamellibranchs, etc. It is also possible that reversal may occur in the ciliated tubes and cav- ities of the higher animals. Thus the transportation of the spermatic fluid in those vertebrates with internal fertilization would be facilitated, if among the glandular secretions of the males material was found that would induce temporary ciliary reversal in the oviducal tracts of The Reversal of Crliary Movement in Metazoans. 11 the females. But so far as I am aware, experimental evidence in these directions is almost entirely lacking. III. THEORETIC CONSIDERATIONS. Although the extreme minuteness of cilia has made a direct inves- tigation of their finer anatomy practically impossible, the study of their action has led to certain fairly well founded conclusions as to their probable structure. The old opinion that cilia are not motile, but are composed of an elastic substance, and are moved by a mechanism at their bases, though lately revived in a measure by Benda (:01), has not appealed as a rule to recent workers. The striking correspondence in chemical reactions, etc., between cilia and living protoplasm, the close relation between protoplasmic pseu- ‘dopodia, flagella, and cilia, and the continued activity of isolated cilia, have all supported the belief that the cilia themselves are actively contractile. Nor is there anything in recent work that especially favors Schiafer’s (91, :04) belief that ciliary movement is produced by the injection of fluid into an elastic sheath and its subsequent withdrawal. In fact, it is now generally conceded that these organoids contain within themselves their own contractile material. Such contractile material would act most efficiently if combined with some skeleton-like structure for mechanical support. As Putter (:03, p. 32) has pointed out, the most effective place for the con- tractile material is the periphery of the cilium, the axis of which would therefore be the most advantageous position for the support. It is probable, as Pitter (:03, p. 37) has stated, that the axial fibrils described by Ballowitz (’88, ’90) in the tails of spermatozoa, and believed by him to be contractile material, are really the supporting substance, and that the more peripheral surrounding layer is the true contractile portion. That an axial portion, such as Ballowitz has identified in spermatozoa, occurs in the minute cilia of ciliated epithelia has never, to my knowledge, been demonstrated; but the action of these organoids certainly suggests the presence of such a part, in which case its occurrence ought to be characteristically different for the two kinds of cilia, reversible and irreversible. In ordinary cilia incapable of reversing, the contractile layer ought to be best developed on the face of the cilium corresponding to the direction of the effective stroke. On the face opposite this, only I2 : G. H. Parker. enough contractile material would be necessary to bring about a return of the cilium to the position at which the effective stroke begins, and even this substance might be done away with provided the supporting structures had enough elasticity for the return of the cillum. The absence of contractile material from the sides of the cilium would restrict its movement to a single plane. Such a cilium has been likened to a vertebrate limb, in that it would possess an internal skeleton and flexor and extensor elements, though it may be that the extensors are replaced by the elasticity of the skeleton. But whether we postulate extensors, or only an elastic skeleton, the ultimate structure of such a cilium must be unsymmetrical in that the contractile material must be gathered especially on one side of the supporting axis. With reversible cilia, the ultimate structure must be quite different. These cilia must consist of flexor and extensor elements placed on the opposite sides of the supporting axis, and about equally developed, but so specialized that, to take an example from the labial cilia of Metridium, in ordinary sea-water, what we may call the flexor ele- menis are stimulated to excessive action, and in sea-water containing potassium ions the extensor elements are more efficient. Reversible cilia, therefore, should have a more symmetrical form than the irre- versible ones, but their contractile materials would be more highly specialized than those of the irreversible cilia. What the exact nature of this contractile material may be is quite unknown, but I see no reason to suppose that it may not be deli- cately fibrillar, as suggested by Engelmann (68, p. 456). Putter (:03, p. 32) has opposed this hypothesis because of the difficulties met with in applying it to the complex and often variable movements of flagella, and has argued in favor of a more nearly homogeneous material in which the particles may temporarily arrange themselves in lines in various directions. But if Engelmann’s hypothesis is not wholly satisfactory, Putter’s seems to me still less so, for it contains no suggestion as to what determines the new linear arrangements of particles, and thus leaves unanswered the real question at issue. In my opinion, it is entirely possible that, while the contractile layer of some flagella may be almost homogeneous, that of the more regu- larly beating cilia is very probably fibrillar, for certainly the move- ments of these two kinds of organoids are different enough to warrant a difference of structure. I therefore do not agree with Putter in discarding the fibrillar hypothesis of ciliary structure because it is . } { The Reversal of Crthary Movement in Metazoans. 13 inconsistent with some of the facts of flagellar movement; but on the contrary I believe that this hypothesis is the most consistent thus far advanced, as an explanation of the more usual type of ciliary movement, — a movement which in its regularity differs widely from that of most flagella. IV. SUMMARY. 1. The labial cilia of Metridium marginatum do not reverse when in contact with carmine, India ink, sand, pellets of filter-paper moistened with sea-water, or with solutions of sugar, quinine, or picric acid in sea-water. They also do not reverse as a rule to crab meat from which most of the extractives have been taken. 2. They reverse to dilute crab-meat juice and slightly to a sea-water extract of rubber. They also reverse to a 2} per cent solution of potassic chloride in sea-water. 3. They will live several hours in a pure 2 # NaCl solution, and reverse in a solution containing 3 7 NaCl and 4 m KCl. 4. Reversal is not due to the osmotic action of the reversing fluid, for they reverse in 8 # NaCl +4 KCl (36.9 atmospheres), and not in 8 # NaCl + 4 # NaCl (36.0 atmospheres ). 5. Reversal is not due to the anions, for it occurs in 3 # NaCl + 4 m KCl, but not in § m NaCl+ 4m NaCl, 3 mNaCl+3m LiCl, or 8 m NaCl + 4m NH,Cl. 6. Reversal is due to potassium ions, for it occurs in 3 ™m NaCl + 4m KNO,, but not in 3 m NaCl + 4 m NaNO,. 7. The reversal due to crab-meat juice is probably dependent upon some organic combination containing potassium. 8. There is no evidence of spontaneous reversal or of reversal through mechanical stimuli. g. In reversal the propagation wave, as well as the effective stroke, change directions. 10. The reversal occurs only where the stimulus is applied, and gives no evidence of involving nervous reflexes. 11. Though crab-meat juice causes the labial cilia to reverse, it does not alter the direction of the ciliary stroke on the tentacles or the siphonoglyphs, thus demonstrating the extreme differentiation of these surfaces to stimuli. 12. Irreversible cilia are probably unsymmetrical, in that they consist of a supporting elastic element, on at least one side of which contractile material is present. 14 Gil. Parken: 13. Reversible cilia are probably more nearly symmetrical, in that there must be contractile material on opposing sides of a supporting element; but these two portions of contractile material must be regarded as chemically different, in that one becomes predominantly active in pure sea-water, and the other in sea-water containing crab- meat juice or potassium ions. BIBLIOGRAPHY. BALLOwWITZ, E. 88. Untersuchungen iiber die Struktur der Spermatozoen, zugleich ein Beitrag zur Lehre vom feineren Bau der kontraktilen Elemente. Arch. f. mikr. Anat., Bd. 32, pp. 401-473, Taf. 14-18. BALLOWITZ, E. ‘90. Fibrillare Struktur und Contraktilitat. Arch. ges. Physiol., Bd. 46, Pp: 433-464- BENDA, C. :o1. Ueber neue Darstellungsmethoden der Centralkérperchen und die Ver- wandtschaft der Basalkorper der Cilien mit Centralkérperchen. Arch. f. Anat. u. Physiol., Physiol. Abt., Bd. 1901, pp. 147-157. ENGELMANN, T. W. 68. Ueber die Flimmerbewegung. Jena. Zeitschr. f. Naturwiss., Bd. 4, pp. 321-479, Taf. 6. ENGELMANN, T. W. 79. Physiologie der Protoplasma- und Flimmerbewegung. /# L. Hermann; Handbuch der Physiologie, Bd. 1, Theil 1, pp. 341-408. ENGELMANN, |. W. ’98. Cils vibratils. Zz C. Richet; Dictionnaire de Physiologie, tom. 3, pp. 785-799. GRIFFITHS, A. B. ‘92. On the Blood of the Invertebrata. Proceed. Roy. Soc. Edinburfh, vol. 18, pp. 288-294. HAECKEL, E. 72. Die Kalkschwamme. Bd. 1. Berlin, 8vo, xvi + 484 pp. Irjima, I. $4. Untersuchungen iiber den Bau und die Entwicklungsgeschichte der Siiss- wasser-Dendrocoelen (Tricladen). Zeitschr. f. wiss. Zool., Bd. 40, Heft 3, Pp. 359-464, Taf. 20-23. KLEINENBERG, N. ’36. Die Entstehung des Annelids aus der Larve von Lopadorhynchus. Zeitschr. f. wiss. Zool., Bd. 44, Heft 1-2, pp. 1-227, Tat. 1-16. Liab) oy toes Re ‘ol. On Differences in the Effects of Various Salt-solutions on Ciliary and on Muscular Movements in Arenicola Larve. Amer. Jour. Physiol., vol. 5, no. I, pp. 56-85. Lhe Reversal of Crliary Movement in Metazoans. 15 Meee. RS. 702. On the Effects of Various Solutions on Ciliary and Muscular Movement in the Larve of Arenicola and Polygordius. Amer. Jour. Physiol., vol. 7, no. I, pp. 25-55- TeenIE, RS. 104. The Relations of Ions to Ciliary Movement. Amer. Jour. Physiol., vol. Io, no. 7, pp. 419-443. MATHEWS, A. P. 704. The Nature of Electrical and Chemical Stimulation. I.—The Physi- ological Action of an Ion depends upon its Electrical State, and its Electrical Stability. Amer. Jour. Physiol., vol. 11, no. 5, pp. 455-496. MIKLUCHO-MAcLay, N. 68. Beitrage zur Kenntniss der Spongien. I. Jena. Zeitschr. f. Naturwiss., Bd. 4, pp. 221-240, Taf. 4-5. Minor, C. S. : 77. Studien an Turbellarien. Arb. zool.-zootom. Inst. Wiirzburg, Bd. 3, Heft 4, pp. 405-471, Taf. 16-20. PARKER, G. H. *96. The Reactions of Metridium to Food and other Substances. Bull. Mus. Comp. Zo6él. Harvard College, vol. 29, no. 2, pp. 107-119. PURKINJE, J. E., et G. VALENTIN. 35. De Phanomeno generali et fundamentali Motus vibratorii continui. .. . Wratislaviae, 4°, 96 pp. PUTTER, A. 703. Die Flimmerbewegung. Ergeb. Physiol., Jahrg. 2, Abt. 2, pp. 1-102. Roux, W. 95. Ueber die Zeit der Bestimmung der Hauptrichtungen des Froschembryo. Gesammelte Abhandlungen tiber Entwickelungsmechanik der Organismen, Bd. 2, No. 16, pp. 95-124. SCHAFER, E. A. ’91. On the Structure of Amceboid Protoplasm, with a Comparison between the Nature of the Contractile Process in Amoeboid Cells, and in Muscular Tissue, and a Suggestion regarding the Mechanism of Ciliary Action. Proceed. Roy. Soc., London, vol. 49, pp. 193-195. SCHAFER, E. A. 704. Theories of Ciliary Movement. Anat. Anz., Bd. 24, No. 19-20, pp. 497-511. SCHULZE, F. E. gt. Ueber Trichoplax adhaerens. Abhandl. Kgl. Preuss. Akad. Wiss., Berlin, 1891, 23 pp., 1 Taf. TorREY, H. B. 04. On the Habits and Reactions of Sagartia davisi. Biol. Bull., vol. 6, no. 5, pp. 203-216. 16 G. Hl. Parker. VALENTIN, G. "42. Flimmerbewegung. /z R. Wagner; Handworterbuch der Physiologie, Bd. 1, pp. 484-516. VIGNON, P. or. Recherches de Cytologie générale sur les Epithéliums, V’appareil pariétal, protecteur ou moteur. Le Réle de la coordination biologique. Arch. de Zool. exp. et gén., sér. 3, tom. 9, pp. 371-715, pl. 15-25. Pee RIMENTAL STUDIES ON THE PHYSIOLOGY OF THE, MOLLUSCS.— FIRST PAPER. BYeeOrAYETTE B. MENDEL anp HAROLD C. BRADLEY. [From the Sheffield Laboratory of Physiological Chemistry, Vale University.] T is probable that no group of the lower animals has attracted so much attention and furnished the means for such numerous inves- tigations along physiological lines as the Mollusca. There is no group of the invertebrates or of the lower vertebrates that lends itself so readily to the experimenter, that is found so widely distributed, and that offers at the same time such a wide degree of variation and specialization. Between the slowly moving, thoroughly protected lamellibranchs and gasteropods, with their rudimentary sense organs and their slow, uncertain response to almost all forms of stimuli, and the cephalopods depending entirely for protection and food upon their highly perfected sense organs and their speed of motion and reaction to stimuli, there exists an enormous range of variation. Indeed the difference between the lower and higher members of this group is scarcely less than that which exists between the lowest and highest vertebrate. No other group has adapted itself so thoroughly to meet the demands of such varied types of environment ; and it is for this reason that a comparative study of its various members in relation to their environment is especially interesting. In the present series of studies on the physiology of the molluscs, Sycotypus canaliculatus has been selected as the subject of investi- gation for several reasons. In the first place, the large size and com- mon occurrence of this gasteropod along the Atlantic coast greatly facilitates the study; and secondly, it represents a type of mollusc that has hitherto been neglected. Considerable work has been done upon the herbivorous gasteropods such as Helix and Limax, as well as upon the pulmonates and lamellibranchs, while still more is known about the physiology of the cephalopods, the most obvious examples 17 18 Lafayette B. Mendel and Ffarold C. Bradley. of carnivorous molluscs; 1! but so far as can be learned no carnivorous gasteropod, of which Sycotypus is one of the most extreme types, has been thus examined. Sycotypus is found abundantly in Long Island Sound and the adjacent waters, frequenting the oyster beds, where it finds its food. When unable to secure this otherwise, it is believed to be able to bore through the shell of its prey and suck out the contents in the same way in which the common “drill” obtains its food. Associated with it, but occurring in smaller numbers, is the closely allied form Fulgur carica. Our studies have shown no distinct physiological dif- ferences between the two; and the descriptions of experiments upon Sycotypus apply equally well to Fulgur. Both are thoroughly pro- tected by their shells and opercula, are extremely slow of movement, and are provided witha simple and comparatively low form of nervous and sensory mechanism. Metabolism of all kinds is slow; digestion, respiration, muscular movement, and reaction are of a-low order, cor- responding to the limited demands of the environment with which the animal associates itself. THE PHYSIOLOGY OF THE ALIMENTARY CANAL OF SyYCOTYPUS CANALICULATUS. Inasmuch as any description of the physiology of an organ implies familiarity with its anatomy, it has been deemed advisable to intro- duce a brief discussion of the anatomy of the alimentary canal of Sycotypus in connec- tion with the study of its functions. For the general anatomical ar- rangement, the diagram will perhaps serve the FiGuRE ].— Diagram of the digestive tract of Sycotypus. pb., proboscis; s. g., salivary gland; c., crop; s¢., stom- PDUrpose better than a ach; /., liver; dv., hepatic ducts; #., pylorus; z., inm- more exact figure. The testine; @z., anus; s.d., salivary duct; ~., radula. names used throughout are those commonly employed by morphologists? for the designa- tion of the various organs. They refer evidently to analogies of posi- 1 The literature on this subject is well reviewed in v. FURTH’s Vergleichende chemische Physiologie der niederen Tiere. 2 Cf. Bumpus: Invertebrate Zoology. _& Experimental Studies on the Phystology of Molluscs. 19 tion merely, and are no indication, per se, of the function of an organ. The anterior end of the alimentary canal of Sycotypus, as of other allied carnivorous gasteropods, is prolonged to form a movable pro- boscis which can be withdrawn within the head proper, or ex- tended from two to three inches beyond. Extension is accom- plished by the turgescence of the tissues, while flexion and retrac- tion are effected by the flexor and retractor muscles of the probos- cis, and by the contraction of the outer circular muscle fibres, by which the blood may largely be removed from the vascular spaces rad. (d.) rad. (v.) of the organ. FiGuRE 2.— Cross section of the proboscis. : } oe., esophagus; s. @., salivary duct; m., The mouth proper 1s provided muscular coat; wv. c.¢., vascular connec- with the chitin-like radula, and tive tissue; ¢c. m. 7., circular muscular the openings of the salivary ducts. layer; 7. m., retractor muscles; c. f/., cartilage plate; vad. (d.), radula (dor- By means of the former, playing sal) ; vad. (v.), radula (ventral). over the end of a cartilaginous plate, and drawn backward and forward by the radulary muscles, Sycotypus is enabled to penetrate the shells of its prey, and to com- minute its food.! The ducts, opening into the mouth on either side, 1 It is a matter of physiological interest to note that although haemoglobin is lacking in the blood of this animal, being replaced by a copper-bearing respiratory proteid, —a hemocyanin, —the muscles operating the radula are brightly pigmented with hemoglobin. With the exception of a slight amount in the heart musculature, and in the ganglia and main nerve trunks, no other tissues of the body contain this latter pigment. Undoubtedly the regular contractions of these muscles for long periods of time, as during the process of shell perforation, or the comminution of food, require a more thorough tissue respiration than the sluggish stream of hemocyanin could furnish. By means of the hemoglobin, these tissues may be enabled, during the periods of rest, to store up oxygen sufficient for the excess of demand over supply during work. In other words, the presence of hemoglobin in these muscles strongly suggests an adaptation to the heightened internal respi- ration attendant upon their unusual contractile changes. This same local pig- mentation with haemoglobin is not confined to this particular species. The “drill,” Urosalphinx cinerea, which operates in much the same way, is similarly provided with hemoglobin-pigmented radulary muscles, while many other molluscs have ganglia and nerve trunks similarly pigmented. 20 Lafayette B. Mendel and Harold C. Bradley. carry the secretion from the salivary glands: the two large, yellow, secreting glands of the head. It has been assumed from the size and position of these glands that they secrete a solution to assist in shell- perforation, — possibly hydro- chloric acid or some similar sol- vent for calcium salts. As will be vt, demonstrated later, this is not the case; the function of the salivary secretion appears to be primarily sz digestive. From the mouth, the cesopha- gus extends along the dorsal side of the proboscis (cf, Higge2ye passes through the circum-ceso- phageal ganglia, bends sharply to follow the general spiral arrange- FiguRE 3.— Section of salivary gland. @, ment of the animal, and enters ote ee t., vascular connective tissue ; theuistomach: scloccuncomnne ‘peri- 5. 4, secreting tubules. cardial sac. Near the ganglia, the cesophagus is invaginated to form the crop, —a small organ of obscure function. Section of the cesophagus anywhere along its length, shows it to be a tube of simple ciliated epithelium, rather rich in mucous cells, and surrounded by a thick muscular layer consisting chiefly of circular fibres. Its function is apparently compar- able with that of the cesophagus in higher animals, — 7. ¢., the transpor- tation of the insalivated food mixture from the mouth to the stomach. As was demonstrated experimentally, 4! aS ES FEN EY SONS the mucosa apparently contains neither enzymes nor activating FIGURE 4.—Section of crop. m. Z, muscular layer; /Z., lumen; JZ. ¢., lumen of crop. kinases. The stomach is that portion of the alimentary canal which lies super- ficially embedded in the liver, or so called hepato-pancreas. It forms a distinct loop, with a shallow cacum at the angle. Its lumen is considerably greater than that of the cesophagus or of the intes- Experimental Studies on the Physiology of Molluscs. 21 tine proper. At the same time a sheath of vascular connective tissue in part replaces the muscular coat so prominently developed in the cesophagus. are employed chiefly in the evacuation of the undigested residues after diges- tion is completed. The most characteristic feature of the stomach mucosa is the number and complexity of the rugz, which nearly fill the lumen of the cardiac portion, where they are most highly developed. Their function is evidently to retard the progress of the digestion mixture through the stomach and to offer a max- imum surface for absorption. They are thus quite analogous to the intestinal Peristaltic movements of the stomach are minimal, and FIGURE 5.— Section of cesopha- gus. m. /., muscular layer ; Zum., \umen. villi of the vertebrates. The products of proteolysis and, to amuch less degree, fats and carbohydrates are presumably absorbed by them. FIGURE 6.— Section of cardiac portion of the stomach, showing the ruge. s.,sinus; ep., epithelial coat; v.c.¢., vascular con- nective tissue; #. /., muscular layer; /.2., liver tubules; fzg.c., pigment cells. As the pyloric end of the stom- ach is approached the character of the rugze becomes much more simple, though still adapted to ready absorption. No secreting structures such as the glands of Lieberkiihn occur in the stomach, and no digestive fluid of any kind is elaborated by it, according to our experience. Three large diverticula — the hepatic ducts — open from the stomach into the liver mass, and rapidly divide into smaller ducts, and finally into the minute tubules of that organ. These diverticula can hardly be regarded as mere ducts for the hepatic secretions, since the characteristic enzymes of the liver are not found in the stomach contents to any large extent. They apparently serve as ceca in which the final hydrolysis of fats and carbohydrates largely occurs, and where the products of 22 Lafayette B. Mendel and Harold C. Bradley. this hydrolysis are absorbed and retained.’ The liver is undoubtedly a secreting organ, as its highly complex glandular epithelium would suggest; but it is equally an organ adapted to digestion, absorption, and retention. A most unique feature of the liver, in this connec- tion, is its function in storing up the metals zinc, copper, and iron. The presence of large quantities of the former two metals in the he- patic tissue has already been noted elsewhere,” and a more complete discussion of the liver must be reserved for a later paper. The pre- sence of the zinc and copper seems to have no effect upon the diges- tive function of the gland with which this paper is more immediately concerned. As will be evident from the experimental data, the liver secretes the enzymes which hydrolyze fats and car- bohydrates, viz., a lipase, an amylase, and an invertin. No proteolytic activ- ity could be detected. The intestine extends from the pyloric sphincter to the anal papilla in the mantle cavity. It is a thin walled tube of simple epithelium, surrounded by a sheath of vascular connective tissue. Absorption may go on here, but only to FicurE 7.— Section of the tip of 4 Jimited extent, since the digestive re- one of the ruge more highly 2 : : : magnified, showing the mech- Sidues are retained in the stomach until anism for rapid absorption. the absorption there is complete, where- ep. epithelium; s., sinus; 47. ypon the contraction of the muscular blood-vessel; z. ¢. 7, vascular = 3 coat of the stomach, and the opening of the sphincter allow the entire mass to be evacuated. No typical digestive enzymes were discovered in the intestinal wall. Digestion experiments. — In carrying out the digestion experiments upon which the preceding conclusions are based, every attempt was made to exclude as far as possible the influence of disturbing factors, such as putrefaction, bacterial contamination, etc. Control trials were conducted in each case with boiled portions of the various connective tissue. 1 It is of interest in this connection to note further, that while glycogen is found in abundance in the muscles of this animal, and is undoubtedly the form in which part of the carbohydrate of the food is ingested, none could be obtained from the liver. Dextrose and fats are, however, always present in abundance. 2 H. C. BRADLEY: Science, 'N. S. 1904, xix, p.-196. atin. rite ek es Experimental Studies on the Physiology of Molluscs. 23 glandular extracts used; and where doubtful results were obtained, the experiments were repeated. The extracts were made with the organs removed from the living specimens. The usual procedure was to grind the gland with fine sand and extract with a 2 per cent sodium fluoride solution, or in some cases with chloroform-water or glycerin, and filter at once. To follow as closely as possible the conditions obtaining in the actual environment of the animal, all digestions were made at a temperature of about 15°C., and in solutions having the ampho- teric reaction of the normal organs. The first experiment was undertaken to demonstrate the nature and reaction of the salivary secretion. A fresh section of the resting glands reacts slightly alka- line to litmus. For the pur- .pose of obtaining the actual secretion, a capillary cannula was inserted into the sali- vary duct of a living animal, and electrical stimulation applied to the gland and to FIGURE S:— Section of liver tissue, showing the Ae d scale od secreting tubules of narrow columnar epithelial aa 2 metho cells, surrounded by connective tissue rich in which Krause! employed the black pigment cells characteristic of this successfully in obtaining the animal and in which a large part of the copper : 3 of the liver is stored. sec., secretion; fig. c., salivary secretion of Octo- nee c eee ; pigment cells; sec. ¢f., secreting epithelium ; pus. The operation was nec- é. v., blood-vessel. essarily a difficult one, and only a small amount of the secretion was obtained. It was a clear liquid, ropy from its high content of mucin, and reacted slightly alkaline to litmus. Several series of experiments were then undertaken to locate the presence of proteolytic enzymes and kinases activating them. Series I.— Extracts of the salivary glands, cesophageal mucosa, liver, and intestinal mucosa were made. Inasmuch as the stomachs of many of the specimens were found distended with a digestion mixture, the contents of several were removed, made up with sodium 1 R. KRAvUSE: Centralblatt fiir Physiologie, 1895, ix, p. 273. 24 Lafayette B. Mendel and Harold C. Bradley. fluoride to a strength of 2 per cent, and tested along with the gland extracts. Further, to discover whether the blood itself contained proteolytic ferments in any quantity, fresh portions of it also were examined. The method used was that devised by Fermi,! and em- ployed later by Bayliss and Starling? in their studies of pancreatic proteolysis. It consists in subjecting tubes filled with colored gelatin —sterile and rendered antiseptic with sodium fluoride — to the action of the various extracts. The gelatin is liquefied whenever exposed to a solution containing a proteolytic enzyme, and the comparative activities of the various extracts may then be expressed in milli- metres of gelatin dissolved, provided the tubes are of the same calibre. The trials were continued for thirty-six hours in the cold (15°C.) with the results summarized below: TABLE I. DIGESTION TRIALS WITH GELATIN. | : | Gelatin || No. Extract used. Gelatin Extract used. dissolved. dissolved. Stomach contents. . a 1 | Mixture of 6+3-+44. Dowboitled@a cma ger || | Do. boiled Salivary extract . . { | Mixture of 1 + 2 IDYoy loyesliexol 5 9 Sh t Do. boiled Stomach mucosa . . i | Mixture of 6 + Ls 2 Domboiledaaiey aati | Do. boiled Intestinal mucosa. . t | | Mixture of 2+ 5 Do} boiled™ = e+. ! | Do. boiled Vere oe } Blood Dos boiled’) .) =.) 2 ! Do. boiled (Esophageal mucosa . Do. boiled 1 A digestion trial made at the same temperature with an active preparation of KUHNE’s dry pancreas powder, showed scarcely more vigorous pie emae action than did the salivary gland extract used above. It is evident that the stomach contents and the salivary extract are both proteolytically active; but since the stomach mucosa, the liver, and the mucosa of the cesophagus failed to indicate any proteo- 1 Ci. FERMI: Archiv fiir Hygiene, 1891, xii, pp. 238-260. * BAYLIss and STARLING: Journal of physiology, 1904, xxx, p. 61. Experimental Studies on the Physiology of Molluscs. 25 lytic enzymes, we conclude that the ferment present in the contents of the stomach was merely that of the salivary secretion present with the food. The method of Fermi may be open to the objection that gelatin is employed, and that different results might be looked for by using a true proteid instead of it. A second series was therefore undertaken to eliminate this possible source of error. TABLE If. DIGESTION TRIALS WITH FIBRIN. “ Non-coagu- lable Nitro- gen” in digestion fluids. Nitrogen in tannic acid filtrates. Nature of the yeaa 7 extracts. Un- | ; boiled) ooo extract. J - . U Wy Boiled | boiled | _— extract. |extract. No. of trial. Amount used. Dry fibrin used. Difference. extract. =yies || Nn? | gram gram gram gram 0.094 | 0.058 0.033 | 0.024 | 0. bil 4 re] B =e (op) . os Difference. ‘OB + Salivary (Esophageal . . . .| “ 0.038 | 0.033 | 0.022 | 0.024 | Stomach contents . .| “ ie 0.018 | 0.027 0.011 | Stomach mucosa . .| “ 0.036 | 0.040 0.011 Intestinal mucosa . .| “ : 0.027 | 0.022 +a: a 0.045 | 0.045 Salivary + intestinal . | 0.042 | 0.000 Stomach mucosa + cesophageal mucosa.| “ | “ | 0.038 | 0.040 Liver + intestinal | MMTEOSAI «4 se | 0.040 | 0.036 PAROCMMEECME at sy ge cee 15 “« | 0.016 | 0.019 Series II.— For these trials extracts with 2 per cent sodium fluo- ride solution were made as before, using, as nearly as possible, com- parable amounts of the tissues, and equal amounts of the solvent. Weighed portions of finely chopped fibrin were added, and the diges- tions continued for twenty-four hours. At the end of the period all the mixtures were brought to a boil to arrest the action of the ferments, the coagulated proteids filtered off, and a measured portion of the filtrate, representing an aliquot of the original digestion mix- 26 Lafayette B. Mendel and Harold C. Bradley. ture, analyzed for (a) total nitrogen in non-coagulable compounds ; (b) nitrogen in compounds not precipitated by tannic acid, 2.é., amido-acids, etc. The actual extent of digestion was ascertained by comparison with figures obtained in a similar way from the boiled (control) extracts. The results are summarized in the table on page 517. As will be seen from the table, 1 and 7 are the only trials in which the amount of non-coagulable nitrogen exceeds that of the control by a significant amount. (Indeed for the purposes of this comparison, the figures beyond the second decimal place are negligible.) In both of these digestions the extract of the salivary gland is evidently the active factor. There is no indication of the presence of an activ- ating kinase in any of the mixtures, since the simultaneous action of various extracts shows no increased proteolysis. A salivary extract was allowed to act for forty-eight hours upon fibrin in the cold to determine whether the enzyme produced leucin and tyrosin, under those conditions. As is indicated by the table above, only very small traces of the amido-acids were formed during a twenty-four hour period, and similar results were obtained from the longer digestion. Presumably the products of proteolysis in Sycotypus are absorbed as proteoses and peptones, and scarcely reach the stage of cleavage in which the amido-acids appear. The digestion mixture TABLE IIT. SHOWING THE RELATIVE AMOUNTS OF STARCH PASTE LIQUEFIED IN TRIALS WITH VARIOUS TISSUE EXTRACTS. (The figures indicate centimetres.) Salivanyores Weel eee : 5 | Intestinal mucosa (Esophageal Stomach mucosa . Stomach contents gave a strong tryptophan reaction at the end of forty-eight hours, the salivary enzyme resembling trypsin in this respect. Series III.— In searching for amylolytic enzymes, extracts were made in the same manner as described above. Tubes of a thick starch paste were prepared and subjected to the action of the various extracts and their mixtures, with frequent shaking. As in the case Experimental Studies on the Phystology of Molluscs. 27 of the Fermi gelatin tubes, liquefaction of the paste was taken as the indication of digestion. From the nature of the experiment, the figures are only approximate. The trials were continued for five hours, and the control tubes (subjected to boiled extracts) failed to show liquefaction of the starch in any case. The liver extract is thus seen to contain the active amylolytic enzyme, while the other extracts gave negative results as a rule. The fact that the stomach contents showed a slight starch-digesting power is attributable to the presence of a small amount of the hepatic secretion; the slight degree of activity is rather surprising and indi- cates that the largest part of the carbohydrate digestion takes place in the hepatic ducts. A further indication of this is to be found in the lack of the hepatic pigments in the stomach contents. As a rule, the digesting mass in the stomach is a colorless, viscid, fluid or semi-fluid material. The contents of the hepatic ducts, on the other hand, are usually colored dark brown from the pigments present in the hepatic secretion. The slight liquefaction seen in No. 5 is due probably to some of the hepatic enzyme retained mechanically from a previously evacuated digestion residue. In a prolonged starch digestion, the liver extract was found to produce a reducing sugar, which gave an osazone whose crystalline form and melting point (205°) were characteristic of dextrosazone. Thus the liver enzyme of Sycotypus is comparable with the dextrose- forming hepatic enzymes of the livers of higher animals, rather than with ptyalin. Series IV. — In another series of trials (cf. p. 28) a weighed amount of glycogen in solution was substituted for the starch paste, and the amount of reducing sugar formed at the end of the period (and deter- mined by the Allihn gravimetric method for dextrose), was taken as the measure of amylolytic activity. The results confirm those of Series III. A qualitative experiment was made to determine whether or not the liver extract contained an invertin, with positive results. The formation of invert sugar from saccharose was so obvious that a quantitative determination was not made. Series V. — Finally the lipase of the liver was demonstrated by a method in common use where comparative rather than quantitative results are desired. A blue litmus-milk emulsion! was made, and 1 This method was first suggested by HEIDENHAIN. Cf ROEHMANN: Chem- sche Arbeiten fiir Mediciner, 1904, p. 53. 28 Lafayette B. Mendel and Harold C. Bradley. to similar portions the various extracts were added. The liberation of free fatty-acids from the neutral milk fats reddens the litmus emulsion ; and in this case the results were so distinct that no more accurate method was applied. (Cf Table V.) TABLE, IV. DIGESTION TRIALS WITH GLYCOGEN. Amount Total dextrose Extract used. falkent (calculated). c.c. gram SHInISG 6 6 6 b 29 0.09 Salivary boiled . . . 25 0.04 [eiverss) sober nee: 25 0.72 Liver boiled ... . 25 0.19 1 Liver + Salivary. . . 25 0.48 1 The presence of a reducing sugar in the liver itself is indicated by the boiled control. It also was found to be dextrose. TABLE V. FAT-DIGESTION TRIALS. Color noted in Amount of blue litmus |—— =e extract. | milk. 1 hour. 3 hours. 6 hours. Extract used. | Amount of | c.c. every cee | Blue Red Red Liver boiled .. . Blue Blue Blue Salivary gland .. | Blue Blue Blue Salivary gland boiled | Blue Blue Blue Summary. — Digestion in Sycotypus is effected by the secretions from (1) the salivary glands, and (2) the liver or hepato-pancreas. The salivary glands, closely resembling histologically the com- parable glands of higher animals, secrete a ropy mucin-laden solu- tion containing a proteolytic enzyme. This acts normally upon proteids in the cold, and in solutions of a-neutral or amphoteric Experimental Studies on the Physiology of Molluscs. 29 reaction. In its behavior and in the characteristic decomposition products which it produces, it resembles trypsin. The liver, a highly glandular organ with very characteristic secret- ing epithelium, elaborates enzymes which effect the hydrolysis of carbohydrates and fats: an amylase, invertin, and a lipase. The amylase is similar to the characteristic amylolytic enzymes of the vertebrate liver. Digestion takes place in the stomach proper, and in the so-called hepatic ducts. In the stomach, the proteids are broken down to proteose and peptone, and the products of digestion absorbed through the rugz. Fats and carbohydrates are largely hydrolyzed in the hepatic ducts, where also the products of their hydrolysis are absorbed and retained. SOME NOTABLE CONSTITUENTS OF THE URINESO@S ‘ie “COV. Oins BY ROBERT E. SWAIN: [From the Chemistry Laboratory of the Leland Stanford Jr. University.] i the course of an investigation of kynurenic acid as a product of excretion in the urine of the dog, an effort was made to find this interesting substance in the urine of some animal other than the dog, to which it has thus far seemed to be wholly restricted. Various at- tempts have been made to detect it in the urine of the rabbit? and of the cat ® under conditions of diet which have appeared to be most favorable to its formation, but without success. Hofmeister * prob- ably made the first careful research for it in human urine with nega- tive results, and Capaldi°® failed to find it in the urine of the related animals, the wolf and fox. Pursuing this line of investigation further, it seemed that good re- sults might come from a study of the urine of the coyote, an animal closely related to the dog, which inhabits the arid districts of the western part of North America. In general appearance the coyote strongly resembles the dog. It is lean and slender, — weighing from 15 to 30 kilograms, — yel- lowish gray in color, and combines the cunning and swiftness and the easy, graceful movements of the fox with the greed of the wolf. 1 The results presented in this paper form a part of a doctoral dissertation on “ The formation of kynurenic acid by the animal body,” embodying work carried on under the direction of Professor LAFAYETTE B. MENDEL and presented to the faculty of Yale University, May 1, 1904. As this investigation of the urine of the coyote was in a measure only incidental to the subject in question, the results are thus given in a separate paper. * MENDEL and JACKSON: This journal, 1898, ii, p. I. ® MENDEL and JACKSON: Loc. cit. 4 HOFMEISTER: Zeitschrift fiir physiologische Chemie, 1881, v, p. 67. 5 CAPALDI: Zeitschrift fiir physiologische Chemie, 1897, xxiii, p. 87. 30 Notable Constituents of the Urine of the Coyote. 31 It is wary and hard to approach by day, but much bolder at night, when it will frequently venture near to human habitation in commit- ting its depredations, the especial prey being domestic fowls, young lambs, and calves. It shows remarkable cunning in avoiding traps, and is rarely caught in this way, even after every precaution is taken by the trapper. Until late years it was customary to regard the coyote as a single species ranging over the whole of western North America, from Canada to the table lands of Mexico, and from the western part of the Mississippi Valley to the Pacific Ocean; but in a recent paper Mer- riam! maintains that the name “ coyote,” as popularly applied, really covers an assemblage of species which comprise three well-marked subordinate groups and a considerable number of distinct geographi- cal forms. According to his classification, the specimen made use of in this investigation belonged to the “ Microdon” group, of the form known as Canis ochropus Eschscholtz, ranging over the interior val- leys and lower foothills of central and southern California. It was a young, but full-grown male weighing 24 kilograms. The animal was confined in one of the laboratory dog cages, which are arranged for the separate collection of urine and feces, and fed 250-350 grams of lean meat for several days, until it had become ac- customed to its surroundings and to the regular diet. The urine was then collected in daily portions in clean, stoppered bottles, thymol added to prevent bacterial decomposition, and the samples thus pre- served for analysis. Kynurenic acid was found in every urine sample examined. Al- lantoin was also present in every case, and in addition to these there appeared a crystalline substance whose composition and properties would go to show it to be some new constituent of carnivorous urine, though it very nearly corresponds with Jaffé’s ‘‘ Urocaninsaiire” in composition and general properties. Three of the samples collected on succeeding days with the animal on a constant diet of 300 grams of meat and 50 grams of lard gave the following results. Kynurenic acid was determined after the method of -Capaldi,? and allantoin according to Loewi’s? method. The 1 MERRIAM: Proceedings of the Biological Society of Washington, D.C., 1897, xi, Pp. 19-33. 2 CAPALDI: Zeitschrift fiir physiologische Chemie, 1897, xxiii, p. 92. 8 Lorwr: Archiv fiir experimentelle Pathologie und Pharmakologie, 1900, xliv, p. 20. 32 Robert E. Swatn. latter substance was identified by its crystal form and its melting point. | Urine Kynurenic : Allantoin. volume. Sample. The remaining samples were combined and a portion used in making a general chemical examination, since the coyote is a representative of a numerous species of carnivorous animal whose urine has probably not heretofore been analyzed. The total nitrogen was estimated according to the Kjeldahl-Gun- ning method; urea, by the method of Morner and Sjoquist (since the presence of allantoin, according to Morner,! does not materially affect the results); uric acid, by Ludwig’s method; allantoin and kynurenic acid as above. The following data were obtained as an average of closely agreeing duplicate analyses: Sear, Go o 6 co o 3 JOR errant & @ o 6 O16 6 Go elitanidbversal Potalnitrofeny 9) 1) Sosoleramssperlitre: Wien Go Wo 6 5 6 & oo 6 (AUHIKEP a Witch pai a 6 a o 5 6 so ~COMDIS & * IMlemen otis 6 6 6 6 a 6 5 Wyte > UGwimem@aercl 5 6 5 6 o 6 a Oeyas oS Motalssulphuriciacid yyy cents 2 99D aes ¥ Ethereal sulphuric acid . ... O127 “ a Allantoin has previously been found in the urine of suckling calves, occasionally in that of the dog, cat, and in human urine. Much in- terest has centred in allantoin as a constituent of animal urine, especially in fegard to its origin. That its amount is appreciably increased by the ingestion of food rich in nuclear material, as thymus or pancreas, has been conclusively demonstrated. The main question now at issue is its relation to uric acid. Although it is easily pre- pared in the laboratory by the gentle oxidation of uric acid by lead 1 MOrNER: Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 297. Notable Constituents of the Urine of the Coyote. 33 peroxide, mercuric oxide, and other oxidizing agents, its origin from uric acid in the animal body is yet a disputed question.! The high specific gravity of the urine is unusual as an average over so long a period. Doubtless the natural habits of the animal con- tribute much to this. Its favorite haunts are the arid desert regions of the Pacific slope, where water is very scarce even in winter. In captivity, with fresh water always provided in the cage, it drank sur- prisingly little. Fat was never touched until the last morsel of lean meat was devoured, and dog cakes, cracker meal, and carbohydrate food in general were not relished and often refused. During the interval of nearly two months between the time of collection and the analysis of the urine a small crystalline deposit appeared in each sample. An examination under the microscope showed a homogeneous mass of thin, crystalline, hexagonal plates, very similar to those formed by cystin. They slowly dissolved in alcohol and in hot water, but not in ether. On adding sulphuric acid or nitric acid to the hot water solution, there separated beautiful colorless crystals, which differed in crystal form for the two acids, and were also unlike the crystals of the original substance. After washing with water and alcohol until no considerable trace of acid could be detected in the washings, the crystals gave strong tests for sulphuric acid and nitric acid, respectively, and evidently were salts formed by direct union with the acids. In order to obtain more of the crystals, two litres of the urine were evaporated to small volume on the water bath, strongly acidified with sulphuric acid and allowed to stand forty-eight hours. The precipi- tate was filtered off and decomposed with hot baryta water. After removing the precipitated barium sulphate, and concentrating the fil- trate, about 600 milligrams of colorless crystals separated on standing over night. These were combined with the first yield, the whole re- crystallized from hot alcohol (60 per cent) and analyzed. The crystals melted at 208°, and at 220° began to decompose rapidly, as shown by the liberation of considerable carbon dioxide, along with other gases. The result of the analysis was as follows: 0.1026 gram dried to constant weight at 110 lost 0.0213 gram = 20.86 per cent. 1 See “Ergebnisse der Physiologie,” i, Erste Abteilung, pp. 624-627, for a general discussion of the subject. Also MENDEL and Brown: This journal, 1900, lii, p. 267; SwAIN: Jdid., 1901, vi, p. 38; MENDEL and WHITE: /dzd., 1904, xii, p- 85; KuTscHER and SEEMANN: Centralblatt fiir Physiologie, 1904, xvii, p. 715- 34 Robert E. Swain. 0.1182 gram dried to constant weight at 110° lost 0.0246 gram = 20.81 per cent. The rest of the substance was dried, combined with the above portions, and analyzed. The following data were obtained: it II. Calc. as Cy>HgN,O, Gre. Pee <. ars 52.60 52.70 Hie CA) 4 Bole (OSE 3.26 3.33 AT) cc Sy co a eae 2009 20.60 20.55 O'(by difjs = =. 2338 23.53 23.42 The crystalline substance would then have the formula Cp NOs 4 @- In many of its properties, and very nearly in its composition, this substance corresponds to a substance found first by Jaffé,! and re- cently by Massot,? in the urine of the dog. At the same time it must be admitted that the difference in composition shown by the elementary analysis is sufficient to indicate that this substance is some new constituent of carnivorous urine. Jaffé gave the name ‘“‘Urocaninsdure” to the substance isolated by him, and estimated its formula to be C,,H,,N,O,* 4 H,O. He was unable to find it in the urine of most dogs, and the natural inference has been that it was a casual constituent of this urine which arose either from some meta- bolic idiosyncrasy of the dog in question, or as a result of nitrototuol feeding, to which the animal had been previously subjected. In regard to the latter possibility, however, Jaffé insists that the dog was perfectly healthy, and that all effects of the nitrototuol feeding had vanished months before. Moreover, other dogs failed to excrete the substance after prolonged administration of nitrototuol. As soon as another coyote is available, it is the purpose of the writer to isolate more of this substance and to determine its properties more fully. 1 JAFFE: Berichte der deutschen chemischen Gesellschaft, 1874, vii, p. 1669; and 1875, viii, p. S11. 2 See SIEGFRIED: Zeitschrift fiir physiologische Chemie, 1898, xxiv, p. 399. me CHEMISTRY OF THE PROTEIN-BODIES OF THE WHEAT KERNEL! PART I.—THE PROTEIN SOLU- fPePeiN ALCOHOL AND ITS: GLUTAMINIC ACID CONTENT. BY THOMAS B. OSBORNE anp ISAAC F. HARRIS: [From the Laboratory of the Connecticut Agricultural Experiment Station. | S a result of an extended investigation of the proteids of the wheat kernel, Osborne and Voorhees? concluded that only one protein substance soluble in alcohol was present in this seed. Fleurent ® also held the same view. . Morishima’s# investigations led him to believe that wheat gluten contained but a single protein, and that the glutenin and gliadin of Osborne and Voorhees were derivatives of one and the same protein substance, which he named artolin. Ritthausen ° again asserted his belief in the existence in this seed of three distinct protein-bodies that were soluble in alcohol, but offered no new evidence of their existence. Kossel and Kutscher,® following Ritthausen’s directions, prepared the proteins of wheat gluten, and determined the proportion of basic products which they yielded on decomposition with acids. They found that the protein insoluble in alcohol, called by Ritthausen gluten-casein, by Osborne and Voorhees glutenin, was sharply dis- tinguished from the protein soluble in alcohol by the fact that, on 1 This paper is the first of a series giving the results of an investigation of the nature and amount of the primary decomposition products of the protein con- stituents of the wheat kernel. The expense of this work has been met by a grant from the Carnegie Institution. 2 OSBORNE and VOORHEES: American chemical journal, 1893, xv, p. 392. 8 FLEURENT: Comptes rendus de l’académie des sciences, 1896, cxxiii, p. 755. 4 MorIsHIMA: Archiv fiir experimentelle Pathologie und Pharmakologie, 1898, xli, p. 348. § RITTHAUSEN: Journal fiir praktische Chemie, 1899, lix, p. 474. 6 KossEL and KuTscHER: Zeitschrift fiir physiologische Chemie, 1901, xxxi, p- 165. 35 36 Thomas B. Osborne and Isaac F. Harris. decomposition, it yields a notable quantity of lysine; whereas all their products derived from the alcoholic extract of the gluten yielded none of this diamino acid. This fact shows conclusively that Mori- shima’s view of the identity of all the protein-products obtained from wheat gluten is incorrect. They further found that Ritthausen’s so-called mucedin, gliadin, and gluten-fibrin, prepared from the alcoholic extract of wheat gluten, yielded somewhat different proportions of histidine and arginine ; but, in view of the methods employed for the determination of these bases, they considered these differences too small to fully justify the conclusion that these were distinctly different protein substances. They conclude their paper, however, with the statement that their experiments confirm the views of Ritthausen respecting the compo- sition of wheat gluten. Kutscher ! extended the investigation of these proteins, and deter- mined in the solutions which remained from Kossel and Kutscher’s determinations of the bases the amount of tyrosine and glutaminic acid. As a result of this work, he concluded that mucedin and gliadin were one and the same protein ; but that the gluten-fibrin yielded an amount of glutaminic acid so much less than that given by the other fractions as to leave no doubt that this was a distinctly different substance. Nasmith ? concluded that only one alcohol-soluble protein was pres- ent in wheat gluten. He also stated that, although this protein, gliadin, invariably contains phosphorus, it is not a nucleo-proteid. In passing, we may say that this statement is erroneous, for an exam- ination of a large number of our preparations has shown them to be wholly free from phosphorus. Very recently Kénig and Rintelen® have published a brief account of the results of an investigation which they have made to determine the nature of the proteids of wheat gluten, and conclude, with Ritt- hausen, that there are three present which are soluble in alcohol. As a result of all these later investigations, it would appear that glutenin is sharply distinguished from the alcohol-soluble protein by the presence of lysine among its decomposition products, and that, 1 KUTSCHER: Zeitschrift fiir physiologische Chemie, 1903, xxxviii, p. III. 2 NASMITH: ‘Transactions of the Canadian Institute, 1903, vii. 8 KOnIG and RINTELEN: Zeitschrift fiir Untersuchung der Nahrungs- und Genussmittel, 1904, viii, p. 401. Chemistry of Protein-Bodies of the Wheat Kernel. 37 o notwithstanding the close agreement in ultimate composition, these two proteins do not have a common origin, as Morishima supposed. Respecting the existence of more than one protein soluble in alcohol, all recent investigators, except Kossel and Kutscher, and Konig and Rintelen, agree with the view formerly advanced by the writer. For the purpose of this investigation of the decomposition products yielded by the proteins of the wheat kernel, it has become necessary, as the first step, to examine carefully the new evidence which has been presented respecting the existence of Ritthausen’s gluten-fibrin and mucedin. The first difficulty encountered lay in the impossibility of following Ritthausen’s directions for preparing the several alcohol-soluble proteins which he has named, since many evidently important details are omitted in the description of these methods. Kossel and Kutscher state that their products were made according to these directions, but give no details, nor do they state which method they employed. Kutscher concludes his paper by the state- ment that “the wheat gluten consists of gluten-casein wholly in- soluble in cold 60 per cent alcohol; gluten-fibrin, but little soluble, and gliadin, easily soluble in cold 60 per cent alcohol.” Although we have made a very large: number of preparations rep- resenting fractions of this protein substance, dissolved by alcohol of various degrees of strength, we have never obtained any that were not either completely soluble in cold alcohol of 60 per cent, by volume, or else contained such insignificant quantities which did not dissolve, that we have found it impossible to make from them a preparation of “ gluten-fibrin”” suitable for further examination. We have, therefore, been unable to repeat the work of Kossel and Kutscher and are entirely at a loss to understand how their preparation of “oluten-fibrin” was obtained. Konig and Rintelen describe their procedure in more detail. These investigators extracted wheat gluten with absolute alcohol, added ‘ether to the alcoholic extract, and united the precipitate produced with the extracted gluten. This latter was then extracted with 65 per cent alcohol, and to the extract alcohol was added until the mixture contained 88-90 per cent of alcohol. After decanting from the precipitate that had formed, the solution was filtered clear and evaporated to dryness on the water bath, finely pulverized, and extracted with ether, to remove fat. As all of the fat could not be thus removed, the mass was again dissolved in alcohol, 38 Thomas B. Osborne and Isaac F. Harris. to which some caustic potash was added, and this solution shaken out several times with ether. The weakly alkaline solution was then exactly neutralized with hydrochloric acid, and evaporated on the water bath. The product thus obtained was their “ gluten-fibrin.” The precipitate produced by 88-90 per cent alcohol was washed with alcohol of the same strength and dissolved in a little 65 per cent alcohol. From this, one-half of the alcohol was distilled off and the residual solution cooled, when a precipitate separated. From this the solution was decanted, leaving a mass of gliadin. The solutions which remained from several such precipitates were united and dis- tilled, until one-third of the solvent was removed. On cooling the residual solution, a deposit formed which they considered to be a mixture of gliadin and mucedin. The solution decanted from this deposit was evaporated to dryness, and yielded a considerable residue of “ mucedin.” The analyses of these products showed that the gliadin thus pre- pared had the same composition as that obtained by us, as well as that made by Ritthausen, while the “ gluten-fibrin” and mucedin contained about 1 per cent less nitrogen and much more carbon. That the “ gluten-fibrin” thus made by repeated evaporation at a high temperature with water,:and solution in caustic alkali, can repre- sent an unaltered constituent of the wheat kernel, seems to us highly improbable, especially in view of the fact that gliadin contains a very large proportion of amide nitrogen ; z. ¢., about one-fourth of its total nitrogen, which would probably be easily split off by caustic alkalies. Furthermore, although gliadin can be heated in strongly alcoholic solutions without apparent change, when heated in aqueous solutions a not inconsiderable part is altered. We have never attempted to produce “ gluten-fibrin” by such a process as here described, as we are convinced that products thus prepared cannot certainly be regarded as unaltered constituents of the wheat kernel. We have, however, many times isolated the pro- teid soluble in the strongest alcohol; but although these preparations have represented very small fractions of the total protein under ex- amination, and have been obtained from solutions in very strong alcohol, they have always shown the same properties and composition as the other purified preparations of gliadin. Konig and Rintelen obtained their mucedin from the nearly aqueous solutions remaining after separating the gliadin by evaporat- ing to dryness. We assume that the residue which remained was Chemistry of Protein-Bodtes of the Wheat Kernel. 39 subjected to some further purification, but concerning this they say nothing. We have always found that this solution contained many impurities, and that when the protein substance in it had been prop- erly purified, this had the properties and composition of gliadin. We have recently made large quantities of gliadin, and subjected it to very careful and extensive fractionation, but, as in the earlier work of Osborne and Voorhees, we have been wholly unable to obtain any evidence whatever of the existence of “ mucedin.” In this work we have used only the purest absolute alcohol, which we have ourselves prepared in large quantities, in order to avoid the presence of acids and other substances which might be otherwise introduced. Itis possible that the American wheats which we have examined differ from the European, but this we consider in the high- est degree improbable. Furthermore, it would seem improbable that our gliadin could be contaminated by “ gluten-fibrin ”” and ‘‘ mucedin,” which we certainly did not succeed in separating from it, and, at the same time, show so close an agreement in composition with that of Konig and Rintelen, from which they suppose that both of these pro- _teins had been carefully removed. It is also improbable that Kjeldahl should have found (a)p uniform for successive fractional precipitations of the alcohol-soluble protein, if the material which he examined was a mixture of three different substances, nor, if this were the case, could his determination of (a)p, —92°, be expected to agree so closely with ours. —92.3°. The composition of both “ gluten-fibrin” and “ mucedin” differs from that of gliadin just as one would expect if these former sub- stances were slightly altered, and somewhat impure products obtained from gliadin. Until more convincing evidence of the existence of “ gluten-fibrin ” and ‘‘mucedin” as distinct protein substances is brought forward, we cannot consider them to be original constituents of the wheat kernel. As fractional precipitation has wholly failed, in our hands, to yield products similar to the “ gluten-fibrin ”” and “ mucedin” of Ritthausen, we have sought to discover, if possible, differences in. the amount of glutaminic acid which our different preparations might yield, and so be able to judge of the value of the evidence presented by Kutscher respecting the existence of gluten-fibrin. 40 Thomas B. Osborne and Isaac F. Harris. EXPERIMENTAL PART. Two portions of different preparations of gliadin, each weighing 100 grams, were decomposed by boiling for fourteen hours with 200 c.c. of concentrated hydrochloricacid. The solution, cooled with ice, was saturated with gaseous hydrochloric acid and, packed in ice, kept in a refrigerator for three days. The entire solution, thus treated, solidified to a thick mass of crys- tals which were sucked out with a pump, thoroughly washed with ice- cold alcohol which, according to a suggestion made to us by P. A. Levene, had been saturated with hydrochloric acid. When dried over sodium hydrate, the two preparations thus obtained weighed, respec- tively 55.15 and 58.33 grams. The filtrate and washings, when con- centrated to asyrup, yielded by a repetition of the preceding process a second crop of crystals, weighing, respectively, 4.91 and 2.68 grams, making the total crude glutaminic acid in each case 60.06 and 60.01 grams. Since gliadin, when decomposed by boiling with acids, yields 4 per cent of nitrogen as ammonia, these products were examined for ammonium chloride. By crystallizing one of them from strong alcohol 9.72 grams of ammonium chloride, and 34.75 grams of nearly pure glutaminic acid hydrochlorate were obtained. As considerable loss of glutaminic acid occurred during this process, the other pro- duct was treated dry with absolute alcohol, and the residue washed thoroughly with absolute alcohol, dried, and found to weigh 5.64 grams. The solution and washings were concentrated to a syrup, taken up in water and freed from color by animal charcoal. The aqueous solution was then evaporated to small volume, and concentrated hydrochloric acid added in quantity. It was then evaporated until crystallization began. On cooling, the solution set to a solid mass of large crystals, which was sucked out with a pump and washed with ice-cold alcoholic hydrochloric acid. When dried over caustic soda, these crystals, fraction a, weighed 39.12 grams. The filtrate and washings from a, when concentrated, yielded 2.15 grams of crystals, 4, which examination showed to be nearly pure ammonium chloride, and the filtrate from this 1.42 grams, c, which was likewise ammonium chloride. Fractions 6 and ¢ were united and dried at 100°. 0.3370 gram distilled with MgO gave ammonia equal to 8.64 c.c. of HCl. 1 c.c. HCl=o,o100 gram N=25.63 per centune Calculated for NH,Cl, 26.22 per cent. ij | Chemistry of Protein-Bodies of the Wheat Kernel. 41 The solution from which a, 4, and c had been separated, was then boiled with a small excess of barium hydrate, to remove the remain- ing ammonia, and the barium removed with an equivalent quantity of sulphuric acid. On concentration with an excess of hydrochloric acid, this solution yielded @, weighing 4.43 grams, and the filtrate from d, on long standing on ice, gave 2.41 grams of e. A part of fraction @ was then distilled with magnesia. 0.6815 gram, dried at 100°, gave ammonia equal to 1.34 c.c. HCl. 1 c.c. HCl=o.o100 gram N. It gave ammonia, therefore, equal to 7.51 per cent of NH,Cl. 0.6541 gram of fraction @ dried at 100°, and treated according to Kjeldahl, neutralized 5.9 c.c. HCl. 1 c.c. HCl=0.0100 gram N. It contained, therefore, 9.02 per cent of N. Deducting. the ammonium chloride thus found, the remaining 92.49 per cent of fraction a contains 7.62 per cent of N. Calculated for C; H,NO,HCI, 7.64 per cent. From the results of this analysis, we find that fraction a contained 36.18 grams of glutaminic acid hydrochlorate. Fractions d@ and e were pure glutaminic acid hydrochlorate, as shown by their crystal- line form and nitrogen content. Fraction @. 0.5544 gram dried at 100°, treated according to Kjeldahl, neutralized 4.32 c.c. of HCl. 1 c.c. HCl=o.0100 gram N, equal to 7.79 percent N. Calculated for C,H,NO,HCI, 7.64 per cent. Fraction ¢. 0.5092 gram dried at 100° gave, according to Kjeldahl, ammonia equal to 3.86 c.c. HCl. 1 c.c. HCl=o.o100 Sram WN, equal to 7.58 per cent. Calculated for C,H,NO,HCI, 7-64 per cent. The amount of the glutaminic acid hydrochlorate thus found, was as follows: grams. itionekeayeiee BVS & 5 5B etopiks) Inefraction a." « sun ol + a= 4:40 Inetractiomie: 88s aos Sel 4 43.02 This is equivalent to 34.42 grams of free glutaminic acid. Since the gliadin decomposed contained 7 per cent of moisture, the amount of glutaminic acid thus found was equal to 37 per cent of the dry gliadin. This may be taken as a minimal figure, for the true proportion is certainly higher, as some glutaminic acid remained in the mother liquors and could not be separated in a pure state. 42 ,. .Lhomas B. Osborne and Isaac F. Harris. In confirmation of these figures, this determination was repeated with two fractions of the alcohol-soluble ‘protein of wheat gluten, which had been separated from relatively strong alcoholic solutions and should, therefore, have contained a large proportion of “ gluten- fibrin,” if the statements respecting the solubility of this substance are correct. Two portions of different preparations of the air-dry substance, equivalent to 18.62 and 14.65 grams dried at 110°, were hydrolized as before, freed from ammonia by evaporating with an excess of baryta, and from barium by an equivalent quantity of sulphuric acid. The solution was then decolorized with animal charcoal, and evaporated with an excess of hydrochloric acid until crystallization began. The glutaminic acid hydrochlorate which separated, when washed with ice-cold alcoholic hydrochloric acid and dried, weighed, respectively, 8.69 and 6.27 grams, equivalent to 37.33 and 34.2 per cent of free glutaminic acid in the protein. Another attempt was made to isolate a fraction of “ gluten-fibrin ” by dissolving 200 grams of a preparation, representing the total alcohol-soluble protein of wheat gluten, in a mixture of 900 c.c. of absolute alcohol and 600 c.c. of water; that is, in alcohol of 60 per cent by volume. Although the solution was somewhat turbid, nothing was deposited, even on long standing. We then added 300 c.c. of water to the solution, making the alcohol 50 per cent; but still nothing separated. The strength of the alcohol was, therefore, raised to 75 per cent, by adding 1800 c.c. of absolute alcohol, and a very large precipitate, A, at once separated. By adding 2000 c.c. of absolute alcohol to the clear solution from which A separated, another large precipitate, 2, was produced, and in the filtrate from A a third precipitate, C, resulted, when a further large quantity of absolute alcohol was added. This last product weighed 20 grams, and consti- tuted only 10 per cent of the total protein. The solution from which C separated contained only traces of protein, and C, therefore, repre- sented the fraction of the whole protein soluble in the strongest alcohol, and should, consequently, contain much “ gluten-fibrin.” In 17.5 grams of this substance, dried at 110°, we next determined the glutaminic acid produced by decomposing with hydrochloric acid. By proceeding in the same manner as in the experiment last de- scribed, we isolated 7.8914 grams of pure glutaminic acid hydro- chlorate, which is equivalent to 6.2131 grams of the free acid, or 35.50 per cent, ° Chemistry of Protetn-Bodtes of the Wheat Kernel. 43 0.7296 grams, dried at 110, gave ammonia equal to 5.66 c.c. Et@ls(1 cc. HC] = 0.0100 gram N) = 7.75 per cent N: Calculated for C-H,NO,HCl = 7.64 per cent N. This result seems to furnish conclusive evidence that there is no fraction, soluble in very strong alcohol, to be obtained from the alcohol-soluble protein of wheat gluten that is characterized by yield- ing a relatively small proportion of glutaminic acid. Since Kutscher decomposed his proteins with sulphuric acid, while we used hydrochloric acid in the preceding experiments, we made the following experiment, in order to determine whether the higher yield obtained by us might not be due to this fact. We accordingly boiled, for fourteen hours, with a mixture of 150 grams of sulphuric acid and 300 c.c. of water, 50 grams of one of the preparations of gliadin from which we had previously isolated 37 per cent of glutaminic acid. The resulting solution was treated with an excess of baryta, and the ammonia expelled by evaporation. The barium was then removed by an equivalent amount of sulphuric acid, and the filtered solution evaporated. Some tyrosine separated, which was filtered out, and the evaporation continued until the volume was quite small. On standing, an abundant quantity of crystals of free glutaminic acid separated, and from the mother liquor, by further concentration and standing, a second crop of crystals was obtained. After recrystallizing several times, 8.48 grams of pure glutaminic acid was obtained, which contained 9.43 per cent of nitrogen. 0.6094 gram substance gave, according to Kjeldahl, ammonia equal meets c.c. HCl-(1 cic. HCl=o:o100 gram N) = 9.43 per cent. Calculated for C,H,NO,, 9.48 per cent. From the mother liquors which yielded this glutaminic acid, we further separated, by saturating with hydrochloric acid, and proceeding in the manner already de- scribed, 4.055 grams of well crystallized hydrochlorate which con- tained 7.73 per cent of nitrogen. 0.4983 gram substance gave ammonia equal to 3.85 c.c. HCl @rc.c. HCl =0.0100 gram N) = 7.73 per cent N. Calculated for ©2H,NO,HCI = 7.64 per cent N. The free glutaminic acid corresponding to this amount of the hydro- chlorate is 3.244 grams, making a total of 11.724 grams. Since the 50 grams of air-dry gliadin were equivalent to 46.33 grams, dried at 110°, this amount of glutaminic acid is equal to 25.3 per cent, or about the same proportion as Ritthausen found in his preparation of 44 Thomas B. Osborne and Isaac Ff. Harriss. ‘“mucedin,” after decomposing with sulphuric acid, but much more than the 19.81 per cent found by Kutscher in the same substance. The amount, however, was only about two-thirds as much as we found after decomposing with hydrochloric acid, and although the separation was not complete, there was no reason to suppose that more remained in the mother liquors in one case than in the other. In conclusion, we here bring together the results of these deter- minations, that they may be more readily compared : PERCENTAGE OF GLUTAMINIC ACID YIELDED BY GLIADIN decomposed by HCl H,SO, Preps” =).ch es ee asi eal Z 3 4 5 37.00 37.33 SE 35.50 25.3 All these are minimal figures, since in each case some glutaminic acid still remained in the mother liquors, but we do not think that more than relatively insignificant quantities were lost, unless in the first separation from the mass of decomposition products. Respect- ing the amount which escaped separation from this solution, we know absolutely nothing. From these results it would appear: 1. That Kutscher’s determinations of glutaminic acid fall far short of the actual quantity of this substance yielded by the alcohol-soluble protein of wheat, and that they, therefore, afford no evidence which justifies the conclusion that this substance consists of two distinct protein bodies. 2. That fractional precipitations of this alcohol-soluble protein yield practically the same large proportion of glutaminic acid, so that, in view of their very close agreement in composition and properties, both physical and chemical, we have every reason to believe that only one such protein is present, for which, we think, the name gliadin should be retained. 3. That gliadin yields a remarkable proportion of glutaminic acid, much in excess of that from any other known protein, and greater than that of any single decomposition product yet obtained in a pure state from any other true protein substance, the protamines, of course, excepted. 4. That this very large proportion of glutaminic acid in a food protein, so extensively used, is a matter of great importance in rela- tion to the food value of this substance, and deserves further careful study. PeePROXIMATELY COMPLETE ANALYSES OF THIRTY “NORMAL” URINES. By OTTO FOLIN. [From the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Mass.] 7 nitrogen, — Kjeldahl’s method was applied as follows: 5 c.c. of urine are measured out by means of small burettes (with glass stopcocks) graduated in twentieths of a cubic centimetre, 20 c.c. sulphuric acid, and a small quantity of copper sulphate + potassium sulphate are used for the decomposition, the boiling being continued about one-half hour. The normal acid and alkali from which the tenth normal solutions are obtained, are prepared with sodium bicarbonate (made from sodium) as control. “ Alizarin red” is used as indicator. Numerous duplicate determinations have convinced me that this procedure gives perfectly reliable results with urine. It may be permissible to point out in this connection that the conversion of the urinary nitrogen into ammonia by means of sul- phuric acid involves very little oxidation. It is essentially a hydro- lytic reaction, and not an oxidation.! Urea.2 — The method used in this laboratory is as follows: 5 c.c. of urine are measured into an Erlenmeyer flask (capacity, 200 c.c.), § c.c. concentrated hydrochloric acid, 20 gm. crystallized magnesium chloride, a piece of paraffine the size of a small hazelnut, and finally two or three drops aqueous, I per cent solution of “alizarin red” are added. The safety tube described in my second paper on the subject is then inserted, and the mixture boiled until each returning drop from the safety tube produces a very perceptible bump. The heat is then reduced somewhat, and the heating is continued for a full hour. Particular attention must be called to the addition of the indicator 1 Matratti: Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 469; Foun: /é7d., 1904, xli, p. 239. 2 Fo.in: Zeitschrift fiir physiologische Chemie, 1901, xxxii, p. 504; 1902, XXXVi, p. 333; 1903, xxxvii, p. 548; MOrner, K. A. H.: Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 297. 45 46 Otto Folin. alizarin red, as its use in this combination has not been described before. This indicator is so stable as to withstand the heat, and it turns bright red in the presence of the smallest trace of alkali. The boiling contents in the Erlenmeyer flask must not be allowed to remain alkaline, and as soon as they begin to turn red, a few drops (but only a few) of the acid distillate in the safety tube are shaken back into the flask. At the end of an hour the contents of the flask are put into a litre flask with about 700 c.c. water and 20 c.c. 10 per cent sodic hydrate, and the ammonia distilled off. This distilla- tion must be continued until the contents of the litre flask are nearly dry, or till the distillate shows no trace of ammonia with delicate litmus paper (about one hour). The distillate is then boiled a few minutes to drive off the carbonic acid, cooled and titrated in the usual manner. Alizarin red is used as indicator. It is very service- able in all cases where moderate quantities of ammonium salts are present. This procedure is believed to give highly accurate and reliable results. The combination of this method with the Morner-Sjoquist method, as described by Morner, Joc. ci¢., must be used with urines containing sugar, because when carbohydrates and urea are heated together they combine, according to Schoorl! into very stable con- densation products, the ureids. Sugar was absent in the urines here investigated. I have, however, several times applied the combination of the two methods, but with practically identical results. The highest degree of accuracy is of course the most important point in connection with any analytical method, but as long as scien- tific investigations are still largely limited to more or less isolated individual efforts, it will be necessary at times to sacrifice the last degree of accuracy for the sake of saving time. The extraordinarily careful work of Mérner would seem to indicate that the combination method which he used was a trifle more accurate than my method taken by itself. I freely acknowledge that this may well be the case, and that Morner’s urea values represent the highest degree of accuracy at present obtainable. The only reason why I hesitate to accept this conclusion as quite beyond question is the fact that the accuracy of Morner’s figures obtained by the combination method depends upon the accuracy of his ammonia determinations. The latter were made according to Schlésing, but the important details upon which the ‘ ScHooRL, N.: Chemisches Centralblatt, 1903, Ixxiv, p. 1079. Analyses of Thirty “ Normal” Urines. 47 ~ accuracy of Schlosing’s method depends cannot be said to have been at all well-known prior to the publication of Shaffer’s paper! on the subject in 1903. The method is at best only moderately accurate, and if used according to the directions given in textbooks of physi- ological chemistry and urine analysis, the results are far from cor- rect. Leaving sugar urines out of consideration, it is therefore not quite excluded that the small differences in the results of the two procedures may at least in part be due to inaccuracies in the ammonia- determinations. The fact that duplicate Schlosing determinations were made, furnishes very little additional guarantee of accuracy even in the hands of such a skilled and careful investigator as Morner. Be this as it may, and accepting the differences which Morner found as being in favor of the combination method, they are in the case of normal urines so small (most frequently no difference at all) that I have not hesitated to continue the use of the urea-determinations as described above. The saving of time and attention is of no small importance when a consecutive series of urines are to be as com- pletely analyzed as has here been the case, and the combination method used by Morner requires much more time, as well as more continuous attention than the simple method here used. Ammonia.? — The ammonia is set free by the addition of a weak alkali (sodic carbonate), is then removed from the urine at ordinary room-temperature by means of a strong air current, collected in tenth normal acid, and titrated. 25 c.c. of urine is measured into an aerometer cylinder (30-45 cm. high), and about a gram of dry sodic carbonate and some crude petroleum (to prevent foaming) are added. The upper end of the cylinder is then closed by means of a doubly perforated rubber stopper, through which pass two glass tubes, only one of which is long enough to reach below the surface of the liquid. The shorter tube (about ro cm. in length) is connected with a ‘‘ calcium chloride tube” filled with cotton, which in turn is connected with a glass tube extending to the bottom of a wide mouth bottle (capacity about 500 c.c.) which contains tenth normal acid (20 c.c.), water (200 c.c.), and indicator for the absorption of the ammonia. The complete absorption of the ammonia is most easily insured by the use of a simple but effective absorption tube which compels a very intimate 1 SHAFFER: This journal, 1903, viii, p. 331. 2 Foutn: Zeitschrift fiir physiologische Chemie, 1902, xxxvii, p. 161; SHAFFER, this journal, 1903, viii, p. 330. 48 Otto Folin. contact of the air coming from the cylinder with the acid and water in the absorption bottle. For a description of this tube the reader is referred to the original paper (Zoc. ci¢., p. 169). It is easily made, but it can also be obtained from Eimer and Amend (New York). The air passing first through the alkaline urine, and then through the standard acid, transfers every trace of ammonia in the course of a relatively short time (14 hours) from the urine to the acid, and is then determined by direct titration. For the accuracy of this method I can vouch. The air current must, however, be rapid (several hundred litres per hour). An ordinary good filtering pump, or better an inverted filtering pump, a blast pump, is sufficient if the water-pressure is not too low. Kreatinin.!-— The principle upon which this determination is based is the color-reaction given by kreatinin, and by no other normal urinary constituent, with picric acid in an alkaline solution. In order to make the color-comparisons accurate enough for quanti- tative purposes a high grade colorimeter is necessary. The French instrument of Duboscq, obtained through Eimer and Amend, is eminently satisfactory. Its construction and workmanship are ex- cellent, the prisms are perfectly flawless, and the height of each solution can easily be adjusted to within an accuracy of a tenth of a millimetre. Half-normal potassium bichromate containing 24.55 gm. per litre, saturated picric acid solution containing about 12 gm. per litre, and 10 per cent sodic hydrate solution are the reagents needed. IO c.c. urine is measured into a 500 c.c. volumetric flask, 15 c.c. . picric acid and 5 c.c. sodic hydrate are then added, and the mixture is allowed to stand for five to six minutes. This interval is used to pour a little of the bichromate solution into each of the two cylinders of the colorimeter. The depth of the solution in one of the cylinders is then accurately adjusted to the 8 mm. mark. With the solution in the other cylinder a few preliminary colorimetric readings are made simply for the sake of insuring greater accuracy in the subsequent readings of the unknown solution. The two bichro- mate solutions must of course be equal in color, and in taking their readings no two should differ more than 0.1 mm. or 0.2 mm. from the true value (8 mm.), leaving out of consideration the very first reading made, which is sometimes less accurate. Four or more read- 1 Fouin: Zeitschrift fiir physiologische Chemie, 1904, xli, p. 223. =< : Analyses of Thirty “ Normal” Urines. 49 _. ings should be made in each case, and an average taken of all but the first. After a while one becomes very sure of the true point, and can take the average of the first two readings. At the end of five minutes the contents in the 500 c.c. flask are diluted up to the 500 c.c. mark. The bichromate solution is thor- oughly rinsed out of one of the cylinders by means of the unknown solution and several colorimetric readings are then made at once. The calculation of the results is very simple. If, for example, it is found that it takes 9.5 mm. of the unknown urine-picrate solution to equal the 8 mm. of the bichromate, than the 10 c.c. of urine con- ; out ae tains 10 X Se = 8.4+ mg. kreatinin.! The amount of urine taken for the determination is usually 10 c.c., but if this should be found to contain more than 15 mg. or less than 5 mg. kreatinin, the determination should be repeated with a corre- spondingly different amount of urine, because outside of these limits the determination is much less accurate. With kreatinin solutions, the results are uniformly surprisingly accurate, and I have as yet found no reason for believing that the method is not equally reliable, at least for normal urines. The color of the urine does not materially affect the result on account of the very great dilution. A kreatinin-determination can be made by this method in less than fifteen minutes. Uric acid.2—- The uric acid determinations are carried out as follows : The chief reagent is a solution of 500 gm. ammonium sulphate, 5 gm. uranium acetate, and 60 c.c. 10 per cent acetic acid in 650 c.c. water. 150c.c. urine is measured into a tall, narrow beaker, or a cylinder, and 374 c.c. of the reagent is added (if the available quantity of urine is not too small, 200 c.c. urine and 50 c.c. reagent are used). The mixture is allowed to stand without stirring for about half an hour. The uranium precipitate has then settled and the clear super- natant liquid is removed by siphoning or by decantation. 125 c.c. 1 The calculation is based on the experimentally determined fact that 10 mg. of perfectly pure kreatinin, give under the conditions of the determination 500 c.c. of a solution, 8.1 mm. of which has exactly the same colorimetric value as 8 mm, of half-normal bichromate solution. 2 Fotin and SHAFFER: Zeitschrift fiir physiologische Chemie, 1901, xxxii, P- 552. 50 Otto Folin. of this liquid is measured into another beaker, 5 c.c. strong ammonia | is added, and the mixture set aside until the following day. The precipitate is then filtered off, washed with 10 per cent ammonium sulphate solution until the filtrate is quite or nearly free from chlorides. The filter is removed from the funnel, opened, and the precipitate rinsed back into the beaker. Enough water to make about 100 c.c. is added, and finally the precipitate is dissolved by means of 15 c.c. concentrated sulphuric acid, and at once titrated with 2% potassium permanganate solution, each c.c. of which corresponds to 3.75 mg. uric acid. A correction of 3 mg., due to the solubility of the ammonium urate, is added to the result. The very first pink coloration, extending through the entire liquid from the addition of two drops permanganate solution, while stirring with a glass rod, marks the end point of the titration. Chlorides. — The chlorides are titrated according to Volhard, and the only point that need be mentioned is the fact that a fair amount of nitric acid must be added in order to get a clear filtrate. Phosphates. The phosphates are determined volumetrically by means of uranium acetate solution, using powdered potassium ferro- cyanide as indicator. Chemically pure crystallized monopotassium- phosphate is used to standardize the uranium solution.’ Total sulphur and sulphates. — A few general remarks concerning sulphate precipitations in urine may not be unnecessary. To.get correct results, it is indispensable that the barium chloride be added very slowly to the hot acidified urine. If the barium is added too rapidly the precipitate will invariably carry down other salts which the subsequent washings with hot water will not remove. To illustrate : Urine A, 0.50 c.c. + 25 c.c. BaCl,, added in about thirty seconds gave 0.2759 gm. and 0.2780 gm. barium sulphate. The same urine, with the barium chloride added by means of a capillary funnel (made from a calcium chloride tube) so constructed as to deliver 25 c.c. in five to seven minutes, gave 0.2713 gm. and 0.2693 gm. barium sulphate. Urine B. gave under similar conditions by the first procedure 0.4032 gm. and 0.4052 gm. barium sulphate, and by the latter proce- dure only 0.3897 gm. and 0.3900 gm. barium sulphate. Among the sulphur constituents of urine there are very minute 1 See SutTon’s Volumetric Analysis, 8th ed., p. 316. 3 WaNino Analyses of Thirty “ Normal” Urines. 51 _ quantities of substances which can scarcely be determined separately, but which are included in the total sulphur, and which must there- fore be counted as inorganic, ethereal, or as neutral sulphur. Such substances are traces of hydrogen sulphide very frequently present in urine, and the sulphocyanides which are probably always present. In the scheme of analysis heredollowed, these substances are included in the ethereal sulphates, not because they belong there, but it is most convenient not to exclude them in the ethereal sulphate deter- minations. This arrangement has ina measure proved unfortunate, because of the unexpected imc which the ethereal sulphates have acquired in the course of these investigations, an importance which was not at all foreseen at the time the analytical scheme was adopted. That the errors in the ethereal sulphate determinations due to this fact are not important enough to vitiate the conclusions drawn, is shown by the following determinations made with and with- out the use of potassium chlorate. The analyses were made in urines representing the eighth, ninth, and tenth days of a starch and cream diet described in the next paper. Urine A, 133 c.c. with the use of KCIO; gave 45.8 mg. BaSQ,. cere 68 without ae os OO ce ss B og with t: ef Aare 4 poe So ce without es ss AL a te” G BU with es CA Baked 4 « & without is es he mE es Inorganic sulphates.!— To 50 c.c. of urine in a (200 c.c.) Erlen- meyer flask are added 5 c.c. of a 4 per cent potassium chlorate solution and 5 c.c. concentrated hydrochloric acid. When this mixture has boiled a few minutes (5-10) it is perfectly colorless. The capillary dropping funnel (described above) is then inserted, and the barium chloride (25 c.c. 10 per cent solution) is poured into the funnel. After a few minutes, the heat is reduced to below the boiling point, and the mixture kept at this temperature for one- half to one hour. The barium sulphate is washed for half an hour with hot water, and at intervals of a few minutes hot ammonium chloride solution (5 per cent) is substituted for the water, so that in all five or six additions of ammonium chloride take place in the course of the first twenty minutes of the washing. Filter-paper and 1 FoLin: This journal, 1902, vii, p. 152. 52 Otto Folin. precipitate are then at once partially dried by folding and pressing gently between dry filter-papers, and are then transferred to the weighed porcelain crucible. 3 or 4 c.c. of alcohol are poured into the crucible and ignited. This dries and partially burns the filter- paper, and the residue is then burned to whiteness, cooled and weighed. The weight of the ethereal’sulphate precipitate, separately determined, is subtracted fromthe result. Ethereal sulphates. — 200 c.c. of the urine, previously diluted to I litre if necessary, is measured into a beaker, and 100 c.c. barium chloride solution (10 per cent) added in the cold. The mixture is set aside for twenty-four hours, and the clear supernatent liquid poured into a second dry beaker by decantation. This preliminary decantation is necessary, because the barium sulphate precipitate will otherwise go through the filter. The decanted liquid is filtered, and 150 c.c. of the clear filtrate, corresponding to 100 c.c. of urine, is measured into an Erlenmeyer flask (capacity 400 c.c.), IO or I5 c¢.c. concentrated hydrochloric acid, and 10-15 c.c. 4 per cent potassium chlorate are added, and the mixture is heated to boiling. The remainder of the operation is similar to that described for the inor- ganic sulphate determination. Total and neutral sulphur. — The total sulphur is determined in the urinary residue burned in the presence of 0.3-0.5 gm. potassium car- bonate and previously titrated for the “ mineral acidity.” It may seem questionable whether the urinary residue can be burned in the presence of such small quantities of alkali without loss of sulphur. The oxidation is, however, in this case a slow process, and is there- fore free from the dangers that prevail when the carbonate-nitrate fusion mixture is used. A series of comparative determinations made to test the accuracy of the method have shown the results to be very reliable, and incidentally have shown that the ignition with the fusion mixture does not give correct results unless large quanti- ties are used. This is shown by the following figures : IGNITION WITH K,COs. IGNITION WITH FUSION MIXTURE. 0.297 gm. = 0.1366 gm. BaSO, About 3 gm. = 0.1335 BaSO, 0/377 ¢m. ="'0.1365 gm. “ “ 2 om.= o.1s5cuuee S700 Oi. —= 0.137 or0m: mre “<6 em. ='0:137 25a 0-342 210, — 6.1366 pm. *. 6 2. = 003720 “ Io pM: = 0:137 5. 9 gm. = 0.1390 et. Analyses of Thirty “ Normal” Urines. 53 If the mineral-acidity determinations are not to be made, the nitrate-fusion-mixture method, or better still, the sodium peroxide method, is the more convenient; but where such acidity determina- tions are made, it would be useless loss of time to ignite a separate sample of urine for the sulphate determinations. (For details of evaporating and igniting the urine with small quantities of carbonate alone, see the mineral-acidity determination. ) From the total sulphate as obtained by this process are subtracted the inorganic and the ethereal sulphates, in order to get the “ neu- tral” sulphur of the urine. Indican. — The indican-determinations recorded in this paper will need a somewhat fuller explanation, because the method used for recording these values is wholly arbitrary. Exactly one-hundredth of a twenty-four-hour quantity of urine is taken for each determina- tion. In this the indigo is developed by the addition of Obermeyer’s reagent, and the indigo-blue taken out by means of 5 c.c. chloroform. With the chloroform solutions are then made colorimetric compari- sons, using Fehling’s solution as a standard. The Fehling’s solution is given the arbitrary value of 100. The advantage of this method is that it takes very little time, and the standard for comparison is one that is kept on hand in every laboratory. ; It has been my intention to work out an exact method on this principle, by using the same colorimeter as is used in the kreatinin- determinations for the color comparisons, and then finding the value of Fehling’s solution in terms of pure indigo. The difficulties in the way of such a colorimetric method are by no means small, the chief of which is to produce a pure blue solution from the indican without any admixture of red coloring matter. The method as described gives sufficiently definite values to enable any one to repeat the experiment, and get comparable values; but it does, of course, in the present form, not possess the accuracy that would entitle it to be classed as quantitative. Total Acidity..— To 25 c.c. of urine are added 15-20 gm. pow- dered potassium oxalate, and one or two drops of a I per cent phenolphtalein solution. The mixture is shaken rapidly for one or two minutes, and titrated at once, z. ¢., while still cold from the effects of the dissolved oxalate, until a faint but unmistakable tint remains permanent on further shaking. 1 Fouin: This journal, 1903, ix, p. 265. 54 Otto Folin. Mineral acidity or the excess of mineral acids or bases.! — Bunge,” Magnus-Levy,? Soetbeer,* and many others have studied the balance of the acid and basic forming elements in urine by means of sepa- rate determinations of all the different metals and acids. Here the same result is obtained by direct titration of the burned urinary residue. To 25 c.c. of urine in a platinum dish is added from 0.3 to 0.5 gm. K,CO;, weighed within an accuracy of two-tenths of a milligram. The solution is evaporated to dryness, and the residue ignited, when perfectly dry, over a radial burner using at first a very low heat, and at no time allowing the dish to become more than faintly red hot. The dish is heated at this temperature for one hour, then cooled, and 10 c.c. of hydrogen peroxide is added and evaporated. The dried residue is ignited as before for one hour. It is dissolved in an excess of tenth normal hydrochloric acid and water (50-75 c.c. 74 HCl), transferred to an Erlenmeyer flask, boiled to remove carbonic acid, and cooled. One or two drops phenolphtalein solution and a few crystals of neutral potassium oxalate (to precipitate the calcium) are added, and the solution is titrated as usual. The ammonia, the acidity of the hydrogen peroxide, and the acidity of the organic sulphur (neu- tral and ethereal, 8 gm. of which are taken to represent I c.c. tenth normal acid) must be subtracted from the result given by the direct titration. These values, as well as the acidimetric value of the potas- sium carbonate, must be separately determined. This procedure gives very reliable results, if proper care is used in the evaporation and the burning of the urine. It is to be used only when the actual excess of mineral acids above that necessary for the neutralization of the mineral bases is to be estimated, or when the total amount of organic acids in urine (whether free or combined with bases) is to be determined. For the determination of the /vee mineral and organic acidity, titrate the total acidity as described above, then determine the phos- phates, and subtract their acidimetric value (1 c.c. tenth normal acid for each 7.1 mg. P,O,) from the total acidity. The remainder is the acidity due to uncombined organic acids, and the difference, 1 Fo.in: This journal, 1903, ix, p. 270. 2? BUNGE: Physiologie des Menchen, 1901, p. 420. 8 MaGNus-LEvy: Archiv fiir experimentelle Pathologie und Pharmakologie, 1899, XXXxii, p. 149. 4 SOETBEER : Zeitschrift fiir physiologische Chemie, 1992, xxxv, p. 96. _ Analyses of Thirty “ Normal” Urines. 55 that obtained from calculating all the phosphoric acid as diacid phos- phate, is the free mineral acidity. If the acidity calculated from the total phosphates is greater than the titrated acidity, then there are practically no free organic acids present, and the titrated acidity represents the amount of phosphates present in the diacid form. Urines of the Jatter kind are turbid, unless they are also practically free from calcium. A more detailed discussion of the acidity of urine will be given in the next paper, entitled Laws governing the Chemical Composition of Urine. IJ. THE ComposITION oF ‘“ NORMAL” URINE CORRESPONDING TO A “STANDARD DIET.” Almost every textbook on physiology and on physiological chem- istry gives a sample table to illustrate the composition of normal human urine. Most English textbooks give the table formulated by Parkes a generation ago, and all textbooks that I have had the Opportunity to examine cite the same kind of tables now as were cited twenty years ago, 7. ¢., before the introduction of Kjeldahl’s method for determining nitrogen. All omit the total nitrogen which the tables are supposed to represent. Yet it was by the help of - Kjeldahl’s method that Pfliiger in 1886 made the important discovery that urea constitutes only about 87 per cent of the total nitrogen, and it is only on the basis of the total nitrogen that percentage studies of the different nitrogenous constituents of urine are possible. In view of these facts, and in view of the fact to which Bunge has called attention in the different editions of his textbook, that in all the voluminous urine literature there is still no record of the complete analysis of any one concrete sample of a normal twenty- four-hour quantity of urine, I present in Tables I-VII the records of thirty fairly complete analyses. In Schafer’s textbook of physiology (Vol. I, p. 573) Hopkins, in commenting on two analyses of Bunge’s, points out that no great importance can be attached to such collective quantitative analyses of urine unless the diet itself has been simultaneously analyzed. The thirty urines represented by Tables I-VII were obtained from six different normal persons, all kept for seven days on one standard uniform diet; and in order that all the urines should fairly represent the diet, the first two twenty-four-hour quantities were discarded and only the last five were analyzed. AS oe ore Otto Folin Ss 6¢ or Ot L'0 9°26 £s'0 aie nin * MLO ~ m0 Qe H+ = SONDNONG NO iS ‘ ¢ 199 c 10 £26 L'0 1°26 L0 0'¢6 8'p 70 £6 is L0 re 116 Taee BN pour -19}apuy) 18 LL8 788 OLS £98 Os'0 Os'0 9F'0 6b'0 140 ‘poe “vTIUO UU vl OLE) + vain PLUOWUY ‘vol | "NHDOULING TVLOT #0 LNAD ‘ws ‘uasoaqtu pour -19}9pu /) ANEPRE ot) ws ‘pro ou) “Vol yy wus ULUIQRILST ‘ad ‘H ['AN—I AIAVL *N-eIU -OWUL YY ‘SH NAF ‘9° BLuOulLU yy ee ONF UNIS >} 2 Vole yo om. I owe oR Om a os q “N-vatQ wi ‘PII ‘ONT ‘uas -O1}1U | (2201 | ‘“aULIN Jo owNnjoOA *——O'l "13 ‘ds *SO[Ly UI WYDIO AY 57 Analyses of Thirty “ Normal” Urines. *"panutuod —TI AIAVL 60 sojeyd -soyd [30,1 OSs PLS STs *pezesyy 1PI0.L ‘OURZIO | ‘TRIDUIW sT0 610 610 020 sT0 90°€ 61'E +0'E ‘WAHATNS ‘IVLOL <0 ene oT 5'D NI ALIGIOV, 10 INAO Wad NJ Panu) —T ALGVo 86 «[PANIN,, Za ve S ” “OS me «a S» "0S easoyyy | rue s10uy nai L0 189 Qc NWN HN —l SL0 abEs $30 0°909 ope NaN u ‘\das-3ny I¢ 6:oL SHO OQ PAWS YQWS || YO 4 ye aS S 3 Le 106 $8°0 Tee eee | es S | , ro | +0 | 980 CF = 6'06 $8'0 as St o-grg Lo€ \ : ap +4 S 9106 ooo | 9Ro | % as or 9S bl O's L'0 6'06 ; £8'0 hk e oocs giof or | Sa : Z'16 oso | OF ue | oak 2 i i +90 LET Ié LS s0 ; 0'06 Y Is'‘0 a Bo: o'99t Ge gol : wN ‘DIOR v *\T-proe *N-RlU z ‘ydag pr |, : viuowmue |. ; ans N-proe N-®! anda Bs ; pads peut | oun SLCHI ESET | ui Sal PIUOWWY ase uA) ayeary | -ouuty N-vo1Q N, //9Utn jo! -aqeq ‘uad |) suIN[OA, -19JOPU YL) pout -O.1}1U | =19}9PU/) “ud “ws HN St :9'9 “wus [BIOL ‘o'r | *SO[Iy Ut "NXOOWLING IVLOT 40 LNGXO WAd NJ | ‘prov ou) | ‘ulunevery | “eluoWUy voy) u3B‘dg | qWSIA We) ll Y ANI] ATAvVL SiS) nes. Ori ” 06'¢ 068 T68 erty “ Normal ses of Lh SSE 89°¢ FS OT + 86°€ C&P 666 617 68S Analy ‘(O0OT = ‘syreut ees “dy 4 . 1494) ueoIpuy ‘ws OULL AY UO) “08d Sv sajeyd -soyd [AOL “OTUR BIC) “[PIOUTTY ‘pazeiqy ‘TeIO.L, -OT .-. 7 oF oo) NI ATIGIOY, ‘WOHdATAS IVLOJ, JO LNAO ¥ panunuo)d —TIl ATAVL dd NJ LOS “OS « [WNIN ,, | [Pa1dyI OS o1ues10uy «S ”» 8OS sv Inydjns TROL asvIIAW lin. +68 10 YO -~ 2 2 2 >) SHOP HO KE 2 MO BANA YH QO nee oe oe Oe CO QA Co | Otto Fo $°06 O'f6 L°6 O'r6 L°C6 168 b'68 T'68 0°06 £68 9F'0 er'0 9F'0 1f'0 LEO -1oyopu /) 60 ‘prow ou) ‘NTOOULIN TVLOJT, AO LNT “UIUIYBIIYT “eLUOWULUe -+ voll “eLUOWILU YY Wad NI "ea1() ‘HW UNA ATAV.L ws ‘uas0.191U pour -19}9puy) *N-uLU “Voy -OWWY Qn Q — coo Ot Nr RA Stoan Ve oR UN 2 OQ oh ot "N-RoI | -o1qru ‘plo ou) tips} “UU BIAS] “HN $F '9°9 “PLCOWU wus | [ZIO.L e217) ‘Quin Jo auinyjo, | | "—O'! “13 ‘dg ‘Sny ‘27eq] *SO[LY ur WSO AY I 6 Urines. “ Normal” ty ar Analyses of Th tre ‘panuyuod —TA ATAVL 210 610 06% £9'¢ bL'e LLE SZ OLE ele 9G SL¢ [6% OIE IT€ OZT ‘syreuw “OY ‘(001 = uor}njos $,3ur “149a) uvoIpuy ‘OUI SOI) Ord se sozeyd -soyd TOL “D1URBIGC | *[RIOULTA, EO L10 ITO 020 £0 610 £10 £¢ 0 80°0 910 *payelqyy ‘TRIOL Ah OO NT AIGIOY: ‘MAHATAS IVLOJ dO LNAD Add NI ‘panu1yuod) —A ATAVL 1 STE IS’ [6% FL? oa “ 8S ” “ aS ” OS “OS « [RNIN ,, | [2919497 io Ss: [Os o1ued10uy OL'06 $816 ZIsl 961T OfFl lin. Otto Fo 0£°26. 00°€6 OL 16 OL'06 O8'06 09°66 ‘GNP pour | ‘plow | ‘uluT} -lojyapuyy | oI, | -varyy ‘BIUOWULULYy -+ vain "eLU -OWWLYy ‘NHDOULIN IVLOT, AO LNAO Aad NJ 62 +Z'0 62S 690 ebb s8'0 S99 *N-Blu ‘uaSo.ytu -ULUT] VAY | -OWWUL pour -19}9pu) c "ul ‘poe og “ULA val Sy “vIMOWU ‘TA-I] SUIAV, 10 SHOVATAY TIA ATAVAL 6£el SPOT ctel bhbl WNWIXe UN UULUryy ‘OSVIDAG [PUL "GAT ‘uaso.1jtu 1P30L aunjoA ‘sOpLy Sysram yentuy Sie Vine r O ” Normal “se ty Yr. ses of Th Analy 6'0L O95 ¥e9 *SsO[L] 610 elo LUO “xe “UN IM. ‘syieu | 9a (O01 = | | UOTJNTOS S, Sul ‘weed | . CSC O9E "us out -O[Y) sayeyd -soyd EO “OIUBBIC *paqeiqy THOU | Geo7 (OL 5:5 NI ALIGIOV ‘panunyuo) —TIAN ATAVL ‘WAHATAS IVLOJ, AO LNYO Wad NIT 60 910 £10 910 60 LUO es ” 20S «eQNdN ,, See Ile Ive oS» “OS [eo19 yyy “OS o1uvs10uy «S ” Og se inydyns TROL 64 Otto Folin. The diet which these urines represent was made up as follows: Wiholesmilk 25 < ...<))geuee Gao Sanne gGoncie: Cream (18-22 percentfat) . . . - - 300CCc. Eggs (white and yolk). . . . . . + 450gm. Horlick’s Malted Milk . =. . . . . 200 gm. SUPALIe ee ew ec sige eo ORO TLE Sodium chioride =, =a) : ae 6 gm. Water enough to make the whole up to two litres. Extra water to drink goo c.c. The ingredients combined into a liquid mixture contained 119 gm. protein and approximately 148 gm. fat and 225 gm. carbohydrates. The following determinations made at different times show that the mixture as prepared had a very constant composition : Cl, SO; P.O; Ne 6.25 3-69 5°75 18.92 6.02 SS a7t 5-54 18.80 3.62 5-92 18.86 3.81 5-92 18.86 3-80 19.05 3-84 19-05 19.02 19.02 19.02 The above diet fulfils sufficiently well the requirements of the so-called dietary standards, considering the class of persons for whom it was originally intended (patients in a private hospital for the insane). In regard to the protein, it corresponds almost exactly with the value demanded by Voit for a man weighing 70 kgm.; and since four out of the six normal persons experimented upon weighed less than 70 kgm. and the other two weighed no more than 70 kgm., the diet may be said to be liberal rather than too scanty with respect to its protein-content. . It is not my intention to devote any space to a discussion of these thirty normal urines, for the very reason that they correspond so closely to what we have been accustomed to consider normal. A number of points could profitably be touched upon on the basis of these tables, as, for example, the quantitative relation between ethereal sulphates, neutral sulphur, and indican; between ammonia, Analyses of Thirty “Normal” Urines. 65 total nitrogen, and the acidities; between chlorides, total nitrogen, and the volume of urine. These and several other more important points can, however, be considered to better advantage on the basis of the experimental material given in the next paper. In order to avoid unnecessary repetitions they may therefore be passed over here. The importance of Tables I-VII lies in the fact that they constitute concrete examples representing the prevailing views concerning the composition of normal human urine. LAWS. GOVERNING THE CHEMICAL, COMPOSIEiGia OF “URINE: BY O27 OSE OUIN: [From the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Mass.] CONTENTS. Page Introduction . . 5 ce neat Ee Ona OS Soo ¢ 66 II. The distribution of fie 1 urinary nitrogen among urea, ammonia, uric acid, and kreatinin . . Si it 5 Oo 6 67 III. The distribution of the sulpbar a urine among inorganic sulphates, etieren sulphates, and “ neutral ” aes we RE a gta be Gc 97 IV. Water and chlorides. .. . 2 oo ee OO) V. The relation of phosphates and organic acide to the acidity ve Wott 3 5 4 Le I. INTRODUCTION. HE term normal urine as ordinarily used, and as defined by the percentage tables in textbooks, seems to me to be largely a tribute to Voit. The dietary standards of Voit, demanding about 118 gm. protein per day in the food, will invariably give, in normal persons, urines of approximately the percentage-composition repre- sented in the tables of the preceding paper, the variations being confined chiefly to uric acid, chlorides, phosphates, and sulphates, these depending on the corresponding variations in the quality of the food. The meaning ordinarily given to the term normal urine implies, therefore, a tacit acceptance of Voit’s dietary standards as substan- tially correct. Numerous analyses of the urines from people repre- senting different conditions of life and different social strata have shown that very great variations occur, but throughout the voluminous literature on this subject there is to be noted a strong general desire or tendency to corroborate the findings of Voit. In most investiga- tions made of certain nationalities, as, for example, the Japanese, or of certain groups of people, as, for example, vegetarians and some rural populations, there are to be noted more or less strained attempts to show either that such people come nearer Voit’s standards than had been supposed, or that they are not properly nourished. 66 Laws Governing the Chemical Composition of Urine. 67 This paper is devoted to a study of human urines obtained from diets which are as different from the so-called standards of diet as they could be made, and the results are believed to furnish a more or less new point of view from which to examine those standards. Further, the analytical data and the generalizations recorded in this paper seem to require a reconsideration of the current theories con- cerning the nature of protein-metabolism. A discussion of the nature of protein-metabolism will be presented in a later paper, and in connection with it dietary standards will be discussed. IJ. THe DistTrRipuTION OF THE URINARY NITROGEN AMONG UREA, AmmoniA, Uric AcID, AND KREATININ.! Few persons probably would be willing to call the urines given in Table I normal, yet they were obtained from an apparently perfectly healthy man who gained rather than lost in weight during the experi- ment, and who, according to his own statement, consumed, while the experiment lasted, neither less nor appreciably different food from that which he had been consuming during the preceding three years. One exception must be made to the last statement. Dr. van Someren was not restricted in diet, the purpose being to discover what peculiarities, if any, could be found in his urines while on his ordinary mixed diet (consisting largely of vegetables, bread, crackers, cream, and candies). When the analyses of the first two days’ urines had shown that it was indeed a case of a very unusual urine, Dr. van Someren was however induced to try the diet described on page 64, but taking only one-half the quantity there given. This lasted only two days, January 30-31. Even half the quantity of the standard nitrogen-rich diet produced, it will be seen, quite a change in the composition of the urine, the change tending to make it much more like the urines shown in Tables I-VII. It may be added that Dr. van Someren lost in weight on the nitrogen-rich diet, and that he again began to gain as soon as he returned to his own mixed diet. ' For the opportunity to examine the urines given in Table I (the starting point of all subsequent experiments recorded in this paper), I am indebted, on the one hand, to Professor Bowpircu, and on the other hand, to Dr. ERNEST VAN SOMEREN of Venice. Professor BowpitcH kindly brought Dr. vAN SOMEREN on a visit to this laboratory, and the latter while here consented to remain the guest of the McLean Hospital long enough to permit the collection of a series of consecutive twenty- four-hour quantities of urine. Dr. VAN SOMEREN is known to many readers of this journal through his close | association with Mr. HorRACE FLETCHER, a popular writer on the value of the thorough mastication of all kinds of food. 68 Otto Folin. No attempt will be made to discuss in detail the figures for the different urinary constituents shown in the table, as this can be done to better advantage on the basis of the next four feeding experiments, planned especially to bring out in other persons. the same character- istics as are to be found in Table I. A few salient points concerning the nitrogenous constituents may, however, be indicated. From Table VII of the preceding paper it will be seen that the average total urinary nitrogen obtained from the protein-rich diet is 16 gm., with a minimum of 14.8 gm. and a maximum of 18.2 gm., and this is certainly what must be demanded from a normal urine of an average- sized man, zf the Voit or the Atwater dietary standards are to be the measure of what constitutes normal diets. As against this, we find in the above table only from 4.8 gm. to 8 gm. total nitrogen, and the latter figure only from a diet to which Dr. van Someren submitted under protest. Looking at the distribution of the nitrogen as shown by these two tables, we find in the former that on the protein-rich diet the urea represents from 86.3 per cent to 89.4 per cent, or again the values which have been considered normal ever since they were first discovered by Pfliiger. But in the latter table the urea-nitrogen has sunk to 62 per cent of the total, and at no time rises above 80.4 per cent. With such a decided fall in the per cent of nitrogen repre- sented by urea in Dr. van Someren’s urines, we must of course look for an increase in the percent of one or more or all of the other nitrog- enous constituents. A glance at Table I will show that this is indeed the case. To facilitate comparison, the figures taken from Table VII of the preceding paper and Table I are here placed together. TABLE VII. TABLE I. Total nitrogen 14.8-18.2 gm. 4.8- 8.0 gm. Urea-nitrogen 86.3-89.4 % 62.0-80.4 % Ammonia-nitrogen 3.3— 5.1" 4.2-11.7 “ Kreatinin-nitrogen Sy aG riay 5.5-11.1* Uric acid-nitrogen 0.5- 1.0 “ 1.2- 2.4“ Undetermined nitrogen = 2.7— 5.3 “ 4.8-14.6 “ The wide variations in the composition of Dr. van Someren’s urines are chiefly due to the temporary change in diet on January 30 and 31_ mentioned above. The urines recorded in Table I constitute, as far as I know, the first instance in which a urine obtained from a normal man was shown to contain but 62 per cent of its total nitrogen in the form of urea, and on the other hand over 11 per cent of the nitrogen in the form of Laws Governing the Chemical Composition of Urine. 69 kreatinin. The cause of this remarkable transposition of the nitrogen compounds was naturally sought for in the fact that the total amount of protein-metabolism was clearly reduced almost to a minimum. In the metabolism literature is to be found considerable evidence tending to show that the percentage composition of urine is not the same on a diet containing but little protein as on one rich in that constituent. The experiments of Sivén! and of Burian and Schur,? showing that the uric acid is toalarge extent independent of the total amount of nitrogen eliminated, point in this direction, though the in- fluence of this constituent is of course necessarily very small. The ex- periments of Gumlich® and of Camerer‘ and Pfaundler have shown that the urea-nitrogen is less in per cent of total nitrogen on a vege- tarian diet than on a meat diet, although the difference noted by these investigators was not very great (about 80 per cent urea-nitrogen on a vegetarian diet, as against 90 per cent on a meat diet). The full significance of these variations, or the possibility of obtaining much greater differences in the percentage composition of the urinary nitro- gen, was clearly not understood by these investigators, or they would have planned some more decisive experiments. Accordingly we find that their results have not noticeably influenced the generally prevail- ing views concerning the normal composition of human urine. This is very apparent, for example, in the writings of von Jaksch on the distribution of the nitrogen in urines obtained in different diseases. As a final conclusion derived from his numerous investigations, von Jaksch ® stated only last year (1903) that “urea is and will remain the chief nitrogenous product in the urine of the sick,” as of normal persons, z. ¢., representing from 83 to 91 per cent of the total nitrogen. Von Jaksch goes so far as to assert in a very positive manner that in all metabolism-experiments undertaken in the future for clinical pur- poses, direct urea-determinations may well be dispensed with, because the correct urea-contents of such urines can be found by simply mul- tiplying the total nitrogen of the urine by the factor 2! Such a rule, to be correct, would require that 93.3 per cent of the total nitrogen in the urine be there as urea! That Camerer, who for many years has 1 SivEn: Skandinavisches Archiv fiir Physiologie, 1901, xi, p. 308. 2 BURIAN and Scuur: Archiv fiir die gesammte Physiologie, 1901, Ixxxvii, Pp. 239. 8 GUMLICH: Zeitschrift fiir physiologische Chemie, 1892, xvii, p. 19. * CAMERER: Zeitschrift fiir Biologie, 1903, xlv, p. 1. § JAKSCH: Zeitschrift fiir klinische Medizin, 1903, l. Lin Otto Fo 7O L£S°0 tO 09°0 +90 610 tr0 65°0 gl 029 61 Bay pourut -lajapuy) “eIUOTUUULe + vaig “ULULJBIL YY ‘RIUOULWY “us ‘uaso01j LU pour “N HOOULIN IVLOL AO LNAO Wad NY -1a}Ja pus) *N-proe ou) *N-uLu -1]ed.1¥J *N-RIU -OWW Y "us “plow oI) *(uostag JVULMOAT) *S *A *S yl Glas teal ‘us “ULLAL Sy Wd Bs OT ey- SHIN 4 *9°9 “PLUOLULUL Ne ‘uas -O1}LU [BIOL “ahh ‘aun Jo auNn[O A 82 Lo 99S qaa-"uef ‘aed —o'r a3 ‘dg *sOpLY UL WSIO AN vik of Urine. ZOn O zz l Compos Laws Governing the Chemica ‘(O01 = uonnyos $,3ul 1424) ueoIpUy 0% ont eT eT el 6l 9¢1— “us “QUILL O14) se sayeyd -soyd 1®10.L “O1URBIO ‘TeIIULYAT *peze.yt} “[e10.L °91 9:9 NI ALIGIOV ‘MAHATAS IVLO] JO LNAO Yad NI ‘panunuod —{ ATAVL ér0 820 cc'0 eS ” BOYS « LPAVNIN |, ics ” Fas [eo10y3q | oues10uy 72 Otto Folin. been engaged in the study of this problem, has not appreciated the significance of some of his own experimental results is definitely shown in his last article on the subject published in 1903.1 Here, not on the basis of his own experiments, but on those of von Jaksch, pub-. lished the preceding year, he makes the “interesting discovery ” that the relative amount of urea-nitrogen “seems” to depend upon the absolute amount of total nitrogen eliminated. Unfortunately the urea-determinations of von Jaksch which led Camerer to the above- mentioned deduction were soon afterwards shown by von Jaksch? to be much too low and altogether misleading, because they had been made according to the directions given in his textbook, instead of according to the original directions of Schéndorff, and the former (textbook) directions contain a misprint which proved fatal to the results. According to the textbook directions, he had in some cases found as little as 66 per cent of the total nitrogen as urea, but could obtain no such results when he had eliminated the mistake in his technique. Since Camerer’s deductions concerning the relation be- tween the absolute amount of tdtal nitrogen and the per cent present as urea-nitrogen were drawn from admittedly erroneous experimental data, there is not much reason for believing that he still considers those deductions sound. As far as I am aware, he has published nothing on the subject since. The results obtained from Dr. van Someren’s urine showed, how- ever, positively either that his metabolism was very different from the normal, or that the absolute amount of nitrogen in urine determines to a heretofore-unsuspected extent the per cent of it eliminated as urea. But feeding experiments with low nitrogen diets on other nor- mal persons not previously accustomed to such diets have given in every case results similar to those shown by Dr. van Someren’s urine. It may, therefore, be positively stated as a principle in the chemistry of metabolism that ¢he distribution of the nitrogen im urine among urea and the other nitrogenous constituents depends on the abso- lute amount of total nitrogen present. After having established by means of preliminary experiments with mixed low-nitrogen diets the correctness of this principle, further studies with uniform diets were undertaken. It did not seem improb- able that such studies would yield considerable definite information concerning the conditions that determine the formation and elimina- tion of each separate nitrogenous constituent. The investigations 1 CAMERER: Loc. czt., pp. 14-16. 2 JAKSCH: Loc. cit. 7 Laws Governing the Chemical Composition of Urine. 73 seemed, moreover, equally promising in connection with the sulphur of urine, for it was found that with respect to this element also there is in Table I an entirely different distribution among the three chief representatives (inorganic sulphates, ethereal sulphates, and neutral _ sulphur) from that previously observed in feeding experiments with _ the standard nitrogen-rich diet. (Compare Tables VII and VIII.) ‘ _» Since these observations have also been verified by the later experi- |. ments, these facts are also held to illustrate a principle in the chemis- 4 try of metabolism. The distribution of the sulphur in urine among the _three chief normal representatives — inorganic sulphates, ethereal sul- phates, aid ‘neutral sulphur” — depends on the absolute amount of total sulphur present. For a more detailed study of the laws or conditions governing the formation and elimination of the several more important urinary con- stituents the reader’s attention is now called to Tables II-V. The four feeding experiments there recorded, all practically alike, are . divided into three periods separated by double lines in the tables. During the first period, lasting three or four days, each person took the standard protein-rich food described in the preceding paper and which forms the basis of the results there recorded. At the end of this period it will be seen that each person eliminated exactly the same kind of urine as has already been abundantly illustrated in the first seven tables. A complete change of diet was then made. The effects of this second diet, containing only about 1 gm. of nitrogen as against 19 gm. in the first diet, are represented in the second period of the tables (the space between the two pairs of double lines). The second period lasted from seven to ten days. At the end of this time a return was made to the first protein-rich diet, but only for one or two days. The diet used in the second period of these experiments is referred to in the tables as a ‘starch and cream diet.” It is by no means an easy matter to provide a diet sufficiently nourishing for the ordinary normal man without including enough nitrogenous material to estab- lish nitrogen-equilibrium, if the food is to be sufficiently palatable to enable a person to take it for more than two or three days. Yet it is only by means of such a diet that one can hope to bring the total amount of nitrogen in the urine down to the lowest possible level. The diet here selected is believed to be well adapted for the purpose, but the quantities used were somewhat too small to yield the most satisfactory results. The diet consisted of 400 gm. of pure starch and lin. Otto Fo 74 LC SL fel O9L VIL c6L 06L DLL $08 6+8 168 tb Lb ‘UIUT VOLS L'06 oe £06 wits 9°68 Le “eLUOWUUe 4 vain | Cuowmy ‘NHOOULIN IVLO]L JO LNAD Yad NJ €19 S19 €"s9 0°09 069 069 S19 Sel O6L ++8 S18 £98 6'S8 ‘eo1Q ‘ws ‘udso1}1U peur -lojapuy -NGPHe oy) “ws “poe otf) *(uosaog JvUuL40N) “TO Tl HIGVL *N-uru 1801 y “ud ‘ULUeaL yy *N-eIu IS FIN 9% 9°9 “eluUOUl uy -owuy | N wy : 2 O EES OM FEN HN YA YA wy RIGY Rip! MICO NO Teta Saar ONC) SS BAD MO SHS Ne Ne 2 -eoIQ) wa "voi 9 OI T9I +1 CCL Byr ‘uas -O.1]1U [e191 ‘oD )“oulIn jo aunjoA ‘—o'l 13 ‘ds *SO|Ey at 14 < BIOA\ in a "20e1} OL LT 1Z¢ Gres 98Z $6 OO! $08 STO 610 : 06'T . 0 0% TI S&z oI Zee. |i Ile isaac bons 2:05 £z'0 60°0 +L0 Ss 0 aA L0 +61 8 70% T+é Sor 909 roa 0) 0) 0) i cL0 S 0 Mal OT 16 S 96 8°22 971 9'b9 LTO Ilo 610 e 0 £2 aL 692 S +92 $97 ra T19 LV‘0 » 60°0 b cL0 S 0 SE Cul Zhe LI 6SZ Glog 90'0 sL'0 B 0 Se lan imate alles we || £2 | 06 10°0 610 S 0 9% TD OST $s Sez [Ash 96 800 £8'0 iss) IS 0 8Zé v1 692 9% S67 +81 OTT 10) 0) i 98'0 8 P) 28) 9% ST LIZ $ST TLE ann T8 80'0 ( 660 N Os Ze FZ 68ST SO€ L6b LEl £8 z10 : StI 8 ee eee \S) Os 09 I+ Lob OLE LOS HE HS STO ) fee % ~*~ OOT 6s (Ae 6b% SO0¢ LSS T+ gs 610 : 9V'€ So 09 $s 67 S9r Ga fbb +S T9 LTO SL'Z 'S S ‘(OOL = I OFq |/OlURSIC | ‘[eIOULTA, ee Bo Bg tg Q rmeerguer | UOLINIOS "ws se Cte) A S $,3ul ‘ours =| Sayeyd £05 yaq) | -o1yo | -soyd Ace aie ‘MNHdATAS ' ‘ . ueoipuy [R10], OL SO NI ALIGIOY mes ROoeeo raat « [etnan ,, | [eo194}q | o1UeS10UT 8 in} NX ‘panunuod —Tl ATAV.L Otto Folin. el a 8s°0 +90 SFI o-ost £9'0 95°0. SQL o' Sor Qt hh ™~ Rey 09'0 crO Og'r o:66z2 190 Is'0. IGT osg€ ~8S'0 Lv0 SSI o'9€E cS'0 or'0 EFT o-zo€é cs'0 + a0) OFT ogr€ ss0 cr'0 OF-r o'gbe 0s'0 Ir0 GET oebe ROO OMI LO POS SSWSA KISH HA | 850 6+'0 EFT ors& 850 £+'0 LST o:S0€ 09°0 £0 EQu oete CO | We Ind w et tor od sooo 00 “AN ‘plore ‘aye “eLUOULUTR + vol) ‘ws CNSRIPE |" Neue Nee ‘CNT ||-autanyo ‘uaBoinu) Cf | Heery | -owuy ue8 ‘ll etnnion POU | Se ae | | OseTN -1ajapuyQ a — RS tie [e10] Spee ; NTH r ‘ ray ¢ 4 x HN 29 L —O'! soyly ar NHISOULIN IVLO], 10 LNID Ud NJ ‘prov aug |-uuneary | “ewomuy “a3 dg YBa A pour “ULUN BIAS -lajapuyy) on ‘eluomury | "vora ‘(uostag pouwony) "FY “Gq ‘H Ad TIT WIV. 77 c2'0 +10 95°0 260 620 ITO L9‘0 LOT PLT 610 sto ra ia LVT a8 2 zo | 270 zv'e San Lt0 | ero 6S2 X N oo | oro | 90 | 9L0 > |g 5 10 | oro | zo | €80 s | 3 . . asl Ge so | 60 | OFO | 240 ats 3 5 0 | oro | oro | to 5 5 a. S 86£ "+ : STO 610 LOE WE +02 } +E0 970 08% Ore OF ; E34 Séo ; 02 0 LT0 thZ 18? ‘JaIp Bde pue ATTN : = : “OIUVSIQ | ‘;e1dUT be ay : : : ie “us [ Oa! HIN [PIO .L aes ” eS ” aS ” se S» s Sur *OuLT ¢ 8 g OS se 1 i 405 Os “OS ; WMeA) | -Olr49 Bien “AHA TAY inyd[ns uvoipuy of 319 NI ALIGION. AVLO, 10 Bede i nq || [B4qNaN ,, | [V219YIG | rURB.10UT [PIOL “SyIeU a “panuyuod — JI AIAVL Laws Governing the Chemical Compos Otto Folin. 6b | ZT Lt 268 GE 3) aR mig pion] ool eel melee ara Woe 0 ri a aces ema sO a eae OL 98 | +I 19 68 BOE RANE TO Tea ee eee te ees ae Leah | BDOens| 6 91t | of | O+FT LOL GH | eC NCO eee oe oleaurip aloes tll eaememl saree ||| es Maye Sor | 6¢€ O21 o'¢L oF O69 | aGeiOr | aero wie, ABU OL | Ce as eyeeea mone L Te feo iG 90L VOR eS NP GO. yl eee ere Alea alcealline le 9 spe aver) one aes Oe Wace econ beer) ee POE ee lea eae ue aac 5G eto €18 Os | ahiekeals ev ee | ROR OO ee eee ea ae e£ | +1 | o8 e€8 ee (SOM Soa eee ee Nee casa € BOT || 10) s0;s O+8 E95 | VEU lV" Yepioy nl taco phe VEO a) 2 oranllicaalae tag Se | 740 82 016 se PI SCION eerie eter a Ue Ie ait Se || LO Le 016 ee HAR) i880) PEO | eve ea eat) OBE || Seare| waoo cen noe 99 | OT hig 968 o¢ OS | 2960) eee eee ay zccomece oth |liepes an lmnez LOU vA o¢ £88 Og. A MOSB HO. ho PO Ot et | ween ciel eee eae ae PIP | -urunvary | POU | uowury | “vag “us ENGR NOE || NGS | Se NEES weeorm@| CON acum jo AieeaL -9,2pug) ou tas 4 + vo1- : ‘uasoytu| FN -Wwory | -owwy 93 SUINIOA 21eq peuta -O1jLU ‘NADSOULIN IVLO], AO INHO Ud NJ fade Ser Sah waters ee a as ne aie *(uosdag jpULMONT) “Y *S “a UC PANIC ELAS Al, ao Laws Governing the Chemical Composition of Urine. r.¥) i 7 2 z s 02 £9 9% rau a 0) 9'8 ez ) Sb 80 Nag S8s| 0 Ls 90 ee ° 2 0 ad 60 9081} o% GAL 4 20811 Sill oT ae ol EE Sioa at og oe SZ 2| 08 +L T'+ a | Soe | 98 eb a5 ie | S081 L9 I+ gg e OST Us ee ‘(00T = 3 "0%d = uortynjos “UL se ob tees $,5ul ‘oult =| sozeyd 149M) | -O1yD | -soyd uvolpuy [PIOL €8¢ Iz +6S es v9 9°88 S9S 0¢e— Ste VL VL oS8 86+ 16¢— LO? eG 601 959 IL ihe LOT STI Z8 0°08 LES 02— LIé 9'€% L El L709 9ST L6 £S@ £91 OOT Leh 8EZ s9 £O€ +'ST Sol SOL 827 L6 See Sel FIT T9L Loe O12 LES Srl LOVE 174 60+ 862 LOL LS 09 £88 O9T oes 069 6+ VL LL8 SI 10+ 6IL L’9 98 L+8 Lve Oe £6S a9 TL L98 , 0 Perea 8 % I MULTI | "[eIIUITAL qeI0], S S S ‘AHATAS *OL +5°D NI ALIGIOV ‘panusuoed — KT ATAaVL IVLO], JO LNAD Ud NJ £10 910 See zT0 aay) 6E°T sT0 L0°0 cr'0 O10 L00 890 zT0 LOO cf'0 O10 90°0 St’0 c10 L0'0 9+'0 ITO oro L9'0 02'0 st0 00'T 9T0 610 99°% +0 e720 8s 020 9¢°0 9S°% 910 slo 12% cliGine apie wor al wectreip “OS OS den oo _{euynan,, | 194g | oues10UT Ol CU 0) se) Ene rs! 00 ‘9yeq, Otto Folin. So 0's8 O'+S T +1 Let oeT £6 0's +9 ob TL9 STL €L9 S'6L L18 £8 6LS L18 C6L Ve £6S VL9 6LS eZ O9L T8Z S'+8 Te Re Sie se “UU RAL 8°06 0'06 2:06 £68 ‘eIUOTWUIe + Ra1Q oe Spee Te oe “eIUOWWLY LS $98 TL8 T98 eal) ‘NHOOULIN IVLO], AO LNAD Ud NJ 98°0 aw +60 ANY gor |) peor 6b0 | Bfo “3 | “N-PIR ‘ussoytu; IFN NO™m NONGDNOGONO Coin N Yo *N-ULU -1]eoly SOI LST Stl bl *N-P LU -OWLWLY peurut -l9japuy) xn} ‘prov ou) tbs} “ULU RIL Yy ‘(uostag jou4ony) “FI ‘OAY Aq ‘A HTE&VL “HNO 90 *RIUOWUWY “GAT ‘uos -O.1}1U [BIOL bo} “auTIN Jo ano A PmmO° “ad ‘ds N Nm +f wm OO KY 6 SZ rol *Atnf—9un ‘a7eq “SONY Ul SIAL 81 S a Ss Ss i) Ww Qo 0 os Ss aaa NS 0 >) 8 9081} Ss *90R1} “9ORI} ‘JoIp WesID pue YoIV}S Os 9EF e9€ 66L +9E +9F 88 Il ese FOL 08 : : $67 Ie 969 ‘(OO = ‘OTURBIC | [Peru eer uoynyos a $,5UL phi Inydyns [yo q) ot MOHdATAS vl re Oh i : payne Ralayyy | ouesiou uvolpul eo IVLOL AO LNAO Yad NI al AL iy Be 1) io Laws Governing the Chemical Compos “ponuruod — NK ATAVL 82 Otto Folin. 300 c.c. of cream, containing 15 to 20 per cent fat. Pure arrow-root was selected for the purpose, not only because it is free from nitrogen, but because it is more palatable than ordinary potato or corn starch. The starch mixed with 1500 c.c. of water, and heated until a uniform paste was obtained, was then cooled to 70° C., and digested at this temperature for about half an hour with 2 gm. of active diastase. The ferment reaction was stopped at the proper point by heating for a few minutes to 90° C. The operation is best carried out by means of a so-called double boiler. In this way one can easily obtain a mixture exceedingly rich in starch which is not sweet, and which is liquid when hot, and semi-solid when cold. The consistency can be regulated at will by means of longer or shorter action of the ferment, but the ferment action must not be carried too far, or.the product will acquire a rather disagreeable sweetish taste. Six grams of sodium chloride was the only seasoning used, and no tea, coffee, alco- hol, or stimulants of any kind were allowed during the experiments. This diet has now been tried on several different normal persons, as well as on patients, and I have as yet found only one who was not able to take it, and he succumbed the first day. No person taking the diet has felt less able than usual to do his daily work, nor has any one reported any particular craving for nitrogenous food, though the diet of course gets to seem less and less palatable, especially when served with such unnecessarily small quantities of cream as were here used. Instead of 300 c.c. of cream, it would have been better to use 600 c.c., as the total nitrogen in the food would still have been below that needed for nitrogen-equilibrium, and the increased caloric value of the food would then in all probability have reduced the urinary nitrogen still further. In the general outline nothing new or original is claimed for the feeding experiments with this diet. The conditions are essentially similar to those prevailing in most attempts that have been made to discover the lowest possible level of nitrogen-equilibrium, as, for ex- ample, in those of Sivén,! and they resemble still more the similar experiments of Landergren,? both of whom have in fact succeeded in reducing the nitrogen-elimination to a lower level than was here attained. But the main purpose here aimed at was exactly the reverse of theirs. Their chief aim was to reduce the protein-metabolism to the lowest possible level, and incidentally they studied somewhat the 1 SIVEN: Loc. cit. 2 LANDERGREN : Skandinavisches Archiv fir Physiologie, 1903, xiv, p. 112. Laws Governing the Chemical Composition of Urine. 83 composition of the urine. In these experiments the attempt was made to reduce the protein-metabolism to a minimum, but only in order to make more instructive the data obtained from a detailed study of the composition of urine. Turning now to Tables II to V, and taking first a general glance on the one hand at the total nitrogen and the total sulphur columns, and on the other at the columns representing the percentage-distribution of the different nitrogenous constituents and sulphur compounds, it will be seen that the two principles stated above with reference to this distribution hold true. With the change from the nitrogen-rich to the nitrogen-poor diet there is a great reduction in the total amounts of nitrogen and sulphur in the urine, and accompanying this diminu- tion we find a most striking transposition in their distribution among the different constituents. The same fact is further illustrated in Tables VI-X. With the reduction of the total nitrogen and total sulphur in the urine, we find a smaller per cent represented as urea and as inorganic sulphates, and a relative increase in the other repre- sentatives of these two elements. Some more detailed questions must, however, now be raised, and the answers sought for inthe tables. Is the changed percentage- composition accompanying a reduced protein-metabolism due only to ’ a reduction in the absolute amounts of urea and inorganic sulphates formed and eliminated? Do the other constituents also diminish, but at a different rate, or do they remain fixed in absolute amounts, or can they even be absolutely greater when the protein-metabolism is teduced? The reader has undoubtedly already asked these questions, and observed that the answers must be different for the different constituents. In a discussion of the relative and absolute variations in the elim- ination of the nitrogenous constituents produced by quantitative changes in the protein-metabolism, it would perhaps seem most suit- able to begin with urea, as is the general custom in all metabolism experiments in which it has been determined. In this case, how- ever, the discussion of the urea will require considerable reference to the other nitrogenous constituents, and it will therefore be taken up only after they have been considered. Kreatinin.— The part played by the kreatinin as a factor in the relative distribution of the urinary nitrogen is perhaps the most interesting fact discovered in the course of these investigations. It has already been pointed out (p. 68), in connection with Tables VII 84 Otto Folin. and I, that when the urea constituted 86-89 per cent of the total nitrogen in Table VII, the kreatinin represented only 3.5-4.5 per cent; while in Table I, when the urea had sunk to 62 per cent of the total nitrogen, the kreatinin-nitrogen had risen to 11 per cent. The importance of the kreatinin as a representative of the protein-metab-_ olism, when the latter has been reduced to a minimum, is even more strikingly illustrated in Tables II-V. In Table II the kreatinin- nitrogen represents 3.4 per cent of the total on the last day of the protein-rich diet, and 15.4 per cent on the last day of the starch and cream diet. The corresponding days in Table III give the values 3.6 and 17.4 per cent respectively. And almost identical changes in the relative importance of the kreatinin-nitrogen are shown by the other tables. Turning from the consideration of the per cent in terms of the total nitrogen to that of the absolute quantities of kreatinin elimi- nated, we find the remarkable fact that the absolute quantity of kreatinin eliminated in the urine on a meat-free diet is a constant quantity differ- ent for different individuals, but wholly independent of quantitative changes in the total amount of nitrogen eliminated. This appears to me to be another fixed principle in the chemistry of metabolism. In verification of it, the reader is invited to inspect all the tables in this, as well as in the preceding paper, representing sixteen different feed- ing experiments, and 163 twenty-four-hour quantities of urine. The statement will be found substantially correct ; certain small variations do occur, but the constancy of the kreatinin, as compared with all other nitrogenous constituents, is certainly remarkable, especially in view of the fact that the muscles and most if not all other organs must at all times contain very considerable quantities of kreatinin. The variations that do occur are as independent of the volume of the urine as of the total nitrogen. To cite but one out of very many illustrations of this to be found in the tables, the reader is referred to Table IV. There we find that Dr. A. eliminated, on June 30, 2300 c.c. of urine containing 15.8 gm. total nitrogen, and on July 6 he eliminated only 505 c.c. of urine and 2.7 gm. total nitrogen. The amount of kreatinin is exactly the same on those two days. If the kreatinin findings here reported are correct, and it would seem as though the experimental material furnished constitutes suf- ficient evidence as to this, we have in the kreatinin by far the most reliable index as to the amount of a certain kind of protein-metab- olism occurring daily in any given individual. Laws Governing the Chemical Composition of Urine. 85 While the amount of kreatinin eliminated with the urine is for each individual practically a constant quantity, independent of the total amount of nitrogen and of the volume of the urine eliminated, the amount may be very different for different persons. Thus we find the average kreatinin-elimination of Dr. H., weighing 87 kgm., to be 1.6 gm. (Table III); while in the next table (IV) we find that Dr. A., weighing about 56 kgm., eliminates on the average about 1.15 gm. kreatinin. The chief factor determining the amount of kreatinin eliminated appears to be the weight of the person. The proportion between the body-weight and the amount of kreatinin in the urine is, however, not very constant. Fat or corpulent persons yield less kreatinin per unit of body-weight than lean ones. This is illustrated in Tables II and V. There we find that F., who weighs 65 kgm., but is lean, eliminates about 1.6 gm. kreatinin (Table II); while Dr. H. (Table V), who weighs 70 kgm., but is rather short and corpulent, yields 1.4gm.kreatinin. Theanalytical data, as far as yet obtained, indicate that moderately corpulent persons eliminate per twenty-four hours about 20 mg. kreatinin per kilo of body-weight, while lean persons yield about 25 mg. per kilo. The above facts show that in metabolism studies of the sick it is not permissible to speak of abnormally high or low kreatinin in the urine, without taking into account, on the one hand the body-weight, and on the other hand the physical condition, z. ¢., the adipose tissue present! In view of the seemingly extremely important significance of the kreatinin in the metabolism of normal persons, it would appear highly desirable that the elimination of this substance in different diseases be more fully investigated. Also it should be worth while to investigate the effect of certain drugs on the kreatinin-elimination, particularly those which are known to affect the uric acid, as, for example, the salycilates. The effect of hard physical exercise on the kreatinin-elimination has already been the subject of considerable investigation. The results may still be considered doubtful, and further experiments are necessary to settle this question. Gregor,? the most recent investigator in this field, came to the conclusion that hard physical exercise does produce a decided increase in the kreatinin-elimination, particularly on the day following such exercise. The extraordinary variations in the normal daily kreatinin-elimina- 1 See MACLEOD, J. J. R.: Journal of physiology, Proceedings of the Physi- ological Society, xxvi, p. vii, 1900. 2 GREGOR: Zeitschrift fiir physiologische Chemie, 1900, xxxi, p. 98. 86 Otto Folin. tion shown by Gregor’s figures render them, however, decidedly doubtful, although I am inclined to believe that his main contention may be correct. The fact that the kreatinin-elimination is normally an almost perfectly constant quantity in any given person is important in connection with metabolism experiments in hospitals, or whenever one must depend on someone else for the proper collection of twenty-four-hour quantities of urine. Any considerable loss of urine is promptly shown by the kreatinin-determina- tion, which can be made in a few minutes. As an illustration of this, attention may be called to the omission in Table VIII of the urine corre- sponding to April 18. 600 c.c. of urine, containing 5.1 gm. nitrogen, 136 c.c. {4 ammonia, and only 0.81 gm. kreatinin, were brought to the laboratory as the full twenty-four-hour quantity. The preceding six days showed, however, that the patient should certainly not yield less than 1 gm, kreatinin per day. An investigation of the ward was at once in- stituted, and it was found that the patient had been without supervision during a considerable part of the day. Uric acid. —The preceding discussion of kreatinin has doubtless reminded the reader of the important papers of Sivén,! and of Burian and Schur? on uric acid. Burian and Schur devote some two hundred and fifty pages to a very able and interesting discussion of the part played by the “ purin bodies” in metabolism. The most important point brought out by them, or at any rate the point most pertinent in connection with these investigations, is their contention that the uric acid eliminated by man on a purin-free diet is for each individual a constant quantity and entirely independent of the total amount of nitrogen eliminated. This contention becomes peculiarly strong in view of the fact that Sivén simultaneously and independ- ently had drawn the same conclusion from a series of experiments which seem particularly well adapted to prove the point. Since the two diets described above (pp. 64 and 82) conform to the conditions demanded by, Burian and Schur with regard to their purin contents, the numerous uric acid determinations recorded in these tables should throw additional light on the validity of the conclusion reached by Sivén and by Burian and Schur. By inspection of Tables IJ-V, it will, however, be seen that the uric acid elimination is by no means as great on the starch and cream diet as on the milk and egg 1 SitvEN: Skandinavisches Archiv ftir Physiologie, 1901, xi, p. 123. 2 BurtIAN and Scuur: Archiv fiir die: gesammte Physiologie, 1900, Ixxx, p. 241; 1901, Ixxxvil, p. 239; 1903, XCiv, p. 273. Laws Governing the Chemical Composition of Urine. 87 diet. In Tables Il and III it sinks on the former diet to only about one-half of what it ison the latter; in Table IV it sinks to about two- thirds, and in Table V to about four-fifths. That is, under the con- ditions of these experiments, the uric acid elimination does not show the characteristic quantitative constancy that we find in the case of the kreatinin. When the total amount of protein-metabolism is greatly reduced, the absolute quantity of uric acid ts diminished, but not nearly in proportion to the diminution tn the total nitrogen, and the per cent of the uric actd nitrogen in terms of the total nitrogen 1s therefore much increased. The last brief statement represents, however, by no means all that there is to be said concerning the conditions governing the uric acid elimination. This has always been a complex problem, and as far as Ijcan see, is so still. In the first place it might be said that the results obtained by Burian and Schur do not necessarily conflict with those recorded in Tables II-V. These authors did not work with urines showing such wide variations in their total nitrogen-contents as is here the case, and they accordingly generally refrain from claiming that the “ en- dogenous ” uric acid is absolutely independent of the total amount of protein-metabolism. It might therefore well be the case that the conditions prevailing in Experiments II-V are simply outside the limits within which the rule of Burian and Schur is true. But the feeding experiment of Sivén represents even wider variations in the urinary nitrogen than are shown in these tables, yet his results fully confirm the findings of Burian and Schur. In order to get light as to the cause of these different results, it seems necessary to sift out the experimental material that really has a bearing on the subject. Notwithstanding the abundance of highly interesting experimental material recorded in the papers of Burian and Schur, it seems to me that their position with reference to the “ endogenous ”’ uric acid is not particularly well fortified. One single feeding experiment, lasting twelve days, and including nine uric acid determinations, constitutes all the experimental data of their own which they have recorded as a basis for their contention that the “endogenous” uric acid is for each individual a constant quantity.! _Their experiment was divided into three four-day periods,.and represents a diminution of 6 gm. in the total urinary nitrogen, z.¢., from 14 gm. to 8 gm. Sivén also offers but one series of experiments in proof of the 1 BURIAN and Scuur: Loc, ctt.; Ixxx, p. 290; XCiv, p. 274. 88 Otto Folin. contention that the uric acid in the urine of man is independent of the total amount of nitrogen eliminated. His is, however, an exceed- ingly effective experiment, and fully justifies the conclusion which he drew from it. The experiment lasted thirty-nine days, and the urinary nitrogen was reduced from 21.1 gm. to 2.8 gm., yet the uric acid elimination remained practically constant. Since the experiments recorded in Tables II to V show an unmis- takable reduction of the uric acid when the total amount of protein- metabolism has been reduced to the extent there accomplished, we must conclude that the endogenous uric acid elimination can be reduced, at least under certain conditions. Burian and Schur must not be represented as denying this. They take due notice, for ex- ample, of the fact shown by Ranke and by Hofmann! that the amount of uric acid eliminated in hunger is actually less than that elimi- nated on a food containing no protein at all, and notwithstanding the fact that in the former case more total nitrogen is eliminated. But Burian and Schur consider all such results abnormal, and the result of a greatly changed metabolism. I have no doubt that they will pass the same judgment on the results recorded here. This seems, however, to me to be decidedly an open question, especially in view of the fact that the uric acid is less and not greater in hunger than on a food containing no protein. ‘The fact that the kreatinin zs a constant quantity in these experiments seems to me to be of no little importance in this connection. It is not easy to understand how ‘“‘Einschmelzung xanthinbasenhaltigen Materials (Muskel u. s. w.) ”? could take place without noticeably affecting the kreatinin-elimina- tion one way or another. If the endogenous uric acid is to be considered as derived from the cell nucleins exclusively, it would indeed seem highly plausible that the quantity should tend to remain constant, even with very great variations in diet. Rigid proof that the endogenous uric acid elimi- nation is for each individual a constant quantity would be strong evi- dence in favor of such a theory. Burian and Schur support the view that the endogenous uric acid is derived from the cell nucleins, but they contend that in man about one-half of the uric acid so derived is destroyed inside the organism, and that only the other half is eliminated. With the introduction of this important modification of the nuclein theory, there is no longer any reason why the uric acid 1 BURIAN and ScuHuR: Loe. cit.; \xxxiv, p. 275. 2 BuRIAN and ScHur: Loc. cit.; \xxx, p. 275. Laws Governing the Chenical Composition of Urine. 89 elimination should not be a decidedly variable factor which might well be susceptible to change under the influence of many different changes in the conditions, among others, changes in diet. In any event it seems clear that the two experiments of Sivén and of Burian and Schur are not sufficient to prove the point under discussion, since these experiments have given different results. Looking only at the analytical data of my experiments I am quite as much impressed by the variations in the uric acid elimination as by the absence of variations. The elimination tends in some cases to be relatively constant, but this tendency is broken into by a number of peculiar changes for which I have as yet found no explanation. To illustrate this, let us run over the different tables. In Tables II and III the uric acid elimination sinks on the starch and cream diet to about one-half of what it is on the “ purin-free” milk and egg diet. Tables IV and V show a similar but less pro- nounced diminution. One particularly interesting fact is to be noted in connection with the latter two tables. These two experiments were; as will be seen from the dates, carried on simultaneously. But on July 7 both subjects consumed about 2 kgm. of potatoes instead of the arrow-root starch. One of the subjects being somewhat tired of the pure starch, this change was tried for one day.! It was thought that this would not materially change the analytical results, except with reference to the acidity columns and the ammonia. But the uric acid elimination was also distinctly affected, particularly in the case of Dr. A. (Table IV). The uric acid elimination in this case was almost doubled on the potato day. The rise in the case of Dr. H. is less pronounced but still noticeable. The smaller rise in the latter case is probably in some way connected with the interest- ing fact that the preceding fall in the uric acid output under the influence of the starch and cream diet was very much smaller than in Tables II-IV. In view of the marked change in the uric acid elimination which accompanied the substitution of potatoes for pure starch, it is interesting and suggestive to note that the diet used in the experiments of Burian and Schur and Sivén contained respec- tively 500 gm. and 600 gm. potatoes. Burian and Schur? assert that 1 Tt may be remarked here that the consumption of about 4oo gm. of starch in the form of potatoes, even for one day, proved a much more difficult task than the taking of the arrow-root, and the next day both Dr. A. and Dr. H. expressed a decided preference for the pure starch. 2 BuRIAN and Scuur: Archiv fiir die gesammte Physiologie, 1900, Ixxx, p. 269. fete) Otto Folin. in ordinary food, except such as contain “ purin,” 2. ¢., chiefly meat products, there are no other constituents which affect the “ purin” elimination. In view of the above-mentioned result given by pota- toes, when compared with the results given by pure starch, the experiments of Burian and Schur do not seem very convincing, although I am as yet not prepared to assert that potatoes do affect the uric acid elimination.’ Tables VI and VII represent no uniform definite diet, but only two preliminary experiments made to verify the findings in Dr. van 'Someren’s urine (Table 1). The diet was mixed, but so selected as to reduce the nitrogen, z. ¢., no meat products were eaten, and only limited quantities of bread (white), milk, or eggs. The uric acid figures here obtained can scarcely be said to disprove the contention that the endogenous uric acid is for each individual a constant quantity. In fact were it not for the experiments recorded in Tables II-V, and for some later ones, these two mixed diet experi- ments might be quoted rather in support of that contention, for the reduction in the uric acid elimination is indeed very small, notwith- standing that the total nitrogen elimination has been reduced by 7 gm. and 10 gm., respectively, z. ¢., from 10.6 to 3.8 and from 15.9 to 6.3. as Let it be understood that I am in no sense denying that a certain part of the uric acid eliminated in human urine is ‘“‘ endogenous” in the sense that it is a product of tissue-metabolism, as distinguished from the products formed in the breaking down of non-living protein in the animal organism. In fact I could not come to any other con- clusion from the uric acid values recorded in all these experiments. They all indicate that the more the total protein-metabolism is reduced, the more prominent in per cent of the total urinary nitrogen becomes the uric-acid-nitrogen, and the most probable explanation of such a relative increase in the uric-acid-nitrogen is, in my opinion, that at least a certain part of the uric acid represents the breaking down of living protoplasm which must be supposed to be essential to the continuation of life. We are, however, here considering whether a// uric acid eliminated on a purin-free diet must be of such origin, and with reference to this question my experiments answer in the negative. If the view advanced by Burian and Schur were correct, we should 1 The uric acid determinations on July 7, in Tables IV and V, were repeated and found correct. Laws Governing the Chemical Composition of Urine. 91 expect -a fairly sharp line dividing their endogenous uric acid from the uric acid produced under the influence of food. They believe that they have found such a line of division. I have not been able to find it. Even in Tables VI and VII it seems to me that a better case can be made for the negative side of the question. If we compare the three consecutive days having the highest total nitrogen in the urine, with the three days showing the lowest values, we find the following figures : TABLE VI. Feb. 1- 3. Total nitrogen, 29.8 gm. Uric acid, 1.63 gm. Kreatinin, 5.0 gm. Feb. 11-13. = it 13.4 gm. Se eleAG ron: 4.76 gm. TABLE VII. April 8-10. Total nitrogen, 46.4 gm. Uric acid, 2.3 gm. Kreatinin, 5.5 gm. April 31-June 2. sf ‘e ADM eyitig RS Merl aaa) 5 5.5 gm. The comparison leads to the same conclusion as that stated on p. 87, namely, that a pronounced reduction in the urinary nitrogen is accompanied by an increase in the per cent of uric-acid-nitrogen, but by a diminution of the absolute quantity of uric acid. Table VIII represents a feeding experiment with a patient, which is essentially similar to those recorded in Tables II-V. The uric acid elimination is here remarkably irregular, being at times as high on the starch and cream diet as on the milk and egg diet. A compari- son of the last eight days with the first four days recorded will, how- ever, show an unmistakable reduction. The result is 1.31 gm. as against 1.09 gm. (2.18 gm. for eight days). A noteworthy fact about this series of urines is that both the kreatinin and the uric acid are on the average noticeably less than the corresponding values ordinarily obtained from normal persons. This is especially true with regard to the kreatinin. Table IX bears out fully all that has been said with reference to uric acid. On the starch and cream diet the uric acid is reduced to one-half, or less, of the amount eliminated on the milk and egg diet. A particularly striking comparison is furnished by the last day from each series, — 0.18 gm. uric acid on the former, as against 0.45 gm on the latter. Yet the subject of the experiment did not even lose in weight on the starch diet. Table X supports the general conclusion with reference to the uric acid. This series is, however, rather more interesting as a study 92 Otto Folin. of metabolism in general paralysis. On June 13, the patient became suddenly almost unmanageable, and on that day there is to be noted a marked increase in the kreatinin, and a still more marked increase in the uric acid elimination. These nine different feeding experiments all show that the uric acid elimination is reduced when the total nitrogen-elimination is very much diminished. The reduction is, however, irregular and different for different persons. The metabolic processes that deter- mine the uric acid excretion may therefore be said to be in a rela- tively unstable equilibrium. Ammonia. — With regard to ammonia as a product of metabolism, the following statement may be formulated as representing the results given by the tables: With pronounced diminution in the protein-me- tabolism (as shown by the total nitrogen in the urine), there ts usually, but not always, and therefore not necessarily, a decrease in the absolute guantity of ammonia eliminated. A pronounced reduction of the total nitrogen ts, however, always accompanied by a relative increase in the ammonta-nuitrogen, provided that the food ts not such as to yield an alkaline ash. | The conditions governing the ammonia-elimination in human urine produce, therefore, results very similar to those which we have already noted in connection with uric acid, though the determining factors themselves are probably very different. In the study of ammonia as a product of metabolism, it must be remembered that this substance is a base, and its formation in the animal organism is therefore prob- ably quantitatively determined by the necessity of forming salts. Since we know that the human organism is capable of converting the ammonium salts of several organic acids into urea, we must conclude that the amount of ammonia appearing in the urine is determined by the character of the acids produced in the metabolism, 2. ¢., they can- not very well be acids the ammonium salts of which can be converted into urea and carbon dioxide. Such acids are, of course, primarily those derived from the sulphur and the phosphoric radicles of the decomposed protein-substances. But from certain forms of abnormal metabolism, and from the enormous increase in ammonia produced by the excessive consumption of fat,! we know that the formation of organic acids also may be equally effective in leading to an increased production of ammonia. 1 LANDERGREN : Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 125; JOsLin: Journal of medical research, 1go4, xii, p. 442. Laws Governing the Chemical Composition of Urine. 93 The peculiar feature of the ammonia-elimination, as illustrated in the tables, is the pronounced individual differences shown under the influence of the starch and cream diet. These differences I am not able to explain. There are no corresponding differences to be found in the sulphate or phosphate elimination, nor do the organic acids in the urines show any corresponding variations. Table II, showing about 350 c.c. tenth normal ammonia per day on the starch diet, and Table IV, averaging a full 100 c.c. less, show the same amounts of organic acids, z. ¢., about 225 c.c.; nor is the total acidity in the two appreciably different. But if the larger quantities of ammonia are not due to increased quantities of acids formed, then the excess of am- monia in one case above that in the other must be due to differences in the amounts of fixed bases eliminated by the two individuals. This much seems tolerably clear, but before attempting any further explanation, more analytical data must be obtained. Undetermined nitrogen. — Nearly all attempts made in recent years to determine the relative distribution of the total urinary nitrogen have been based on the use of phosphotungstic acid as a precipitant. This reagent is supposed to precipitate, quantitatively, all the nitrog- enous products except urea and traces of monoamido acids, and these are determined in the filtrate. Total nitrogen, ammonia, and uric acid are determined separately, and by a series of additions and subtractions it is held that reliable figures for the so-called diamido acids are obtained. I have never found phosphotungstic acid a very satisfactory reagent for quantita- tive urine analysis,’ and it must be admitted that the extensive inves- tigations pursued during the past ten years, in accordance with the scheme outlined above, have not materially advanced our knowledge of the composition of human urine.” By inspection of the columns giving the absolute amounts of nitro- gen left undetermined in the present investigation, it will be seen that the direct determination of the four most important constituents, urea, kreatinin, ammonia, and uric acid, makes a much more satisfac- tory system of analysis than the elaborate indirect system based on 1 FoLIn: Zeitschrift fiir physiologische Chemie, 1901, xxxii, p. 509. See also MOrnNeER: Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 330, and espe- cially the results of von JAKSCH: Zeitschrift fiir physiologische Chemie, 1903, Zipp: 123. , 2 It was on the basis of results obtained by this system of analysis that von JAKSCH came to the conclusion that the amount of urea present can be calculated directly from the total nitrogen. 94 Otto Folin. the phosphotungistic acid precipitation. With but two exceptions (Tables VIII and IX), we find that the undetermined nitrogen is always less than I gm., even when the total urinary nitrogen is as highas18em.or1ggm. A detailed study of the different constituents that go to*make up this nitrogenous residue would be very difficult in connection with such a comprehensive investigation as was here carried out. The only additional determination that could be con- sidered practicable, would be that of the xanthin bases, and even their determination is far from satisfactory, unless a considerable fraction of the daily twenty-four-hour quantity of urine can be used. From the variations of the undetermined nitrogen under the con- ditions prevailing in these feeding experiments, it seems clear that it must contain substances which, like kreatinin, or like ammonia and uric acid, increase in importance as the total amount of protein-metab- olism is reduced; z. ¢., the absolute quantity of undetermined nitrogen decreases under the influence of the starch and cream diet, butin per cent of the total nitrogen there is always an increase. When the total uri- nary nitrogen has beén reduced from 15 gm. to 3 or 4 gm., the unde- termined nitrogen has decreased from 0.8 gm. or 0.ggm. to 0.3-0.5 gm. Whether or no this relative increase in the undetermined nitrogen is due only to the xanthin bases must be left for future experiments to determine. It seems probable, however, that other substances show- ing a similar variation are present in this residue. Urea. — On the preceding pages it has been shown that kreatinin, ammonia, uric acid, and the undetermined nitrogenous residue all become relatively more and more important as the total nitrogen in the urine decreases. We will now consider wrca, the only nitrogenous substance which suffers arelative as well as an absolute diminution with a diminution in the total protein-metabolism, Urea, always consid- ered the chief nitrogenous substance, and supposed to represent almost 90 per cent of the total nitrogen in normal human urine, as well as in nearly all pathological urines, has here been reduced in ten different cases to about 60 per cent. It seems rather remarkable that this interesting behavior on the part of urea was not discovered long ago, because urines representing greatly reduced protein-metabolism have been studied by many differ- ent investigators. As far as I have been able to determine the reason seems to be purely a matter of analytical technique. This is clearly the case, for example, in the experiments of Landergren! made 1 LANDERGREN: Loc. cit., p. 112. Laws Governing the Chemical Composition of Urine. 95 only last year. The conditions of his experiments were almost identical with those here prevailing, yet when the urinary nitrogen had been reduced to 4 gm., the urea-determination made according to the Morner-Sjéquist method, gave 84.8 per cent of the total nitrogen as urea. But the subject under investigation weighed 71 kgm., and the urine must therefore have contained about 1.6 gm. kreatinin or about 15 per cent of the total nitrogen. Hisown determination gave the ammonia as 5.6 per cent and the uric acid as 3.2 per cent. The undetermined nitrogen must have represented not less than 10 per cent. The méthod used in these experiments for the determination of urea would therefore almost certainly not have shown over 68 or 69 per cent of the total nitrogen in the above-mentioned urine as urea. Had the diet used in Landergren’s experiments not contained potatoes (200 gm.), the ammonia-nitrogen would have been increased by about 5 per cent, and the urea-nitrogen would then have represented at the most 65 per cent of the total nitrogen. The reason why the remarkable variability of the urea-nitrogen, as per cent of the total nitrogen, was not discovered before, is therefore, I think, chiefly due to the fact that the earlier methods of determina- tion gave erroneous results. An interesting question to be considered here is, does 60 per cent represent about the lowest limit to which the urea-nitrogen can be reduced, or should it be possible to reduce it very much further? The answer to this question depends, it seems to me, wholly on the answer to another question, namely, whether it is possible to reduce the protein-metabolism to a still lower level? The results recorded in this paper show unmistakably that with every decided reduction in the total amount of urinary nitrogen the per cent of that nitrogen in the form of urea becomes less and less. If the total nitrogen could be reduced still further, there is every reason to believe that the per cent of urea-nitrogen should not only continue to diminish, but this diminution should become more and more rapid. To illustrate: If the Jast day on the starch and cream diet in Table IIT had given 1 gm. less of total nitrogen, there is no reason for supposing that this diminution should not have fallen almost if not wholly on the urea- nitrogen. We should then have had 2.8 gm. total nitrogen, and 1.3 gm. urea-nitrogen, z. ¢., less than one-half of the nitrogen would have 96 Otto Folin. existed as urea. In Sivén’s experiments! the total nitrogen was reduced to this figure, and his subject weighed almost exactly the same as the subject represented by Table I]. Another reduction of I gm. in the total nitrogen would under some conditions diminish the urea-nitrogen to less than 16 per cent of the total. Is there any reason for believing that the urinary nitrogen can be reduced to this extent? This can, of course, only be decided by actually accomplishing it. The analytical results recorded in this paper seem to me to indicate clearly the lowest possible limits that we can hope to’ attain for the nitrogen-elimination in man. It is conceivable that so long as any urea is eliminated it may still be possible to reduce the total nitrogen- elimination; but when the urea is absent, then the lowest posszd/e limit has been reached. Is it probable that any such peculiar state of protein-metabolism can be produced in feeding experiments with normal men? I am not in position to express an opinion on this question. It depends on whether all the urea eliminated is of the same origin, or whether a certain part of it is produced in the same metabolic processes that give rise to the kreatinin. Of great interest in this connection is the analytical result which I obtained a short time ago from a patient of this hospital. The patient had for several weeks taken extremely little food, and for five days preceding the collection of the urine he had taken no food of any kind, and in addition had abstained from drinking any water. 125 c.c. of urine was obtained, and this contained .85 gm. of nitrogen. Two closely agreeing duplicate urea-determinations were made. They showed that in this urine only 14.7 per cent of the total nitrogen was there as urea. 40 per cent of the nitrogen was, however, present as ammonia. A still more interesting urine is one analyzed by Morner? last year. This urine contained only 4.4 per cent of the total nitrogen as urea, with 26.7 per cent of nitrogen present as ammonia. Of course both these urines were pathological, and for that reason cannot be used as conclusive evidence; but they suggest the possi- bility indicated above of reducing the urea in the urine of normal men to a much greater extent than has yet been accomplished. ESSIVEN :, 0c. (Ctl. 2 MoOrNeR, K. A. H.: Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 314. Laws Governing the Chemical Composition of Urine. 97 IlI. Tue DISTRIBUTION OF THE SULPHUR OF URINE AMONG INORGANIC SULPHATES, ETHEREAL SULPHATES, AND ‘“ NEUTRAL” SULPHUR. We have seen from the preceding tables that the relative distribu- tion of the urinary nitrogen and sulphur among the different con- stituents depends upon the total amounts present, and we have discussed the nitrogenous constituents somewhat in detail. The sulphur constituents must now be considered. The sulphur in urine is frequently used as a measure of the total amount of protein katabolism, and not infrequently we find that only the sulphates (z.¢., the mineral and ethereal sulphates) have been determined because the so-called neutral sulphur is supposed to represent but a small per cent of the total. But it is also recognized that the sulphur-elimination is much less accurate than the nitrogen as a measure of protein katabolism because of the well-known fact that different albuminous substances contain very different per- centages of sulphur. Aside from such general studies of the sulphur in metabolism, many special investigations have been pursued since Baumann made the important discovery that the sulphur in urine is not all in the form of inorganic sulphates. Baumann’s own investigations were very extensive and resulted in the definite identification of a number of different conjugated or ethereal sulphates occurring in normal urine, and he also showed that they are all derived from aromatic products formed in the intestines by the action of bacteria on the protein of the food. The ethereal sulphates in general, and the indoxyi sulphate (indican) in particular, are therefore supposed to indicate the degree of intestinal putrefaction, the only subject of controversy being whether the per cent of ethereal sulphate or the absolute amount is to be considered the more important factor. More recently Blumenthal believed that the indican is a product of tissue-metabolism ; but this view Ellinger ! showed to be erroneous, so that the problem is at present supposed to be more or less definitely settled. The origin of the “neutral” sulphur, on the other hand, is still not understood, although this sulphur is supposed to be more or less directly derived from the bile acids. As a part of the normal metabolism, the sulphur derivatives are 1 ELLINGER : Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 44. 98 Otto Folin. therefore at present believed to offer comparatively few and unim- portant problems for further investigations. But the analytical results obtained in the course of these metab- - olism experiments show that in the sulphur, as in the nitrogen- metabolism, there are certain regularities in the distribution of the waste-products. In Tables II-V it is to be observed that as the total urinary sulphur is reduced, the per cent represented by the inorganic sul- phates sinks from about 90 per cent to less than 60 per cent. The similarity of the inorganic sulphate elimination to that of the urea is too striking to be overlooked. But the reduction in per cent of inor- ganic sulphur must, of course, be accompanied by an increase of the other forms of sulphur, and by inspecting the tables it will be seen that both the ethereal sulphates and the neutral sulphur are in- creased in percentage of the total, that of the former being more than doubled, and that of the latter increasing four or five fold. Turning from the percentage figures to those giving the absolute quantities, we find that the ethereal sulphates diminish as the total amount of sulphur diminishes, but to a much smaller extent, and that the neutral sulphur is not visibly affected by the diminution of the total. In other words, the ethereal sulphate elimination is analogous to that of ammonia, uric acid, or undetermined nitrogen, and the elimination of neutral sulphur resembles that of the kreatinin. The total sulphur-elimination, it will be seen, is, however, not nearly so constant as that of the kreatinin. J am inclined to ascribe this in a considerable extent to a lack of accuracy in the technique, because I have reason to believe that accurate sulphate determinations in urine are beset with greater sources of error than is generally recognized. In looking over a number of neutral sulphur-determinations recorded in the literature, I find even greater variations, variations so great, in fact, that they must certainly be due to analytical errors.’ These remarkable results can, it seems to me, be interpreted only on the basis of general processes of metabolism and not on the basis of intestinal putrefaction. It will be noted that under the influence of the starch and cream diet the indican invariably disappears altogether from the urine, while the absolute quantity of ethereal sulphates is reduced only to 1 See, for example, those of StvEN: Skandinavisches Archiv fiir Physiologie, LQOl,-X1, Pi 525- Laws Governing the Chenwcal Composition of Urine. 99 about one-half of the amount eliminated on the nitrogen-rich diet. Moreover, under the influence of the latter diet, the indican and the ethereal sulphates do not vary together. In Table IX, for example, we find the indican less than “50,” with the ethereal sulphates amounting to 0.32 gm., while in Table V the indican is “150” and the ethereal sulphates amount to only 0.22 gm. Although the indican undoubtedly exists as an ethereal sulphate in urine, the experiments here recorded clearly speak against the view that the ethereal sul- phates are all due to conditions identical or similar to those which give rise to the indoxyl] sulphate. On the basis of these analytical data the following conclusions seem warranted: (1) The urinary indican is not to any extent a product of the general protein-metabolism, is therefore probably, as is generally supposed, a product of intestinal putrefaction, and may consequently be assumed to indicate approximately the degree of putrefaction in the intestinal tract. (2) The ethereal sulphates can only in part be due to intestinal putrefaction, and neither their absolute nor their relative amount can be accepted as an index of the extent to which the putrefaction is taking place in the intestines. (3) The ethereal sulphates, on the contrary, represent a form of sulphur-metabolism which becomes more prominent when the food contains little or no protein. (4) The neutral sulpbur is not at all due to processes identical or similar to those which give rise to indican. (5) The neutral sulphur represents products which in the main are independent of the total amount of sulphur eliminated or of protein katabolized. If the deductions here drawn are correct, if the inorganic sulphates, like urea, represent products that are chiefly involved in quantitative changes of the total sulphur and total nitrogen eliminated, and if, on the other hand, the ethereal and still more the neutral sulphur have an origin similar to that of uric acid and kreatinin, then we must expect that the amount of ethereal and neutral sulphur eliminated is more or less dependent on the kind of food taken. We know that meat products, in contradistinction to most other forms of protein, contain precursors of uric acid and kreatinin, and that on meat diets the elimination of these substances is much increased. Are the ethereal sulphates and neutral sulphur also greater on meat diets than on diets rich in protein but containing no meat? I have not been able to find any literature bearing directly on this important point. Neutral sulphur-determinations have been made frequently 100 Otto Folin. enough, and the effects of different diets have been investigated. But the different percentages obtained can clearly not be interpreted as being due to the diets, so long as the effects of mere quantitative changes in the total sulphur were not recognized. The relative increase of the neutral sulphur on a bread diet, for example (Heffter), is thus to be explained, and does not at all warrant the conclusion that the sulphur of cereal proteins gives rise to different waste pro- ducts in the metabolism. Experiments intended to show whether meat extracts contain sulphur constituents analogous to the nitro- genous purin bodies and kreatin are in progress. IV. WATER AND CHLORIDES. Water. — The volume of urine eliminated depends directly upon the amount of water consumed. This self-evident fact needs no dis- cussion. But the variations in the volume of urine which cannot be traced to changes in the intake of water are of considerable interest. Such changes are at times exceedingly pronounced in the same individual, and different persons will, on the same diet, habitually eliminate very different amounts of water with the urine. Changes of this kind cannot be accounted for by changes in weight, and do not correspond to changes in the atmospheric conditions. For example, on comparing the first periods of Tables II, III, and X, we find that on exactly the same amount of water consumed the volumes of urine eliminated during the three days are 2555 c.c., 3770 c.c.,and 5500 c.c., respectively; yet the loss of body-weight is in each case about 300 gm., and the first two cases represent the same dates, thereby excluding the influence of the atmospheric conditions. Similarly in Table IV (first period) we find the volume of urine vary- ing from 780 c.c. to 2300 c.c. without change of diet and with but a slight loss of body-weight. And in Table III (second period) we find the volume of urine varies from 385 c.c. to 1880 c.c. I have observed that the greatest volumes of urine frequently occur on days when the body gains in weight. . But changes in body-weight occurring in the course of feeding experiments with normal persons are unquestionably largely due to cains or losses of water. That this is so is shown in the second periods of the feeding experiments Tables II to X. The starch and cream diet contained not over 2200 calories ; any extensive retention of fat is therefore practically excluded. Yet on this diet, involving Laws Governing the Chemical Composition of Urine. 101 the loss of considerable quantities of nitrogen, we find frequently no loss of body-weight. In Table IX, for example, we find a loss of about 40 gm. of nitrogen with the urine alone, in the course of thirteen days, yet the body-weight remained perfectly constant. In Table X, 33 gm. of nitrogen were lost with the urine during seven days, but the body-weight increased 600 gm. _ If this nitrogen should be calcu- lated in terms of muscle substance, it would amount to about I kgm., and this person would therefore have had to retain more than 1600 gm. of fat to explain the gain in weight. Positive proof that this person could not have retained this quantity of fat is obtained from the caloric value of the food compared with that of the fat. 1600 gm. of fat contain about 15,000 calories, while the entire food supply furnished during the seven days amounted to about 15,500 calories. In short, gain or loss of water is, within wide limits, an independent factor which may or may not coincide with the loss of nitrogen.' The volume of urine eliminated by normal persons is, therefore, largely a personal peculiarity, and is probably to a great extent inversely in proportion to the amount given off through the pores of the skin. The latter, the insensible perspiration, must therefore also not only be highly variable under different atmospheric conditions, as we know is the case, but it must normally be quite different for different individuals. Chlorides. — The chlorine determinations made in connection with. these feeding experiments have not yielded anything particularly new. It has long been known that the chloride-elimination varies chiefly with the volume of the urine, provided, of course, that the chlo- rine intake is constant. This conception is correct in so far as it is applied to any one individual, z.¢., on a uniform diet a person’s chlorine-elimination varies to a considerable extent with the volume of the urine. This rule, however, does not hold for urines from different persons. For example, in Tables II, III, and X (first period) we found the total volume of three days’ urine 2555 c.c., 3770 c.c., and 5500 c.c.; the total chlorine eliminated is, however, very nearly the same in all three, namely, 17.3 gm., and 18.6 gm. and 18.7 gm., 1 That muscle-preparations can, under different conditions, retain very different amounts of water, was shown several years ago by J. Logs: Archiv fiir die » gesammte Physiologie, 1894, lvi, p. 1. It is, therefore, not at all strange that the human organism should show a similar behavior, but it is important to know that a considerable loss of body-weight, as well as of nitrogen, can occur without signi- fying loss of muscle-substance or an impairment of vigor. 102 Otto Folin. respectively. This is only what we must expect in view of the fact that the volume of urine corresponding to a constant intake of liquid is different for different persons. V. THE RELATION OF PHOSPHATES AND OKGANIC ACIDS TO THE ACIDITY OF URINE. In the course of my metabolism experiments and acidity determi- nations I have arrived at a new point of view concerning the condition of the urinary phosphoric acid and the nature of the acidity of human urine. The accepted view concerning the subject may be briefly stated as follows : The normal elimination of acids and bases with the urine gives a mixture having an acid reaction, but on account of the tribasic character of phosphoric acid, and on account of the vari- able ammonia-production there is never any free acid present. The average normal result is an acidity equal to and due to a certain part (about 60 per cent) of the phosphoric acid present in the form of diacid- phosphate. Normal variations in the acidity are represented by vari- ations in the amounts, absolute and relative, of the diacid-phosphates. Monoacid-phosphate is a slightly alkaline salt, diacid-phosphate is a rather weak acid, and their constant presence together constitutes an admirable arrangement for preserving the acidity in the form of the comparatively weak and harmless acid-phosphate. The numerous attempts that have been made to procure a suitable and practical method for determining the degree of acidity have, accordingly, chiefly been directed toward a study of the acid-phos- phate mixtures in the presence of the disturbing alkaline earths (calcium and magnesium). The most interesting of such attempts. in connection with the present discussion, are those of Freund? and Lieblein.?, They determine directly the total phosphoric acid, then precipitate the monoacid-phosphate by the addition of barium chlo- ride, and, by a second phosphoric-acid determination in the filtrate, get figures from which they calculate the acidity. The interesting part of this procedure is that it would seem to constitute valid proof that the phosphates in urine always do exist as a mixture of the mono- and diacid salts,—a view which, moreover, as far as I know, has never been questioned. I believe, however, that I can now prove not only that the above method of Freund and Lieblein does not even approximately indicate 1 FREUND: Centralblatt fiir die medicinischen Wissenschaften, 1892, p. 689. 2 LIEBLEIN: Zeitschrift fiir physiologische Chemie, 1894, xx, p. 52. Laws Governing the Chemical Composition of Urine. 10 3 the acidity of urine, but that the results obtained by the procedure of adding neutral barium-chloride to urine have practically no con- nection whatever with the acidity, and further that the phosphate precipitate obtained on adding barium chloride to urine does not at all prove that a certain part of the phosphates is in the.dibasic form. The principle of the method is of course approximately correct. When a soluble calcium or barium salt is added to a dibasic phos- phate solution the insoluble di-calcium or di-barium phosphate is pre- cipitated, and the presence of monobasic phosphate does not materially interfere with the reaction. The principle is correct, but the pro- cedure as applied to urine is nevertheless grossly erroneous. The chlorides of calcium and barium show a similar behavior toward dibasic phosphate solutions; but in the study of the urinary phos- phates, barium chloride is always specified as the reagent to be used. This choice seems strange. One would think that calcium chloride would be much more suitable, because the barium salt gives also a very abundant sulphate precipitate which so hides the presence of precipitated phosphate that mere inspection does not show whether or not any has been formed, and one must rely exclusively on the subsequent phosphoric-acid determination in the filtrate. The reason for this peculiar preference for barium is, however, not very difficult to find. Bariam chloride does, calcium chloride does not, give a phos- phate precipitate with normal, clear, acid urines. Why this important difference has been passed over in silence by the numerous investi- gators who must have observed it in connection with their studies of the acidity and the phosphates of urine seems almost inexplicable. The idea that dibasic as well as monobasic phosphates are present in normal urine is so firmly fixed that the failure of calcium chloride as a precipitant probably has seemed to each observer accidental, and due to the presence of some disturbing factor that he did not care to search for. Asa matter of fact, I myself failed to quite comprehend the full significance of my own analytical results in my paper on the acidity of urine because I could not doubt the correctness of the seemingly unquestionable existence of dibasic phosphates in urine. But having once fairly questioned the correctness of this view, the whole problem of the behavior of the urinary phosphates and their relation to the acidity appears in a new and seemingly clear light. Calcium chloride does not give a precipitate with clear acid (litmus) urines, because the total acidity in such urines is equal to or greater than the acidity of the total phosphoric acid present, ex- 104 Otto Folin. pressed as diacid salts, and therefore also there is no dibasic phosphate there. The results obtained with barium chloride in such urines repre- sent nothing more or less than errors due to the presence of sulphates. Barium-sulphate precipitates have long been known to carry down all kinds of impurities, and I find that this is particularly true with reference to barium phosphate. If barium chloride is added to a pure standard monopotassium phosphate solution, no precipitate is formed ; but if some ammonium sulphate is first dissolved in the solution, the barium sulphate formed on adding the barium chloride can at once be removed by filtration (in contradistinction to pure barium sul- phate), and 30 to 40 per cent of the phosphoric acid is missing in the filtrate. This is evidently the reason why the Freund-Lieblein procedure always shows the urinary phosphoric acid to be present as a mixture of monoacid dibasic phosphates. Soluble phosphates do, as a matter of fact, exhibit a very strong tendency to form precipitates with calcium as well as with barium, even when the solution contains no dibasic phosphate. For example, if to a pure monopotassium-phosphate solution is added first sodic acetate, and then calcium chloride, a precipitate is at once formed, and this precipitate dissolves only to a very small extent on adding acetic acid. If the acetate and acetic acid are added first, no precipitate may be formed when the calcium chloride is added, but it will appear when the solution is heated. This phenomenon is occasionally observed when the heat test is made for albumin in urine, only here the amounts of calcium and of organic salts present are small, and the phosphate dissolves more or less readily when acetic acid is added. Even in such cases there is clearly no reason to assume the presence of dibasic phosphates. Sometimes almost perfectly clear urines are obtained which give an abundant precipitate with calcium chloride. Such urines are, however, not ‘acid to litmus, and the reason why they are clear is that they contain practically no calcium. They give no turbidity when sodium hydrate is added. From these facts we may make the following summary: The phos- phates in clear acid urine are all of the monobasic kind, and the acidity of such urines is ordinarily greater than the acidity of all the phosphates, the excess being due to free organic acids. The current attractive, and in a measure plausible, belief that the acidity of urine is regulated by variations in the relative proportion of the two forms of ‘“ acid-phosphates”’ is therefore erroneous. If urine does at no time contain comparatively strong acids in the free Laws Governing the Chemical Composition of Urine. 105 form, the reason is in part the variability of the ammonia formation and in part the presence of salts of organic acids. Ina mixture of salts containing an excess of acids it is the weakest which will remain uncombined, and the strongest organic acids will therefore exist as salts; but if the total amount of acidity becomes abnormally great, the quality (the strength) of the free acids may change. That this may actually occur in certain pathological cases (kidney diseases) is indicated by the hydrogen ion determinations of Héber.! For all ordinary studies of the acidity of urine, the direct titrations of the total acidity and of the phosphates give the necessary informa- tion. The excess of the total acidity above that calculated from the phosphates (7.1 mg. P,O;=I c.c. 7 acid) gives the total free acids present. In the tables forming a part of this paper, the data obtained by such a calculation had to be omitted for want of tabular space, The data in question can, however, easily be obtained as indicated. Here it was thought more important to show, on the one hand, the balance of the mineral acid and basic elements, and on the other, the total amount of organic acids present. The excess of the inorganic acids expressed as “ mineral acidity ” decreases, it will be seen, under the influence of the starch and cream diet until it frequently becomes a minus quantity. The organic acids do not diminish to nearly so great an extent. Of special interest in connection with the organic acids are the results of July 7, in Tables IV and V when the pure starch was replaced by potatoes. The total acidity of the urine is but slightly reduced, and the total amount of organic acids eliminated is greatly increased, yet the ammonia is still considerable. This would seem to indicate that a part of the acid and ammonia formation within the organism is not affected by the alkalies of the food. 1 HOBER: Beitrage zur chemischen Physiologie, 1903, ili, p. 539. en er tt 7. 106 Otto Fol Ww 8 FOL v9 9G Lil £6 8 Oc es 6¢ Orel $6 Sie Stl 08 oe Oc OTL CC SL $8 o£? 0'8 cs VG 66 +6 9% U8 LL 9? U8 SL aS £8 6'9 0% 0's LL 0’? 89 L’8 ST 6S CL Dat Wy) eure on UIULJVALST -19j}9puy) ‘NIDOULIN TVLO], JO LNAO WAd NJ 108 Vm Boilie SSO" ora) MI eeyyeel epee Ina 0°08 DEE | 360221 SeS0) 9) CL ye ieee ae een! eee S61 Sy | OP a S180 8) geo rere Ie eens leas aL ees MIME NCE OL al) Son eae eee Tn eel FET Me sg9 Woo. I Noe meneame | ee OLE SE ace eesOr ol = eyo Wegnce ee yare |may 0°64 Sz ZOb|| 26200 \\ ee. hg eal oe amok eg Qe geen cieeso || eee rele eee S'6L OF See Mev Oral ge Nie ence We ee | eee 66L OF 6'SL LS°0 lec oe ee a Lg BASIE TENS Natal ee AC wll os s'T8 SSR (OGY | OM abuse are cS Ne aa Tes Uv O'6L c5'0 be wo oe Bo) S€8 O'+ S6L bL'0 on Hie Doce aoe ores ieee aie |lWmese On, meee peal Merete goer | eee O'+8 a 96L 0L'0 ee oy OT SOE By, <= ens a Re bale os. pourra : = -19}9pUuU /) “ug wus ‘us ppe oA | wunrery | viuowMy | “vag ‘(uosdag JoUuso0ny7) “WT *O ‘IA WIAVL ~N ‘uas -O.1QLU [PIO S{6 oT € rL9 is SI ee i Hee wall Gi ( ee ZI 0¢9 oe Ul SNe 0 | OS ok 6 cea eee OSS be | L SQ/ oLo | : gf We Sel Ze + ss “ Peet =e eam 6 049 I Eé 9°09 0 | gos ‘outIn jo | ‘ayrq auINnyoO A | ‘- Re a pete wa 13 ‘dg WYSII AK ™ eg i) Be g's él IS+ +8E— LST ELI ST 604 610 ra 0) 8L'0 COLL |r ; "90Rl] 8s eT OSE bEo— ral S+1 L6 SSL STO a0) +6'0 til el Wee 8 "20R1} os rT £6 vee— 6ST ell 66 bbl 610 co 060 I@'T ial S *90C1} 09 UT ISt sLe— 9L FST 16 SSL LTO +; OTO £80 OP Lt acer =~ “20e1} SP L0 bOb 60¢— S6 SLT fOr CCL LTO Ovo 0L°0 L6'0 él S "29¥.1} os OT 90+ EOGes 101 SSI +8 TEL LTO 80'0 890 £6'0 ll S$ "B0R1] i) IT IIt FLo— Let o'eT Sl OSL £10 cl0 9L'0 I0'T Ol & "90B1} See OT +e Sto— OL ee! Sel TOL +0 a0) 190 LS‘0 6 S "908.1 6% Lp! C6¢ c8— O1Z est Ved TTL stO 90 +8'0 ST 8 = "908.14 I+ oat 994 wOE— oO Lv TOI SL 90 Il0 c8'0 60'T L 8 Ol ok cl Ott 08¢— 991 SFI 901 OL sv0 £0 160 col 9 § S 6+ ol LOE 89¢— 6c £1 L'8 008 £0 0) 160 valle s S) Ol 6+ Sal See OST— Ss 66 OL oCS8 0) $0'0 £80 10'T 14 S Ol TL ST 60€ tcl 98T FOL 96 0°08 LV0 910 Oe T £9'T € Sy) our L's ST elt Soles 8Lé FOL | 16 $08 LV0 STO cel val c K soos os LT Sth LOT— SHE PAO! SL TI8 cc'0 a0) OFT "08" I 8 ® “(001 = So%g ||mesio| ‘rerun | TTP || 8g | Fg | tg r 4 Hee. bac So QS sg yreur | URNS “wi se 1P90.L aS 55 Pe Seime | SES ; . g [Pen | oa |e sereas lela] abe rs uvoIpuy [e10, % OO NI ALIGIOV IVLOL dO LNXO Wad NY « [eNIN ,, | [Pa194Ig | OtUeS.1OUT [e10.], “panuyuod—TA ATAVL ' n Lin. Otto Fo ‘ploe ota) ‘UTUTVIIST “PLUOWWUe + vag “eLUOWWUU YY “NGOSOULIN’ TVLOT, tO LNYO Wad NT udsoO1}IU pourut -19y9puy) *(uosdag yousony) * *N-urU “years -owwly a S +] re ™~ O Nin Nin tin ™~ RN O6 5 H NGS “wd “plore omy) TIA WIdViL “ws “UUIRaLYy lel tak pt eG | HN $f ‘0/9 “PruowWlW ‘eT ‘u23 |) suinjoa -O1}1U [R10], 109 of Urine. 70n O 20, Laws Governing the Chemical Compos ‘paonper yonut Alaa peoiq pure s[eva199 [TV 0 0 ‘9081 9081) OLF OFS LSv ICS 2 lm! ~ @ o. < © OQ ) or 2 a. p =} Q @ sD *payoiyjsor AyueN?)® 'syieu ay | *2081} "2081 *20R1} ‘901 9081} 9081} *9081} 2981} 2081} 2081} 9081} ‘9081 69+ Sb +8+ £S8h L6S bLy FL £19 OSb ITS LSb for ((0;0) uolqyn[os $,3ul -1494) uevoIpuy ‘OIURSICO | ‘[eIOUL *pazegy [TRIO L “9X 5°90 NI ALIGIOV “MNHdTNS IVLO]T, JO LNAD Yad NI (Panu eoD — TN SAL Los « [eQNaN, ,, FOS a aos [eo19 yy q | Otues10uy «S ” ‘Og se Inydjns TPIT Otto Folin. ITO. ete aes a OL1 6'S9 +8 S109 |), $8605 oho ahaa eee eee Gn 4558 O91 1°89 +8 poo irocan | CRORE ChE) 20 ee FOr | 2e FOI 694 6+ OCP AGHOL tons © meee ek. Moca | nee 6zt | 02 Sel ELL D RICO IP OPO. tl sere] eral tee a7 atc Ge elece ie Lei ers TO) Prey Ue eae a ea et Coun mie s0l LLL £6 Vell ZOOM tee rina veal pean lino 691 | siz ZI 0'89 9 GND) HO je eee eet eee 98 | 62 Tal OL $9 669) |) dz. 4) SOO ees a CeO eae Om ale 2 el S'€8 OF Ohh PEO” It son nyl apes coe aeemleg g9 | ST UL esg og O28 NE CBOs ee ees ailenene Ge Si 99 $88 6% SEB h I MZ Hl "Pcie een | oneal th meee "oy | LT L's L'18 6€ Beet || PEO MN Gree lh epee le are euler #9 «| 80 SE 0°68 oe y see cOO |) Seep | ele ae ee ee Ly | 80 org S16 Be ges || 190 | NRO ee eae ie Bele ez0 Sz Z'€6 bt S Ge Il Sv: ||) eee ee eel ye ze | so as 116 Le Gee oi SG ieee eee ee | serene || ae am) Bi) Ze $88 org Seco NM cOnt, vy SOO | Ud eee |) Sel spo | 29° |uuneony | ome | ouomny | eon | | APY A | AY coin = pourut ‘NHDOULIN IVLO], dO LNXO Wid NT ees Pea Sanaa aN ar *(70a10T) “DH ‘'N “AW ‘THA WIdVL “SN ‘uas | 20D, -O.1QTU | oss 6¢ Zr org ale OLZI ae Lé mL | 6 gre O90T pa +2 SLY 92 £7 SOL 7 By CC OFLT ref IZ O€Tl gr 0% | OTOL gr 6l eee OFS ¢ ie IT | ozor | st | ée rt OLOT | 92 bT || O9€T . Wl oae ST | O88. |) 21 | ox ot9 1] —- | ‘o"9 udy | aurIN Jo | ‘a1V(] awnyo A or | sony ur “3 -dg IYBIIA\ — . o oie i icc | BO || ean AVAIJ | : WSS) | “99e1y j 9T0 60 . bz'0 a0 “s SS 02 0 600 +20 600 > . : 1Z'0 100 aS Z10 60°0 sS mal £20 60°0 ££°0 0) zo | 610 zo | +20 %0 | +20 OLE 3 | | 60 8c'0 ‘Jelp WeIID pue YOIeP}S QtZ 820 620 | SSI 6£0 £0 ee : 697 |, 2 8£°0 ze'0 wm |) gre | 9¢5 ; o¢'0 620 ‘yoIp 33a pue ILI paqeny 2 ( = : ‘OIURSIQ | ]e10UL ; pe, ri (OOT RA ASON NES SUNT ionap cigs pn. WOoLjN[OsS $,8ul 1494) Prorcee ueolpuy u Psy ur 8 8 Sana ys OS se she : OS OS OS ‘ wand TAs «WANN, | ~eor9y7%q | OTURT10UT anydins SE een IVLO], 40 LNHD Add NJ MIO Laws Governing the Chenical Compost ‘panuyuo) —TITIA ATIAVL ut li Otto Fo I2 SOND GG GOO OnMAN OStFMAAPMAHP AI POMOM x ~N NN coise oxen Ns d pie WO _ > NEN o) io N i. wun hy NONDNONONSG HUD com NVuy 2 2 Ha Hea He oes WS TUHOKNDS ) ne 2 NE MIO OOOM~ a 8 Scel or Stel Or OSL §f 0202 6 OSSL or Sle OL OFSL &f OFST ZI Sé9l or ONT pouruu -lajapu yl) ‘plore OL) ULUT] BOTY “PLUOWUULe + vai “eIUOWWY ‘NEDOULIN IVLO]J, dO LNAD UAd NI ‘udsO0.1QIU pour -1ajapuy Pl "N-proe O11) *ay-ULU AMAL Dk "N-PRLU -OWWY “us “plow Of) ‘(7uarvd) “I *H ‘N ‘YN ATIVL ‘ws “ULU BIL “fe 7 OT +m: HN % ‘99 "BLUOWUW ‘N-vOIQ) ‘ON ‘usd -O.TU [P1901 “919 pour jo suNnyO A ‘oun[ ‘aye "—O'! | ud ‘ds *SO[IY UL WYSIA MA ihe) S aS S eS) = NY ms) 8 ss 6 & S S Bo 8 ws ~S N S OS & S x ~N S iS) s S 6 3 8 N) 0 ie 60 CLI oI sot || sist | vet | 649 ZU0 sto | oso 110 || 91 nm | 0 Os | f0'r ZZ co- | wan || vor | sun | sig9 +0 sto | ss0 $80. || ST a alee re | O01 262 g- | “er || 92 | OF | S29 0z'0 er0 | 950 680. || 41 all to as Oa zz | os- | aziz || ote | ett | wes 00 +10 | oso +60 || ST 0 OF | TT Sez w— | 961 || 9'e¢ | 941 | 819 1z'0 ero | sso 6s0 |! Zt Peel eee |) mee Gye || e— || exe ll cet | Fi |) Te9 ST‘0 ro | s9°0 +60 || TL % | -aoen | 92 | gt yee | PSL goc || soz | 26 | eos 220 oro | 920 sot || ot Eppa || Oka Weste 612 0sz 6p || woz | sor | ¢69 Ig0 910 | 901 eo Teal ae ee] wecee lie vc) eee see £0 Sevimiliesoe || ce |\mreigs 9¢°0 910 | +e ore || 8 a ce OO | er Ie al mGez geo || wor | +h | 6+8 IFO FO) oa8 we || 4 & | o¢ | as | z+ Ze tee er. || ss | +9 | ss +0 so | ese sse || 9 a Oza Sispell) 0% ssh | gs2 Gi, |e nz On 008 0e0 0 | I¢¢ Gas pS = sor) 5-1. Olt cog SW Woe ee || Yacis 6Z'0 sco | +1¢ ze || + ‘(oor =| 4o%g |fomesio| tesoun| Vor || *s- | #8 Hs ee ee | oS me cae : eee soreud ye 8 2 FOS se || “ered fab | M9 | iS |) ak oo xr xstaOy — |faycoy METIS a yy EON | eo aruediouy| “EU *‘panutjuod — XT ATAVL Otto Folin. 114 G3 | y2z aa T'SL OL TU? | SOP ores. of netee lO ganas late elmer Pee a 7 me | ai LOt €'18 HOW GOL NF TSO” I eon oll apg | eens | meayes NeisG 9||ecome SI . * . Gs Se) |) el 90L O'8L GL DOLE MPO Eee | pee Neenneee ean, allay Nl nore HI . Q' . Cc | Ciara Gil O01 6.6L OL SCL | GO. “Ne Ce RE Ae a ees oon ipaceas eT 69 | OT L9 +'S8 rs OOS |i SFO: «|! ANP eee re eee 69 GOs ZI ies) IP Ow 6+ +68 Us SH | OS Oe ae au ree le ae OF O68 cog ee Ganall ie a ST S0 re 9'+6 or¢ O16. I ze | TER. | eep | Saegt|| coer | cornell iaeeneen (amine . at sc . eiee | ey) rs | 90 org 106 o¢ A Oe ss pe | er ee | S90 ll eee 6 = . . . . : 7) e) ho | LO | 9 0°86 Ly Geer Nae |e Ce ne woe m eye eae I cet | Oeste |e aa (RIOe “ULUTR 9.05] eee ‘eIUOWI YY | “vor “Ws CE ba ee AEN ‘N-eo19Q | “@N ‘air jo ae f -rajapug | 24 a1 os ° + Pat ; uesonm| IF) Heol | -owwmy | ~ uad |/ounjo, |. wd | ees ine ae _ pourut ie : “OI || an ats ; | A979puy} “us "wd "8H N SE -o'9 "wd THIOL || -on *SO[LY UI ‘NADOULIN ‘IVLOL tO LNAO Yad NY] ‘uunvary | emowmy | -voaq | | aa-dg | ays “plor O11) ‘(sestpvavd posouas quayvgd) *{ *g *) “aIN *X HTa VL i ih 8 Ss Ss wn 80'0 éS'0 ~» ~ 3. 9T0 0 0) 650 98'0 Ss S. LV0 +10 cs'0 68°0 BS g +20 £10 LOT SET S = 60 ITO 160 ial S & 910 910 00 eT S) . S |ez ze0 €0 | Ze LS'€ aS seed ; Soules 262 : ; WO nec | te | are (ie) - < 7B) OF eeallaeere sto zo | ez | eee i) *peyeiqy SS = i . ty ‘TRIO , . = . ~ sort D1URSIC | ‘[RIDULTL [210], 8S» S55 IS «So» s,3uly : sajeyd 8 8 2 Osi se Sul Os OS OS -14y9q) -soyd Site ‘ToaHdaIO c inydins uvoIpUuy [e10L, “YO NY ALIGIOV IVLOJ, 40 cao uaa NT ||“ POUND,» Tede He MCL [e101 (POU UOD Xe Alay sls Laws Governing ¢ AWebibORyY OF PROTEIN, METABOLISM, le LOMNUO) IONE [From the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Massachusetts.) N the preceding paper! on “ Laws Governing the Composition of Urine,’ it was shown that urine obtained from normal persons does not necessarily exhibit any such constancy in composition as has been supposed to be the case. The analytical results recorded in that paper prove that quantitative changes in the daily protein katab- olism are accompanied by pronounced changes in the distribution of the urinary nitrogen and sulphur, and that the variations occur according to laws that can be formulated with a fair degree of pre- cision. The nature of these variations may be expressed by the following generalizations: 1. Kreatinin. The absolute quantity of kreatinin eliminated in urine on a meat-free diet is a constant quantity, different for different individuals, but wholly independent of quantitative changes in the total amount of nitrogen eliminated. 2. Uric Acid. When the total amount of protein metabolism is greatly reduced, the absolute quantity of uric acid is diminished, but not nearly in proportion to the diminution in the total nitrogen, and the.per cent of the uric acid nitrogen in terms of the total is there- fore much increased. 3. Ammonia. With pronounced diminution in the protein metab- olism (as shown by the total nitrogen in the urine) there is usually, but not always, and therefore not necessarily, a decrease in the abso- lute quantity of ammonia eliminated. A pronounced reduction of the total nitrogen is, however, always accompanied by a relative increase in the ammonia-nitrogen, provided that the food is not such as to yield an alkaline ash. 4. Urea. With every decided diminution in the quantity of total nitrogen eliminated, there is a pronounced reduction in the per cent of that nitrogen represented by urea. When the daily total nitrogen 1 In vol. xiii, p. 98, line 24, read “neutral” instead of “ total.” ? FoLin : This journal, 1905, xili, p. 66. 117 118 Otto Folin. elimination has been reduced to 3 gm. or 4 gm. about 60 per cent of it only is in the form of urea. 5. Inorganic Sulphates. Decided diminutions in the daily elimi- nation of total sulphur are accompanied by reductions in the per cent of that sulphur present as inorganic sulphates. The reductions are as great as in the case’ of urea. 6. Neutral Sulphur. The neutral sulphur elimination is analogous to that of the kreatinin. It represents products which in the main are independent of the total amount of sulphur eliminated or of protein katabolized. 7. Ethereal Sulphates. The ethereal sulphates represent a form of sulphur metabolism which becomes more prominent when the food contains little or no protein. In order that the reader may recall the extent of the changes in percentage composition of urine from normal persons shown by the experiments of the preceding paper,a part of Table III is here reproduced. juExeds: JuLyY 20. Volume or urine — 1170 c.c. 385 c.c. Total nitrogen 16.8 gm. 3.60 gm. Urea-nitrogen 14.70 gm. = 87.5% 2:20 em: =(61-ng, Ammonia-nitrogen 0.49 gm.= 3.0% 0.42 gm. = 11.3% Uric acid-nitrogen ONS iome — was 0.09 gm.= 2.5% Kreatinin-nitrogen 0.58 gm.= 3.6% 0.60 gm. = 17.2% Undetermined nitrogen 0.85 gm.= 49% 0:27 emiy—s su, Total SOs 3.64 gm. 076 gm. Inorganic SO; 3.27 gm. = 90.0% 0.46 gm. = 60.5% Ethereal SO, O19 pm = 5:24, 0.10 gm. = 13.2% Neutral SO, 0.18 gm.= 4.8% 0.20 gm. = 26.3% It is clear that the laws governing the composition of urine repre- sent only the effects of other more fundamental laws governing the katabolism of protein in the animal organism. From the variations in the percentage composition of urine described by the generalizations, it would seem therefore that some conclusi be drawn in regard to the nature of protein metabolism. It purpose in the present paper to attempt an interpretation of prot metabolism on the basis of observed variations in the percentage composition of urine. j We have at present two fundamentally different theories concern- ing the nature of protein metabolism, namely, that of Pfluger and that of Voit. The theory of Pfliiger is essentially a modification of an A Theory of Protein Metabolism. 119 earlier one advanced by Liebig, and is very old. The theory of Voit was first formulated in 1867, after the original theory of Liebig had become untenable, and for a long time Voit’s theory enjoyed almost universal acceptance, although it, too, had to be modified in order to be consistent with the facts brought out by Pfliger. Since 1893 Voit’s theory may be said to have lost ground. In that year Pfliiger? published an exceedingly searching criticism of nearly all the facts which Voit had advanced in favor of his theory, and showed that they were either erroneous or capable of a different interpretation. This criticism, accompanied as it was by the experiments of Schén- dorff, has never been refuted. The two theories are briefly as follows: According to Voit, the protein of the absorbed food passes through the blood to the different tissues and cells, and is there katabolized under the influence of the living protoplasm, but without first becoming an integral part of the latter. Voit’s fundamental conception seems to me to be that the living protoplasm is in a state of suspension, the “ circulating protein” is in solution, and the chemical decompositions that con- stitute protein katabolism take place only in solution. The small amount of living protoplasm which dies in the course of twenty-four hours is at first only dissolved, thereby becoming a part of the circu- lating protein derived directly from the food. Pfluger, on the other hand, believes that there is a very decided chemical difference between circulating protein and living proto- plasm. The former is comparatively stable toward oxidizing re- agents, while the latter is in a very unstable equilibrium and is particularly susceptible to oxidation. All the protein katabolized is first transformed into bioplasm, becomes an integral part of the living tissue, and only as such undergoes the oxidation that is supposed to constitute the most fundamental chemical decomposition of protein katabolism. It must be stated, however, that Pfliiger can scarcely be represented as unqualifiedly committed to this theory. He is generally so represented, but chiefly on account of his ex- tremely hostile and uncompromising attitude toward the theory of Voit. In the latter part of the paper published in 1893 (pp. 414- 419) he practically admits the possibility that a certain amount of protein may be katabolized in solution, z. ¢., without having acquired the structure of living protoplasm. The chief, if not the only positive evidence seit Voit’s theory, 1 PFLUGER: Archiv fiir die gesammte Physiologie, 1893, liv, p. 334. as 120 ; Otto Folin. and in favor of Pfliiger’s, rests upon the experiments of Schéndorff.! These showed that when the blood from a starving dog was passed through the hind legs and liver of a well fed dog, the urea contents of the blood were increased, while no such increase was found when blood, whether of starved or well fed dogs, was passed through the hind legs and liver of a dog that had previously starved for several days. This result obtained from sixteen experiments has been generally accepted as proving the point which Pfliiger and Schérndorff intended it to prove, namely, that it is the state of nutrition in muscle cells and not the circulating protein which is the determining factor in the protein katabolism. It seems to me, however, that the evidence furnished by these experiments is by no means unassailable. After a dog’ has starved several days his protein katabolism has of course been reduced to an extremely low level. When such a dog has been bled to death, and his legs are washed out with normal salt solution, it does not seem very peculiar that the passing through of defibrinated blood should not revive the protein katabolism in the legs to a demon- strable extent. Schéndorff’s well fed dogs, on the other hand, had been taking enormous quantities of meat, 1000 gm. or 1200 gm. a day, and special efforts were made to have them at the height of digestion when killed. In these dogs, therefore, we must assume, according to both Voit and Pfliiger, that the protein katab- olism taking place in the muscle tissues of the legs at the time of the killing was enormously increased above.that in the tissues of the starved animals. That a certain amount of nitrogenous katabolism products should be obtained on passing blood through the legs of such animals seems to me by no means to prove Pfliiger’s point. The evidence obtainable from these experiments depends not on whether there was an increase in the urea contents of the blood passed through the muscles of the well fed animals, but on the quantity of that increase. Schéndorff does give the increase in terms of the urea present in the blood at the beginning of each experiment and thereby gets figures which seem very striking indeed (100 per cent or more), but that is only because the original urea content of the blood was insignificant. In his last experiment (No. 16, p. 481), however, Schondorff gives in addition the absolute quantity of urea which he believes must certainly be held to represent the protein katabolism occurring in the muscles of the dog while the blood was * ScHONDORFF: Archvv fiir die gesammte Physiologie, 1893, liv, p. 420. A Theory of Protein Metabolism. T20 passed through. This dog weighed 18.5 kgm. During the last four days preceding the experiment he had taken 1200 gm. meat per day, and during the last nine hours immediately preceding the experiment he had taken 700 gm. of meat. In the course of four and a half hours the blood was passed through the legs sixteen times. The urea obtained amounted to 53 mg., the urea nitrogen therefore to 25 mg. This dog had during the preceding four days katabolized in the neighborhood of 35 gm. of nitrogen per day. Expressed in terms of this nitrogen, the two hind legs had during four and a half hours katabolized less than one-tenth of 1 per cent! Considering the numerous sources of error and uncertainty necessarily attached to an experiment of this kind, it seems very strange that the extraction of 25 mg. of urea-nitrogen from the hind legs of a dog killed while engaged in digesting 700 gm. of meat should be accepted as proving not only that protein katabolism did occur during the experiment, but also that it occurred in the bioplasm and not in the circulating protein. The chief positive evidence in favor of Voit’s theory not disproved by Pfliiger is the well known fact that extremely large quantities of food protein are more or less completely katabolized in the course of a few hours. It is considered incredible that such enor- mous building up of bioplasm, accompanied by immediate destruction of it, can take place in so short a time, especially since the previous state of nutrition of the organism has practically nothing to do with it. While the theory of Voit is no longer so generally accepted as was once the case, it undoubtedly still has a greater number of adherents than has the theory of Pfliiger. . The analytical data obtained in connection with the metabolism experiments described in the preceding paper seem to me to have a direct bearing on this general fundamental question concerning pro- tein metabolism. No fundamental theory concerning that metabo- lism can be accepted unless it can be made to harmonize with the laws governing the composition of urine. If heretofore no objection has been raised against either Voit’s or Pfliiger’s theory on such grounds, it is only because the accepted views concerning the com- position of urine agree sufficiently well with either. At the time when these theories were advanced, nothing inconsistent with their correctness was known in connection with the composition of urine, ‘nor are the views concerning that composition to-day essentially dif- ferent from what they were then, z. ¢., quantitative changes in the 122 Otto Folin. total protein metabolism have been supposed to have no appreciable effect on the percentage composition of the urines representing such metabolism. But this is clearly not correct. Quantitative changes in the protein metabolism, or in other words, changes in the total amount of nitrogen and of sulphur eliminated in the urine, are accompanied by pronounced and unmistakable changes in the per- centage composition of the latter. But this fact constitutes, it seems to me, definite proof that neither the theory of Voit nor that of Pfliiger can be correct. We have seen from the tables that the composition of urine, representing 15 gm. of nitrogen, or about 95 gm. of protein, differs very widely from the composition of urine representing only 3 gm. or 4 gm. of nitrogen, and that there is a gradual and regular transition from the one to the other. To explain such changes in the composition of the urine on the basis of protein katabolism, we are forced, it seems to me, to assume that that katabolism is not all of one kind. There must be at least two kinds. Moreover, from the nature of the changes in the distribution of the urinary constituents, it can be affirmed, I think, that the two forms of protein katabolism are essentially independent and quite different. One kind is ex- tremely variable in quantity, the other tends to remain constant. The one kind yields chiefly urea and inorganic sulphates, no kreatinin, and probably no neutral sulphur. The other, the constant katab- olism, is largely represented by kreatinin and neutral sulphur, and to a less extent by uric acid and ethereal sulphates. The more the total katabolism is reduced, the more prominent become these representa- tives of the constant katabolism, the less prominent become the two chief representatives of the variable katabolism. The fact that the urea and inorganic sulphates represent chiefly the variable katabolism does of course not preclude the possibility that they also represent to some extent the constant katabolism; but I have reason to believe that it is possible to plan feeding experiments which will yield urines containing very much smaller per cents of these two constituents than I have yet obtained. We know from the experiments of Sivén that it is possible to reduce the total protein katabolism still more, and I am confident that in such cases the per cent of urea-nitrogen will sink still lower, and that the nitrogen of the other constituents, particularly of the kreatinin, will again show a corresponding increase. If there are two distinct forms of protein metabolism represented A Theory of Protein Metabolism. 123 by two different sets of waste products, it becomes an exceedingly interesting and important problem to determine, if possible, the nature and significance of each. The fact that the kreatinin elimination is not diminished when practically no protein is furnished with the food, and that the elimination of some of the other constituents is only a little reduced under such conditions, shows why a certain amount of protein must be furnished with the food if nitrogen equilibrium is to be maintained. It is clear that the metabolic processes resulting in the end products which tend to be constant in quantity appear to be indispensable for the continuation of life; or, to be more definite, those metabolic processes probably constitute an essential part of the activity which distinguishes living cells from dead ones. I would therefore call the protein metabolism which tends to be constant, tissue metabolism or endogenous metabolism, and the other, the vari- able protein metabolism, I would call the erogenous or intermediate metabolism. The endogenous metabolism sets a limit to the lowest level of “nitrogen equilibrium attainable. Just where that level is fixed will depend on how much, if any, urea is derived from the same katabolic processes that produce the kreatinin. If this can be determined, we shall have a formula expressing more or less definitely the point of lowest attainable protein katabolism, because at such a point the per- centage composition of the urine should be practically constant. The total nitrogen eliminated when this constant composition of the urine has been reached will indicate the lowest attainable level of nitrogen equilibrium. Whether or no such a level can actually be attained, or whether a certain amount of protein must not always fall prey to the exogenous metabolism, can only be settled by a great deal of experimental work. In connection with such work the significance and the importance of the exogenous protein metabolism might also be settled. Is any of it indispensable, and if so, how much? The analytical data recorded in the preceding paper make it clear that protein sufficient to main- tain the endogenous protein metabolism is indispensable. And the chemistry of the exogenous protein metabolism ought also to furnish evidence as to its importance to the organism. A fair amount of such chemical evidence is, I believe, already available, but before con- 1 These names are chosen because they have already been frequently used as indicating the origin of special metabolism products. 124 Otto Folin. sidering it, or rather in connection with it, I desire once more to refer briefly to the above mentioned theories of Pfliiger and Voit. The question that naturally arises is, do these two forms of protein- metabolism, 7.¢., exogenous and endogenous metabolism, represent the views held by Voit and Pfliiger? Does Pfliiger’s view cover the tissue, the endogenous metabolism, and does Voit’s view correctly represent the variable, the exogenous metabolism? The endogenous metabolism is of course more or less identical with the protein metabolism of Pfliiger, since it represents the decompo- sition of living protoplasm. Viewed more in detail it is, however, different, because it yields end products in proportions entirely differ- ent from those which Pfliiger knew. The exogenous protein metabolism has, I believe, nothing in com- mon with Voit’s metabolism of circulating protein. The difference between the theory of Voit and that of Pfliiger is not unimportant; but on a matter of far greater importance they agree, namely, on the universally accepted assumption that the entire katabolism of protein takes place in the same tissues and by means of katabolic processes similar to those which bring about the decom- position of the non-nitrogenous food, z. e., the fats and carbohydrates. In other words the protein katabolism is supposed to be essentially an oxidation, and the greatest amount of protein katabolism is sup- posed to take place where the greatest amount of oxidation occurs, z.€., in the muscles. F But the nitrogen of protein is invariably linked to carbon on the one hand and to hydrogen on the other, z.¢., it exists as amido or imido groups, and from a chemical point of view no oxidation is _| necessary for the splitting off of the nitrogen of such groups. Such splittings are more easily accomplished by hydrolytic reac- tions than by oxidations. For example, ina = CNH, or =C=NH group, the nitrogen can be split off as ammonia by the simple addi- tion of a molecule of water. There is, therefore, a przort no reason for assuming that the katabolism of the protein-nitrogen is brought about by the same chemical processes as those which decompose the fats and the carbohydrates, Have we any valid reason for believing that these two sets of chemical decompositions occur simultaneously or even in the same tissues? I do not think that we have. On the other hand, when this question is once fairly raised there is, it seems to me, considerable evidence against it. A Theory of Protein Metabolism. 125 The fact that urea, the chief representative of the exogenous pro- tein metabolism, is not to be found in the muscles (except in infini- tesimal traces) acquires a new significance from this point of view. The absence of urea in muscles where the greatest amount of general katabolism takes place is to this day an unexplained fact. Many attempts at explanation have been made. It has been thought that the protein decomposition in the muscles does not proceed as far as the urea, but only produces certain precursors of it which are after- wards elaborated in the liver into urea. But the only precursor which is found in quantities that could be considered at all adequate is kreatin, and it has been shown that the liver is not capable of con- verting kreatin into urea. Moreover the perfect constancy in the kreatinin elimination shown in these experiments under such different conditions excludes, it seems to me, positively the possibility that the kreatin produced in the protein katabolism of the muscles is after- kreatinin, and the presence of such considerable quantities of kreatin in the muscles, coupled with the absence of urea, as well as of any known precursor of urea, constitutes (in the light of the manifestly different laws governing their elimination) exceedingly strong evi- dence against the view that the nitrogen katabolism represented by urea takes place to any considerable extent in the muscles. If this is so, it explains why the experiments of Schondorff referred to above gave practically negative results. But if the nitrogen katabolism represented chiefly by urea does not occur in such important tissues as the muscles, then we are practically forced to the conclusion that this nitrogen katabolism is brought about by special chemical processes in special places, and that this katabolism is not of any such general fundamental importance as is the katabolism that yields kreatinin. Several other facts point in this direction, and I know of no facts contradicting it. We know that in the protein digestion ammonia is formed directly by the action of the proteolytic ferments, and we know that the liver is capable of convert- ing that ammonia into urea. That the ammonia formed during diges- tion is not recombined into protein compounds we know from the investigations of Nencki,! which showed that the blood coming from the intestines at the height of digestion contains two or three times as much ammonia as the blood on the other side of the liver. My _ '* NENcKI and ZALESKI: Zeitschrift fiir physiologische Chemie, 1901, xxxiil, m p» 200. wards converted into urea. The muscle kreatin is eliminated as | — 126 Otto Folin. own ammonia determinations published two years ago have confirmed this result. Here we have therefore positive evidence that a part of the food nitrogen is eliminated as urea without having ever passed through the muscles. The recent discovery by Cohnheim ? of erepsin in the mucous membrane of the intestines, points in the same direc- , tion. Cohnheim was looking for an enzyme which should be able to. \ recombine peptones or albumoses into albumin, but found instead an “Gears which splits them still further, z. ¢. into amido acids. The / work of Kutscher had previously showed that trypsin is capable of ' completely splitting protein products into relatively small amido acid molecules. The prevailing idea that these splitting products before entering the blood are recombined into albuminous bodies has'never been proved. The chief attempts made in this direction have not taken the deep seated decompositions of trypsin and erepsin into account; /and the failure to find peptones or proteoses has been accepted as showing that these products are at once in the mucous membrane converted into albumin and only as such enter into circulation. To be sure, Kutscher also failed to find the amido acids in the blood, and was himself therefore forced to assume, in accordance with earlier investigators, that they must be recombined into albuminous products before they get through the mucous membrane of the intestines.® Bergmann and Langstein* were also unable to show that protein digestion in dogs is always accompanied by an increase of non-albu- minous nitrogen in the blood, although their figures indicate that this is usually the case. It seems to me that the experimental conditions chosen by Bergmann and Langstein are in a measure inconsistent with the theoretical part of their paper. In their experiments they determine the non-coagulable nitrogen in the blood of fasting dogs and of dogs at the height of digestion in the hope of finding a demonstrable difference due to digested food protein But they use only ro c.c. or 15 c.c. of blood for the determination of this — nitrogen, yet they show by calculation that 30 gm. of nitrogen could be absorbed from the intestines of a man in the course of four hours with- 1 FoLin: Zeitschrift fiir physiologische Chemie, 1902, xxxvii, p. 174. * COHNHEIM: Zeitschrift fiir physiologische Chemie, 1901, xxxiii, p. 451. 8 KUTSCHER und SEEMANN: Zeitschrift fiir physiologische Chemie, 1902, xxxiv, Pp. 529: * BERGMANN and LANGSTEIN: Beitrage zur chemischen Physiologie und Path- ologie, 1904, vi, p. 27. A Theory of Protein Metabolism. 127 out the total increase of nitrogen in the blood amounting to more than 5 mgm. per too c.c. Their calculations are, however, equally applicable to dogs, and it is therefore not in any sense remarkable that they could not always demonstrate such an increase in 10 c.c. or 15 c.c. of blood. That they actually did find such an increase in some cases is, however, decidedly interesting. The failures that have attended the attempts to show that protein digestion is accompanied by an increase of the non-proteid nitrogen of the blood must as yet be ascribed to unfavorable conditions or to lack of sufficient accuracy in the technique. The determinations of the ammonia show that there must be such an increase. The chief reason why the nitrogenous splitting products produced by the digestive enzymes are universally assumed to be reconverted _into albumin is the teleological one. The food proteins are tissue builders and the organism must therefore not waste them. The fact that the muscle tissues of normal men do not increase when the protein of the food is increased, but that all of the nitrogen of such protein is at once eliminated, has not been sufficiently considered in this connec- | tion. The only adequate teleological explanation of this fact is that this nitrogen is not needed for the building of new tissues. It is not needed because the organism cannot enlarge indefinitely, and because after it has attained its full growth the daily waste of tissue is small. Yet when more nitrogen than the organism needs is fur- nished with the food, we find that the protein containing it is still absorbed up to the limit of the digestive capacity. What is the teleological explanation of this fact? In order to explain this phenomenon Pfliiger and Voit have ad- vanced the now generally accepted hypothesis that the organism uses protein by preference, even when an abundance of fats and carbohy- drates is available. But this hypothesis again is almost indissolubly connected with the assumption that it is because of its nitrogen that protein is preferred. Yet nobody, not even Voit, believes that a diet giving 25 gm. of urinary nitrogen is to be preferred for the average man to a diet giving 15 gm. or 16 gm. of such nitrogen. Why this arbitrary limitation? It is not indicated by the chemistry of protein metab- olism. When aman has established a nitrogen equilibrium, whether on IO gm. or on 20 gm. of nitrogen, an increase of nitrogen in the food is immediately followed by an increase in the elimination. At 20 gm. as at 10 gm. the organism seems therefore by preference to use any additional protein that is furnished with the food. 128 Otto Folin. There is abundant evidence that extensive hydrolytic decom- position of protein occurs in the digestive tract, and from the pres- ence of the enzymes, pepsin, trypsin, and erepsin, we must assume that the decompositions which they can accomplish are for the advan- tage of the organism. The usefulness of the proteolytic enzymes has always been admitted, but their usefulness, it has been thought, must be restricted to the mere solution of coagulated proteins. This re- striction is probably in a large measure due to the influence of Liebig, who held first that protein is the sole source of muscular work and, secondly, that the animal organism must obtain all its protein in pre- formed condition because it is unable to synthesize it out of simple crystallizable products. The first idea is now no longer believed to be correct, and the second is seriously questioned; but'the forma- tion of amido acids in protein digestion is still considered as so much waste of valuable material, because the organism is still supposed to need those products in the form of protein. It is clear that the teleological argument in favor of the immediate recombination of protein digestion molecules into albumin has no more value than the same argument earlier advanced against the view that the protein digestion could give rise to any considerable quan- tities of crystallizable amido acid products. Such an argument may be applicable to a part of the protein metabolism, but clearly cannot be true of the whole. The well established fact that the animal organism tends to maintain nitrogen equilibrium within such tre- mendously wide limits as those indicated on the one hand by “ forced” feeding, and on the other by the low nitrogen experiments, is utterly inconsistent with any such teleological argument in favor of albumin formation. Such an extensive fornaatiorrof Voit’s “ circulating pro- tein,” followed by a second immediate decomposition and elimina- tion as urea, seems to me almost if not quite as improbable as the corresponding formation and decomposition of Pfliiger’s organized protoplasm. According to the views here presented on the other hand, only a small amount of protein, namely, that necessary for the endogenous metabolism, is needed. The greater part of the protein furnished with standard diets like Voit’s, z. e. that part representing the exoge- nous metabolism, is not needed, or, to be more specific, its nitrogen is not needed. The organism has developed special facilities for get- ting rid of such excess of nitrogen so as to get the use of the car- bonaceous part of the protein containing it. The first step in this A Theory of Protein Metabolism. 129 “process is the’ ‘d omposition of protein in the digestiv@ tract into ° proteoses, amido acids, ammonia, and possibly urea! The hydro lytic decompositions are carried further in the mucous mémbrane of the intestines, and are completed in the liver, each splitting being such as to further the formation of urea. | In these special hydrolytic decompositions, the result of which jis to remove the unnecessary nitrogen, we have an explanation of yy and how the animal organism tends to maintain nitrogen equi- ibrium even when excessive amounts of protein are furnished with the food. ~ This excess of protein is not stored up in the organism, as such, because the actual need of nitrogen is so small that an excess is ; always furnished with the food, except, of course, in carefully planned periments. The ordinary food of the average man contains more rogen than the organism can use, and increasing the nitrogen still further will therefore necessarily only lead to an immediate increase in the elimination of urea, and does not increase the protein katab- ‘olism involved in the kreatinin formation any more than does an increased supply of fats and carbohydrates. The normal human organism can be made at almost any time to ‘store up fats and carbohydrates. The katabolism of these products jconsists chiefly of oxidation, a decomposition which sets free large § itities of heat which can be converted into mechanical energy use- red the organism. The hydrolytic removal of nitrogen from the tein involves by comparison a very small transformation of energy 1 yields a non-nitrogenous rest of great fuel value. This non- rogenous rest derived from protein may partly be directly trans- Te ; ed to the different tissues, and thus at once supply oxidative terial where needed, but in all probability is partly converted into 3, or at least into carbohydrates, and then becomes subject to the ws governing the katabolism of these two groups of food products. [ fully believe that the view here developed concerning protein abolism is substantially correct. More detailed studies may neces- ite minor modifications and additions. For example, it may well that the digestive tract and the liver are not alone sufficient to t off all the unnecessary nitrogen. Other glands may play a part, e know that proteolytic enzymes resembling trypsin are present ch glands. And it may be that a detailed study of the urines of 1ivorous animals, as the cat and the dog, may indicate a somewhat -Kosset and Dakin: Zeitschrift fiir physiologische Chemie, 1904, xlii, 130 Otto Folin. different condition. We know in fact that these animals can store — up great quantities of protein (in the form of increased muscle sub- stance?). Their endogenous metabolism may therefore be much lems) ; constant than that of man. The urines of such animals contain nitrog- | enous products not found in human urine. Their protein katabe olism is therefore qualitatively as well as quantitatively more or less | different. Similarly with reference to herbivorous animals. They also yield nitrogenous waste products quantitatively quite differently distributed than are those found in man, and detailed studies of those products under different conditions will undoubtedly throw additional! light on the whole subject here under discussion. But I do not think that such studies will materially affect my main contention that pro- tein katabolism in the animal body is of two kinds, that one of these” only is true tissue katabolism, and that the other consists of a pre-~ liminary removal of unnecessary nitrogen by means of hydrolytic. splitting processes which are distinct and separate from the decom-— \ positions associated with subsequent oxidations. The exogenous or intermediate protein katabolism is here conceived as consisting of a series of hydrolytic splittings resulting in a rapid elimination of the protein-nitrogen as urea. This view does not, however, preclude the possibility that a certain amount of oxida is associated with this form of katabolism. In fact, we must assume ~ that this is to some extent the case. The hydrolytic removal of the| — nitrogen would leave hydroxyl groups in the positions previously occu- pied by the amido groups. But since it is almost absolutely certain: that the non-nitrogenous rest is in part converted into carbohydrates, — a certain amount of oxidation and of so-called aldol condensation — must constitute an important part of the chemical transformations necessary to the formation of glucose.!_ The splitting off of the pro- tein-nitrogen as ammonia, and its combination with carbon dioxide, and subsequent conversion into urea, demands no oxidation. But just as oxidation is necessary for the formation of carbohydrates out — of the non-nitrogenous rest, so a certain amount of oxidation is — necessary to bring about the conversion of the protein-sulphur into sulphuric acid. . ae a es eee Se ee eres 1 A full discussion of this question cannot be attempted here. That the animal organism can form glucose out of protein derivatives containing no preformed ear- bohydrate groups is wellnigh positively demonstrated. See LEO LANGSTEIN’S in- teresting article on this subject in Ergebnisse der Physiologie, iii, part 1, p. 453 (1904). See also Lituyr: Archiv fiir die gesammte Physiologie, 1904, cvi, p. 160. . A Theory of Protein Metabolism. 121 DEDUCTIONS CONCERNING SOME MoRE SPECIAL PROBLEMS IN METABOLISM. The theory developed on the preceding pages concerning protein metabolism would cause a number of important problems to appear in a new light. Nitrogen equilibrium. — We have already seen how the theory ex- plains the persistent tendency on the part of the organism to main- tain nitrogen equilibrium, even when this involves the formation of excessive quantities of urea. A rather interesting and instructive illustration of this phenomenon is shown in the third periods of the feeding experiments represented by Tables II-V of the preceding paper. The period represents a return to a nitrogen-rich diet after the subjects had been losing considerable nitrogen during a week or more when on a diet containing less than 1 gm. of nitrogen. On the very first day that nitrogen was obtained with the food, the elimination of this element was more than doubled, and on the sec- ond day more than three times as much nitrogen was eliminated as on the last day of nitrogen starvation. Sivén observed the same phenomenon, and found that after having lost 32.4 gm. of nitrogen on a protein-poor diet, he regained only 20.6 gm. in the course of thirteen days on a diet very rich in nitrogen. Sivén interprets this as speaking against Pfliiger’s and in favor of Voit’s theory of pro- tein katabolism.! He explains the phenomenon by assuming that the loss of nitrogen represents loss of muscle substance, and that the subsequent rebuilding of the lost muscle tissue is a slow pro- cess. But the detailed analytical work carried out in connection with my feeding experiments clearly indicates that the loss of nitro- gen does not represent loss of muscle tissue, and Sivén’s uric acid determinations certainly point in the same direction. I would offer the following explanation: All the living protoplasm in the animal organism is suspended in a fluid very rich in protein, and on account of the habitual use of more nitrogenous food than the tissues can use as protein the organism is ordinarily in possession of approximately the maximum amount of reserved protein in solution that it can ad- vantageously retain. When the supply of food protein is stopped, the excess of reserve protein inside the organism is still sufficient to 1 SIVEN: Skandinavisches Archiv fiir Physiologie, 1901, xi, p. 320. 139 Otto Folin, cause a rather large destruction of protein during the first day or two of’ protein ‘starvation, and after th&t the protein katabolism is very small, provided sufficient non-nitrogenous food is available. But even then, and for many days thereafter, the protoplasm of the tissues has still an abundant supply of dissolved? protein, and the normal ac- tivity of such tissues as the muscles is not at all impaired or dimin- ished. When 30 gm. or 40 gm. of nitrogen have been lost by an average-sized man during a week or more of abstinence from nitrog- enous food the living muscle tissues are still well supplied with all the protein that they can use. That this is so, is indicated on the one hand by the unchanged kreatinin elimination, and on the other by the fact that one experiences no feeling of unusual fatigue or of in- ability to do one’s customary work. Because the organism at the end of such an experiment still has an abundance of available pro- tein in the nutritive fluids, it is at once seemingly wasteful with nitro- gen when a return is made to nitrogenous food. This is why it only gradually, and only under the prolonged pressure of an excessive supply of food-protein again acquires its original maximum store of this reserve material. Standard diets. — If the interpretation just given for the phenome- non of nitrogen equilibrium is correct, it constitutes at the same time a definite reason why the so-called standard diets are unnecessarily rich in protein. Nitrogen enough to provide liberally for the en- dogenous metabolism and for the maintenance of a sufficient supply of reserve protein is shown to be necessary. But it ought neither to be necessary nor advantageous for the organism to split off and remove large quantities of nitrogen which it can neither use nor store up as reserve material. In the case of carnivorous animals, the uncertainty of the food supply has evidently led to the development of the ca- pacity to store a certain amount of protein in the form of increased muscle substance, but in man this capacity seems not to exist. The slowness with which the normal human organism stores nitrogen after having lost only very moderate amounts, does not mean that the human organism can replace lost muscle tissue only slowly and with _ difficulty. When the organism really has suffered a loss of such tissue, as, for example, during typhoid fever, we know that during convalescence there is an astonishingly rapid recovery of weight 1 Whether such solution is colloidal or otherwise, and whether such dissolved reserve protein is stored in the blood and lymph, or also, as seems to me highly probable, in the cellular fluids, is immaterial in this connection. . | | . | | | | A Theory of Protein Metabolism. ne 3 a and a correspondingly extensive retention of nitrogen.! The loss of 25 gm. or 30 gm. of nitrogen, due to the withdrawal of nitrogenous food, does not involve loss of muscle substance, Ist, because the endogenous metabolism is not affected; 2d, because it is not ac- companied by loss of strength and ability to do work ; 3d, because from a teleological standpoint it seems highly improbable that the consumption of nitrogenous material in a system consisting of living protoplasm immersed in a highly nitrogenous solution should be at the expense of the protoplasm rather than at the expense of the solu- tion. The chief reservoir for nitrogenous reserve material in the human organism seems therefore to be the unorganized protein in the fluid media rather than the protoplasm of the muscles. This is why there is normally in man a comparatively sharp limit to his capacity for storing protein, and why all excess of nitrogen given with the food is promptly eliminated. The immediate elimination of the greater part of the nitrogen con- tained in 118 gm. of protein by means of the exogenous katabolism would seem to constitute very strong evidence in favor of the view that the protein so katabolized can without harm, if not with advantage, be replaced by an equivalent quantity of carbohydrates. The above interpretation of the phenomenon of nitrogen equilibrium is based on the generally accepted view that the human organism ordinarily can- not be made to increase its store of nitrogen. Recent experiments by Liithje ? and others * have, however, shown that this view is not strictly correct. By means of short feeding experiments with extraordinarily large quantities of protein (40 gm. to 60 gm. of nitrogen or more per day) or with diets containing enormous fuel value (7o calories per kilo of body weight) the human organism can evidently be made to retain for a time considerable extra quantities of protein. ‘These experiments are highly interesting, but they seem to me to have less bearing on the ques- tion of nitrogen equilibrium than may at first sight appear to be the case. There must be a limit to the amount of food, particularly of protein, which the organism can dispose of by means of normal physiological processes. ‘The peculiar feature of such forced feeding experiments is the introduction of exormous quantities of protein for a short time (40-60 1 LUTHJE and BERGER: Deutsches Archiv fiir klinische Medizin, 1904, 1xxi, mp. 278. SeLUTHIE: Loc. ctt. 8 See KAUFMANN: Zeitschrift fiir diatetische und physikalische Therapie, _ 1903-4, vii, p. 355. See also ALBU: Berliner klinische Wochenschrift, '904, _xivii, p. 1228. SL * . 134 Otto Folin. gm. of nitrogen means from eight to fifteen times as much as is actually needed for the maintenance of nitrogen equilibrium). We have but to assume that the digestive tract is able to handle such quantities for a few days, but that the subsequent physiological processes involved in the katabolism are more limited, and we are forced to conclude that the organism is then compelled to dispose of a part of such nitrogen in a manner more or less abnormal. The fact that a part of such excess of nitrogen is retained for a time at least, is interesting, but this may indicate only that the organism is less damaged by retaining it in one form or another than by eliminating it in pathological forms through the kidneys, as would be the case if it is flooded with protein derivatives which it cannot use. Detailed study of the nitrogen that is eliminated under such extraordi- nary conditions might throw some light on the processes involved. If 50 gm. of urinary nitrogen is eliminated in twenty-four hours from a diet containing no meat, this nitrogen should contain not less than 96 per cent of the total as urea and ammonia in case the physiological katabolic processes involved are adequate. If they are not adequate, if the excess nitrogen is stored because it cannot be normally eliminated, then we should probably (though not necessarily) find that less than 96 per cent is present as urea and ammonia, and that the deficit is accounted for, not by an increase of kreatinin and uric acid, but by the presence of other nitrogenous constituents not ordinarily present in human urine. In other words the inadequacy of the physiological katabolic processes may result in the escape of some nitrogen without the preliminary conversion into urea. The two chief arguments which have been advanced in justification of large amounts of protein (118-130 gm.) are first that people who can afford it actually do frequently consume such large quantities of protein, and secondly that Voit and his co-workers thought they had shown that nitrogen equilibrium cannot be permanently maintained on smaller amounts. The first argument has, however, no more value than the same argument would have if applied to the question of the daily use of wine at the table. Like most statistical arguments this one is capable of quite different interpretations. The second argument has carried incomparably greater weight. It has repeatedly been shown that nitrogen equilibrium can be maintained on less than one-third the nitrogen demanded by the standard diets; but Voit came long ago to a different conclusion, and the experiments of the more recent inves- tigators have not been able to awaken any general doubt as to the A Theory of Protein Metabolism. 135 substantial correctness of the conclusions of Voit. Even the men who have successfully made low nitrogen equilibrium experiments have hesitated to draw conclusions essentially different from those of the older master. ‘‘ Doch ware es sicherlich ibereilt,’ says Sivén, “von einem Ueberschusse zu sprechen bevor man den Zweck kennt, wozu der Organismus das Ejiweiss verwendet, welches es taglich verzehrt. Was wissen wir dariiber’”’? “Wir wissen, dass das Eiweiss in der Nahrung fiir die Muskelar- beit nicht notwendig ist; wir wissen dass ein gewisses Minimum nicht unterschritten werden darf, ohne dass der Organismus darunter leidet, aber wir wissen nicht weshalb dem so ist; wir wissen nicht, welchen vitalen Processen das Eiweiss im Allgemeinen, und speciell diese verhaltnissmassig geringe Eiweissmenge dient.’ ! The appearance of Chittenden’s book a few weeks ago, on Physio- logical Economy in Nutrition, seems, however, to me to indicate the beginning of the end of Voit’s nitrogen-rich standards of diet. Here we find feeding experiments numerous enough and prolonged enough to constitute incontrovertible evidence against the necessity of con- forming to the standards of Voit. And Chittenden moderately but firmly takes the position indicated by the experiments that the cur- rent dietary standards can safely be divided at least by two as far as their protein contents are concerned. The conclusions here derived from detailed studies of the composi- tion of urine corresponding to different amounts of protein katabolism are therefore in harmony with recent low nitrogen experiments made purely from the nitrogen equilibrium standpoint. Theoretically, as well as practically, 118 gm. of protein have therefore been shown to be far in excess of the needs of normal man. Can anything more be said? Can it be shown that such large quantities of protein are not only unnecessary but in addition detri- mental? This is undoubtedly a much more difficult question to answer. It has frequently been argued that the excessive produc- tion of uric acid, xanthin bases, kreatin, etc., in short, of the so-called extractives, is responsible for the gouty tendencies so common among people known to consume large quantities of protein. This may be true. There may be, probably is, some genetic relation between the production of uric acid and of gout, but this cannot be admitted as showing or indicating that the use of excessive quantities of protein is responsible for it. Such arguments can evidently be applied only 1 SIVEN? Loc. cit., p. 332: 136 Otto Folin. to meat eating, because the analytical data recorded in the preceding paper show that the excess of protein consumed is converted into harmless urea. The protein does not lead to an excessive production of uric acid, and does not at all increase the kreatinin. Crude meat, on the other hand, contains considerable amounts of such extractives or substances out of which they can be formed, and since meat con- stitutes the chief source of concentrated protein food, the argument may have a certain practical value. But from a purely scientific point of view it has no bearing on the question raised at the beginning of this paragraph. Yet it would seem that since 118 gm. or 130 gm. of protein are not necessary, the daily use of such large amounts may sooner or later lead to metabolism disorders, and thus be detrimental. The prevailing views of the phenomenon of nitrogen equilibrium are based on the supposition that the organism uses protein by prefer- ence, instead of fats and carbohydrates, and that protein is therefore the kind of food most suitable to the needs of the organism. This view, if pushed to its logical conclusion, is equivalent to saying that the more protein the food contains, the better it is; in other words, there can be no such thing as “ luxus-consumption ” of protein. But Pfliiger is, as fav as I know, the only one who in recent times has held this extreme position. To my mind this interpretation of the protein consumption is simply a more exact statement of the popular idea that large quantities of meat increase the strength and endurance of healthy men. In the light of the theory developed in this paper concerning the double nature of protein metabolism and the explanation of the phenomenon of nitrogen equilibrium, the following objection can perhaps be made to the use of large quantities of protein. The excess of nitrogen furnished with the food is normally quickly converted into urea and eliminated, and is, therefore, normally harm- less. The continuous excessive use of protein may lead, however, to an accumulation of a larger amount of reserve protein than the organ- ism can with advantage retain in its fluid media. But it is entirely possible that the continuous maintenance of such an unnecessarily large supply of unorganized reserve material may sooner or later weaken one or another or all of the living tissues. At any rate it seems scarcely conceivable that the human organism, having all the time access to food, can gain in efficiency on account of such an excess of stored protein. The carrying of excessive quantities of fat is considered as an impediment, the carrying of excessive quantities : i 5 i A Theory of Protein Metabolism. 137 of unorganized protein may be none the less so because more common and less strikingly apparent. ; Diets in diseases. — The objections that can be raised against the use of too much protein are rather meagre and indefinite with refer- ence to perfectly normal persons, but in the case of the sick they become clear and definite. In diseases of the liver and the kidneys, in diseases affecting the formation or elimination of urea, important practical conclusions must be drawn on the basis of our understanding of the function of protein as a food constituent. If the conclusions arrived at in this paper are correct, it is clear that the exogenous protein katabolism should in such cases be reduced to the lowest practicable level. This is not accomplished by substituting a milk diet for a meat diet, for such substitution affects only the purin bodies, only the extractives. If low nitrogen experiments have any practical value, they should be tested in such cases; in other words, diets having sufficient fuel value, but containing only 3 gm. or 4 gm. of nitrogen or less should be tried. And, in special cases, as in typhoid fever or where the onset of urzemia is threatened, a diet like that used in Ex- periments II to V, containing practically no nitrogen, and leaving practically no undigested residue in the intestinal tract, ought to be of value. The store of nitrogen is abundant, and the use of it is lessened if non-nitrogenous food is available. Soluble starch or dex- trin solution ought, it seems to me, to be more suitable than solid food, milk, or even starvation, in such cases. The use of non-nitrog- enous food would seem to me to be contra-indicated only in cases where the urea elimination is enormously increased and out of pro- portion to the caloric needs of the body. In such cases there has probably been an extensive destruction of living protoplasm which has led to such an increase of the dissolved protein that the organism cannot retain it. While this condition lasts there may be no need of supplying non-nitrogenous food. Weakened kidneys and the elimina- tion of albumin may perhaps, in some cases, be directly the result of an abnormal production of dissolved protein rather than the result of toxins. \ The effect of work on protein metabolism. — The effect of physical work on protein metabolism, or rather the absence of any such demonstrable effect, becomes, it seems to me, for the first time in- telligible in the light of the theory concerning protein metabolism advanced in this paper. The fact that protein metabolism is independent of muscular work 138 Otto Folin. seemed at first so contrary to Liebig’s theory of the source of muscu- lar energy, a theory which he defended as long as he lived, and seems still so contrary to the deep rooted general idea that meat is the most efficient kind of food for all who labor physically, that it was accepted only after having been repeatedly demonstrated by a number of dif- ferent investigators. Even yet this fact has not become common knowledge. The rank and file of the people, especially the people of England and the United States, still hold the view that an abundance of meat is essential to a good day’s work. The fact that moderate, or even severe muscular work does not in- crease the katabolism of protein is indeed highly remarkable, if the prevailing views concerning the nature of that katabolism are correct. If the katabolism of protein consisted essentially of oxidations, and if these oxidations occurred most extensively in tissues where the great- est amount of energy is developed, 7. ¢., in the muscles, then it would seem almost incomprehensible that the tremendous zzcrease of oxida- tion, and of energy production which accompanies muscular work, should not include an increase in the oxidation of protein: If the organism oxidizes food protein for the purpose of producing heat and energy just as it oxidizes fats and carbohydrates for that purpose, and if it in addition uses protein by preference, we should expect that the increased energy production of physical labor should increase the ka- tabolism of protein even more than that of the non-nitrogenous food. According to the theory developed in this paper, on the other hand, the protein katabolism, in so far as its nitrogen is concerned, is inde- pendent of the oxidations that give rise to heat or to the energy that is converted into work. The hydrolytic splitting off of ammonia, and its elimination as urea, is independent of oxidations except, of course, to the mere extent of using a relatively small amount of the carbon- dioxide that is produced by means of oxidation. According to this theory, it would therefore appear clear why work does not appre- ciably affect the katabolism, or, to be more specific, the exogenous katabolism of protein. Moderately severe physical labor may, however, have more or less of an effect on the endogenous metabolism. Whether or no this is © the case cannot be shown by means of urea, or even by total nitro- gen.determinations, because the effect of work on these constituents is too slight ; determinations of kreatinin, uric acid, and neutral sul- phur are necessary for the study of this question. f ; i MEASUREMENT OF ELECTRICAL CONDUCTIVITY FOR CriInNicCAlt PURPOSES. By T. M. WILSON. [From the Hull Physiological Laboratory, University of Chicago.) CONSIDERABLE amount of work by Stewart,’ Tangl and Bugarsky,? Roth,® Rollett,* Hamburger,’ and Viola,® has been done on the electrical conductivity of the blood and serum of lower animals, a physical quantity which gives us a measure of the con- centration of ions in serum. The number of observations on human serum is very small, owing to the difficulty of obtaining sufficient quantities. In order to overcome this obstacle, I have made a small apparatus which requires only four or five drops of blood for each experiment, and is thus adapted for clinical purposes. It also may be of use in obtaining physiological and pathological data in animals when only small quantities of blood or other liquid are available. When the conductivity of the blood is measured, in addition to that of the serum, an observation requiring only a few minutes more, the relative volume of corpuscles and serum can be at the same time deduced either from Stewart’s formula (2) ‘p= 2 (174.5 — Ap) ry = 5 or the more complex one ~= a (180 — A,— VA,) formula (1),” s where 7 is the percentage volume of serum, A, and d,, the conductivity of the blood and serum respectively, expressed in reciprocal ohms at aoe. X TO%. Of course the hzmatocrite gives us an easy, rapid, and when un- diluted blood is used, but not otherwise, a fairly accurate means of determining the relative volume. The additional information derived 1 STEWART: Journal of physiology, 1899, xxiv, p. 356. 2 TaNGL and BuGarsky: Centralblatt fiir Physiologie, 1897, p- 332. Bore 2 /i2d., p. 271. 4 RoLLeTT: Archiv fiir die gesammte Physiologie, 1900, Ixxxii, p. (99. 5 HAMBURGER: Osmotische Druck und Ionenlehre, 1902, i, p. 474. VIOLA: Revista veneta di scienze mediche, xviii, Fasc. viii, 1901, cz. HAM- BURGER, p. 486. 6 139 140 T. M. Wilson. from the conductivity readings as to the approximate number of ions present in unit volume of the serum, may in the future be of equal or even greater clinical importance in certain diseased conditions. COMPARISON OF ELECTRICAL AND HZEMATOCRITE METHODS OF DETERMINING THE RELATIVE VOLUME OF CORPUSCLES AND SERUM. The electrical method of measuring the relative volume of corpuscles and serum was compared by Dr. Stewart with Hoppe-Seyler’s 1 chemi- cal method and also with a colorimetric method of his own. A satis- factory degree of agreement was found for dog’s blood, for which the formule were originally constructed. Recently P. Fraenkel? has compared the method with Bleibtreu’s ? chemical method) with a like result. He recommends it for scientific purposes, but his statement that one determination can be carried out in an hour with 20 c.c. of blood and éven with 12 c.c.to 15 c.c., when the blood is rich in serum, is far too unfavorable to the method, as will appear from my own results. Such quantities, of course, cannot in general be obtained in clinical examinations. So far as I am aware, the electrical method has not hitherto been systematically compared with determinations by the hzmatocrite. Hamburger,‘ indeed, gives a table comparing the percentage-volume of corpuscles as determined by centrifugalization of a few specimens of blood with that deduced by the electrical method. But he appears to have used an ordinary centrifuge and there is considerable variation in the degree of agreement of the methods in different experiments. It was, therefore, thought worth while.to begin by instituting a com- parison for dog’s blood between the haematocrite readings and the electrical determinations. About 500 c.c. of fresh defibrinated blood from a dog was centrifugalized and the serum removed. ‘The percentage of serum still left in the sedi- ment as determined by the hamatocrite was 25.5. In all the observations the hzmatocrite was turned for five minutes at an average rate of 217 times a second. The serum separated from the blood was recen- trifugalized until it was entirely free from erythrocytes. Fourteen mixtures 1 Hoppe-SEYLER: Handbuch der physiologisch- und pathologisch-chemischen Analyse, 1893, p. 441. 2 FRAENKEL: Zeitschrift fiir klinische Medizin, lii, p. 476. 8 BLEIBTREU: Archiv fiir die gesammte Physiologie, li, liv, lv, and Ix. 4 HAMBURGER: Loc. cit., p. 528. Measurement of Electrical Conductivity. 14! of the sediment, with various quantities of the clear serum, were made as shown in Table I, the amount of serum added to the sediment being 5 per cent of the first mixture (Experiment 2), ro per cent of the second mixture (Experiment 3), and so on, increasing by 5 per cent in each successive mixture. ‘The resistance measurements were made at about o C., the tube being immersed in melting ice. A thermometer with its bulb close to the tube indicated a temperature which only varied one or two-tenths of a degree from 0° C. ‘The resistances were reduced to 5° C. by multiplying by a constant. ‘The resistance capacity of the tube was determined by filling it with 5 per cent potassium chloride solution, and measuring its resistance at 18° C., for which temperature the specific conductivity of potassium chloride solution is given in Kohl- rausch’s tables. From these data it.is easy to calculate the conductivities of the various blood-mixtures expressed in reciprocal ohms multiplied by ‘ro® at 5° C., by multiplying the reciprocals of the corresponding resistances by a constant 208,607, which is the same for all measurements made with one and the same tube. If a large series of observations is to be made, it saves time to construct once for all a table giving conductivities directly, as recommended by Dr. Stewart, and then the working out of the result from the formula entails but little labor. ‘The small apparatus was not used for these comparative experiments, but, instead, a U-tube with a capacity of 7 c.c. A constant quantity, 5 c.c., of the mixture®to be tested, was always introduced into the tube. It will be seen that for mixtures with 67 to 74 per cent of serum, as determined by the hzmatocrite, there isea very close agreement between the results of the hamatocrite and the electrical method. With mixtures richer in serum, the electrical method gives results somewhat higher, and with mixtures poorer in serum, results some- what lower than the hematocrite. At both extremes the numbers deduced from formula (1) are nearer to the hematocrite readings than those deduced from formula (2). But in the middle range, which would correspond to the majority of specimens of dog’s blood, formula (2) gives a somewhat closer approximation to the hematocrite deter- minations. As in Dr. Stewart’s! tables when the percentage of serum is low, formula (2) gives rather smaller values than formula (1), but somewhat larger values when the percentage of serum is high. Some serum is, of course, present in the interstices of the sediment in the hzematocrite tube even after prolonged centrifugali- zation. No attempt was made to determine its amount. But the 1 STEWART: Loc. cit. Wilson. Tt ee: 142 ‘47Q SVM UINIAS 94} JO APAVONpUOD OY, 06'6L 88 OLL OS OL OSL 6EL 00 eL “tL COL 02'S9 oL9 o99 OF'99 s'+9 L709 Ole9 O19 06S OLS OLS cuss OLS a Salis 00'es FOS S'Lh 09'6b 69h tr Of 9b ech b Ob 00'¢h S'6E 9195 09'6£ e9E 6CE Of 9F Lee C66 S67 260 S'S? *yuao rad or "yua9 aod ¢ ‘OUON WUnAIs JO SUIL}UOD JUAWIPas 9OOJeUAY yeyy uodunsse uo payepnoyvo ‘sornjXxIl oy} UL WiNndas JO 95vIU90.10,] 00 28 09'S 09°94 00 £4 O¢ OL OL 99 0049 0609 OFSS 09 OF 09'9¢ $6 cf OLE BS 7 62 008 OL O'rL O'0L 0'L9 O'e9 0'09 995 $'0s rere 0 Cb SSE OE 9'6¢ GSC ‘yuo aod QT ST jE }vy) ou -snd109 at ruinsse ‘sao |} UADMyOq [[GS andtos 10F poe99t1O9 SSUIPva 9} LIDOPOWML YY *yuao 10d ¢ *941190) -viuey Aq SSUIPVaI winds Is £8h orl SOL L'99 elo 6LS oes PLP oCb OL 0 +2 (Z) BIN ULO NT 912 e+? ‘(T) BINWIO YT ‘poyjyoul eo1y99[9 Aq wndes Jo asv}UI010q I dlavL 0s'9S os'¢ ores 0s'9 09'S¢ 00°€ 06'tb eee OUlt s4 09 9¢ 00'¢ 06'¢€ 00'2 oe 0g SLT sey4 OST 00'F@ OST 0812 00'T OL61 001 Og 91 0s'0 SO +1 0s°0 Og ZI 000 ORIN ‘O°O Ul sonpuday | Baeee Wns OST Os’ 00° 00'e Os é eit O0'e 00'S OPONUT Udy} yuowt “P9S Ww Measurement of Electrical Conductivity. 143 magnitude of the errors which may possibly arise, neglecting this quantity, ate displayed in the table, where the percentages of serum actually present in the mixtures are calculated on the assump- tions: (a) that no serum was still present between the corpuscles in the hematocrite sediment, (b) that 5 per cent of the hamatocrite sediment consisted of serum, and (c) that ro per cent of it was serum. The actual hematocrite readings are given, and also readings corrected in accordance with assumptions (a), (b), and (c). The computed percentages are uniformly lower than the corresponding hzematocrite readings, which seems to indicate that there may be some truth in the criticism directed against the hzematocrite method that, even when undiluted blood is employed, liquid may be squeezed out of the corpuscles during centrifugalization. The common method of diluting blood with artificial solutions before centrifugalization is pretty generally recognized as inadmissible when any great degree of accuracy is required, since, in general, the diluting solution will not be exactly isotonic with any particular specimen of blood. The calculated percentages agree considerably better with the electrical determinations than with the corresponding hamatocrite readings for mixtures relatively rich in serum, and for those contain- ing such a proportion of serum as is generally found in normal blood. For mixtures relatively poor in serum the computed percentages diverge considerably from the actual hzmatocrite readings, though much less from the corrected readings. The electrical results for mixtures poor in serum agree fairly with the percentages calculated on the assumption that the serum in the interstices of the haematocrite sediment is negligible in amount, but diverge rather widely for per- centages calculated on assumptions (b) and (c). DESCRIPTION AND TESTING OF APPARATUS FOR MEASURING ConpucrTivity oF Human BLoop AND SERUM. The small apparatus for human blood is shown in Figure I. Its length is 50 mm. so as to be adapted to the ordinary hematocrite frame. It is 6 mm. in outer diameter and 3 mm. bore. The ends are ground on an emery wheel and made to fit into the sockets of the frame. More than half of the length of the tube is plugged with a glass rod and sealed tightly with sealing wax. The capacity of the part remaining, that is from C to A, is 0.1322 c.c., as determined by weighing the tube filled with dis- 144 IT. M. Wilson. tilled water and then weighing it empty. The platinum electrodes are fixed in a glass piece which can be inserted in the liquid filling the tube. In Fig. 2, G is the glass electrode head surrounding the platinum wires, W, which are about 0.5 mm. in diameter. ‘These wires are bare at P and G cee ee Tt FIGURE 1.— 7 = tube; A =open end of tube; B =closed end; C= upper end of glass plug. iV. The whole figure represents a section of the electrodes. The part from /to P is sufficiently small to enter the bore of the tube'T (Fig. 1) without coming in contact with the sides. JV and P in Fig. 2 are the only parts of the electrodes which are not insulated with glass. FicurE 2.— W=wires; G= glass electrode head; #= flange of glass; VW = upper electrode ; P = lower electrode. Figure 3 is a section of the complete apparatus. C is an aluminium cap which is firmly sealed to the electrode head 4 The cap is so bev- elled internally as to resemble two truncated cones with apices adjoining. The tube 7’can be thrust into one of these conical spaces as is seen in the figure. When thrust home, the tube 7’fits into the cap in a definite position such that the insulated portion of the electrode head which enters fT GOVE Th FIGURE 3.— H = glass electrode head; C = aluminium cap; V = upper electrode ; FP = lower electrode; 7'= tube. the bore does not touch the sides of the tube, so that when the tube is filled with blood or other fluid, a layer of it intervenes between the elec- trodes and the tube. When in position, electrode VV is 2 mm., and electrode P 10.5 mm. from the open end 4 (Fig. 1). Measurement of Electrical Conductivity. 145 Figure 4 shows the apparatus in a vertical position. The tube T fits so tightly into the cap C that it is held suspended. Its constancy of posi- tion is assured by making the marks AZ and JZ’ on the cap and tube coincide. The whole apparatus, after the tube is filled with blood, is suspended in a narrow metal tube which is surrounded with pounded ice. In the bottom of the aluminium tube was placed some mercury to aid in more rapid cooling. The apparatus was pushed so far down into the metal tube that the top of it was 4 cm. below the surface of the ice. The electrodes were covered with platinum black in the usual way, by means of a galvanic current passed through a 2 per cent solution of platinic chloride. A medicine-dropper drawn to a fine point was used to fill the tube 7, and afterwards to wash its bore and head with water. A pair of bellows was employed to thoroughly dry the whole apparatus after each experiment. The first test made with the apparatus was to compare its readings with those obtained with the U tube employed in the experiments shown in Table I. ‘ Two samples of dog’s blood A and B were used for these tests. The U tube and the small apparatus were filled at the same time and placed in position in the same vessel con- taining pounded ice, and readings were made after six minutes’ immersion. Table II shows |. 4g 9a mae the values obtained. For sample A, the ratio electrode head; C = alumi. of the average of all the readings with the U_ niumcap; 4/’/=adjusting tube to the average of all the readings with lines: 7= tube; @ = glass the small apparatus was 0.5108, and for sample sh B, the corresponding ratio was 0.502, so that the agreement is satisfactory. As it is convenient to defibrinate the blood before filling the small apparatus, especially for clinical work, the question arises whether defibrination causes any difference in the conductivity of the serum. G. N. Stewart! found that the conductivity of the serum of clotted blood is approximately the same as that of the serum of defibrinated blood from the same animal. Oker-Blom? states that the serum 1 STEWART: Centralblatt fiir Physiologie, 1897, p. 332. 2 OKER-BLOM: Osmotischer Druck. HAMBURGER: 1902, p. 477. 146 T. M. Wilson. separated from the blood clot has a greater conductivity than the serum obtained from defibrinated blood; Hamburger! asserts that the TABLE Il: BLoop A. Buioop B. Small apparatus. U tube. | Small apparatus. 8,000 4,090 7,900 4,090 8,000 4,090 8,000 4,090 8,100 4,080 8,000 4,090 8,000 4,080 Average 8,000 4,087 Average 7,973 freezing point of the serum of coagulated blood is lower than that of the serum from defibrinated blood. I made the following experiments in the hope of clearing the matter up. A dog was anzesthetised by ether. Blood was removed by inserting a cannula in the carotid, and one portion was allowed to clot and the other defi- brinated. Both samples were placed in the refrigerator for six hours. The serum of each was then obtained. Both sera were slightly tinged with blood-pigment, but the tint in the serum from the defibrinated blood was a little deeper. The coloring matter proved, on standing, to be still in the corpuscles in both sera. The deeper tint in the defibrinated serum was due to it having been left for a shorter period in the centrifuge. A colorimetric comparison showed that the defibrinated serum had o.1 per cent of corpuscles and the serum from the clot 0.095 per cent. As can easily be calculated, the difference in conductivity which could arise from this difference in the content of corpuscle is quite negligible, since 1 per cent of corpuscles would only alter the conductivity about 1 by 108 in the unit in which conductivity is expressed in the paper. 1 HAMBURGER: J6id., p. 477. Measurement of Electrical Conductivity. 147 Four samples of each serum were taken and the conductivity deter- mined by means of the Utube. Freezing-point readings were also made. The results are given in Table III. They indicate that the TABLE III. CLoT SERUM. DEFIBRINATED SERUM. ae LOE: . Novel Tes 87.2 0 668 87.2 0.660 87.2 0.665 87.6 0.670 876 0.660 87.2 0 665 87.2 0.660 87.2 0.660 conductivity and freezing-point of the serum are the same whether it is derived from the clot or from the defibrinated blood. In close connection with the above may be mentioned an experi- ment with the small apparatus, on blood obtained from a healthy student, to ascertain if there was any difference in the conductivity of TABIGE, LV- Ratio of Constant cor. Conduc- Percentage | Percentage resistance of for 0° C. giving tivity of serum by | by hamato- serum and blood. | values in 5°C. | of blood. | formula (2). crite. 3860 : 8360 395890 47.4 S¥she 58.0 3870 : 8300 47.6 Shey 57.4 3860 : 8400 : 58. 58.3 3870 : 8830 ; § 57.9 3840 : 8700 , 56. 58.0 3840 : 8500 : Wie 58.0 3840 : 8570 ; 58.0 Average S00C ; : 57.9 defibrinated and non-defibrinated blood, and between the conductivity of plasma and serum from defibrinated blood. The conductivity for 148 T. M. Walson. unclotted blood was 37.5; for defibrinated blood, 37.7; for serum from defibrinated blood, 91.2; for plasma, 91.6. These results would indicate that the conductivities of defibrinated blood, or unclotted blood, and of plasma, or defibrinated serum, may be taken at will without introducing any appreciable error in our ’ calculations. ° Conductivity and hzmatocrite determinations were now made with the small apparatus on dog’s blood, the same precautions being observed as with the Utube. The results (in Table IV) again show a fairly satisfactory agreement. There is, however, a greater varia- tion in the individual electrical determinations than would now be considered justifiable after the apparatus and technique have been improved in various ways. EXPERIMENTS WITH THE SMALL APPARATUS ON HuMAN BLOOD. Blood from healthy students was obtained by puncturing the volar surface of the thumb with a sharp scalpel. The incision was about TABLE V. Giidents Conductivities Conductivities Percentage of Sober of blood at of the plasma plasma by 5°1C. x 10% || atiS° C. X10" || formulay@y 101.1 48.1 107.0 22 111.0 yA) 95.0 3915 101.0 62.3 8+.0 623 92:9 55-2 95.8 625 89.0 57.6 89.5 57.8 109.0 54.3 102.6 56.8 109.2 56.9 Measurement of Electrical Conductivity. 149 2 mm. deep, and five drops of blood were caught in a crucible resting on powdered ice. The tube was filled at once without defibrinating the blood and adjusted to the electrodes. The apparatus was inserted six minutes in the outer metal tube, surrounded with powdered ice, before the readings were taken. The results are given in Table V. The above results give a range in conductivity of serum from 84 to I1I, of blood from 35.3 to 53.8, of percentage of corpuscles from 48.1 to 62.3. Viola! gives a range for the conductivity of serum of healthy individuals from 106.18 to 119.12 (expressed in reciprocal Siemen’s units at 25° C.) The series was obtained from the blood of eight individuals. Converting these values into reciprocal ohms at 5°C., we get a range of from 82.6 to 92.7, which is considerably narrower than mine. Table VI gives results obtained on a healthy man when food and drink were taken. ‘The blood was defibrinated by means of a needle. An interval of seven minutes was always allowed before making a reading, as there was some question whether the six minutes allowed in the preceding series was sufficient to permit the blood in the tube to cool completely down too® C. A shorter time did not seem to give quite constant readings. Doubtless the use of a still narrower outer metal tube would lessen the time required. The range of the serum conductivities in this individual over a period of more than five weeks, when observations were made under conditions calculated to bring out the influence of the ingestion of food and water, was from 86.5 to 111, which is very nearly the same as the range for single observations on the series of healthy students in Table V. The range of conduc- tivities for the blood in Table VI was much narrower. In general a pretty constant inverse ratio is seen between the percentages of the serum and its conductivity, that is to say, when the percentage of the serum is high its conductivity is relatively low and vice versa. It is worthy of remark that the results in Experiment 12 of this table, in which a pathological condition (diarrhoea) was present, vary most from this inverse ratio. This is well illustrated by the last column in Table VI, which gives the products of the percentages of serum into the corresponding conductivities. It will be seen that there is a much closer approximation between the products than between the conductivities or percentages. The meaning of this would seem to be, that when the relative volume of serum is increased, ¢. g., by 1 VioLA: HAMBURGER: Loe. cit., p. 484. 150 ZT. M. Welson. drinking water, the serum becomes more dilute as regards salts, and therefore has a diminished specific conductivity. When the serum diminishes in amount, water seems to pass out of it in greater propor- TABLE VI. Per. | Percentages of serum multiplied ‘ Date of Conduc-| Conduc-| Per- | ciiace fe Sate Conditions of | tivity of | tivity of | centage b 5 abae. experiment. blood | serum | by for- fees by the con- san at.o° €.| ats° €; \mular(2): a ductivities tocrite. of serum. RR Niel | 4. 54.0 5831 4 hr. after meal : 5403 | Just after dinner : iil 5960 | 1 hr. after dinner : 5 5658 3 hr. after meal : 52: 5834 Drank 250 c.c. ! i 5664 distilled water ; waited 30 min. 1 hr. after dinner : Stele 5592 Drank 500 c.c. : hiss 5640 distilled water; waited 30 min. Drank 250 c.c. distilled water; waited 1 hr. | Drank 500 c.c. distilled water; waited 1 hr. | Mild diarrhcea Drank e250): distilled water ; waited 20 min. Drank 250 c.c. distilled water; waited 30 min, tion than salts. The conductivity of blood varied very slightly, only 5.7, which is much less than the variation for healthy individuals in Table V. The first five experiments also give the percentages of serum obtained by the hamatocrite, and differ from 0.6 to 8.4 per Measurement of Electrical Conductivity. 151 cent, the hzmatocrite readings being invariably lower. I am unable to suggest the reason why, in the observations in Table VI, in which the hzmatocrite readings were taken, these readings are in- variably lower than the percentages of serum obtained by the electri- cal method, whereas, in the observations on dog’s blood (Table I), the hzematocrite readings for corresponding percentages of serum were a little higher than the electrical determinations. It may be that the centrifugalization was not so complete in the observations in Table VI, as the hzmatocrite was not turned so long. But as I always cen- trifugalized till two successive readings showed no appreciable differ- ence, probably this is not the correct explanation, Whether a greater proportion of serum may not remain in the sediment of human blood than in that of dogs, I am unable to state. This is possible, but per- haps not likely. Finally it may be that the formula deduced from conductivity observations on dog’s blood is not quite accurate for human blood. Fraenkel! indeed, states that he obtained a close agreement between Bleibtreu’s method and the electrical method (using Stewart’s formula) for human blood. But he only used one specimen of healthy human blood. [am unable here, as in Table V, to compare my results with those of other investigators, since I find no published results on this subject. Table VII gives a few results on clinical cases. Hamatocrite readings were obtained by using the resistance tube as a hematocrite TABLE WIL. Number of | Conductivity of Conductivity of Bee eae ieee Ce case. blood at 5° C. | serum at 5°C. formula (2): Bp ain oven aati 1 45.2 96.6 60.5 53.0 2A 82.6 93.0 $1.3 90.4 92.4 82.0 90.0 105.6 68.8 62.0 97.6 58.8 50.9 52.8 tube, the length of the columns of corpuscles and serum being mea- sured by means of a pair of calipers and by the aid of a hand lens. 1 FRAENKEL: Loc. cit. 152 T. M. Wilson. These readings differed by less than 2 per cent from those obtained by the regular hamatocrite tube, and of course by etching a scale on the tube, still greater accuracy could be obtained. There is, moreover, the advantage that the tube is more easily filled and cannot possibly leak, as so often happens with the ordinary tube. Case 1.— Blood from a boy, twelve years of age, with adenoids ; temperature IOI°. Case 2.— Determinations were made on two specimens of blood (A and B) taken from the heart of a man four hours after death from pernicious anzmia. Clotting did not take place until two and one- half hours after removal. The blood count just before death was 816000 reds and 2900 whites, hemoglobin, 18 per cent. Case 3. — Blood obtained from a woman two days after, labor. Case 4.— Blood from a woman in the last stages of Bright’s disease. Case 5.— Patient was suffering from acute gastritis, and had been given four drachms of Epsom salts and two grains of calomel within the previous four hours, but without being purged. In the above five cases the blood was defibrinated, hermetically sealed in a fine tube and thus conveyed to the laboratory. Table VII shows a greater variation in the conductivity of blood than that found in normal individuals, the range being from 37.7 to 82.6. There was no corresponding variation in the conductivity of the serum. Viola! gives a range for the conductivity of serum in blood obtained from a few post-mortem cases of 98 to 140 expressed in reciprocal Siemen’s units at 25° C. or 76.3 to 109 in reciprocal ohms ats5° C. SUMMARY. 1. The electrical method of determining the relative volume of corpuscles and plasma gives results (in the case of dog’s blood) which upon the whole agree fairly well with the hzematocrite determinations. 2. The conductivity of serum from clotted blood is within the limits of error, the same as that of the corresponding plasma and of the corresponding serum from defibrinated blood. The same is true of the freezing-point of serum from clotted blood and from: the corresponding defibrinated blood. 3. The small apparatus described enables us to determine within less than I per cent the electrical conductivity of serum and less 1 VIOLA: HAMBURGER: Loc. cit., p. 484. Measurement of Electrical Conductivity. 153 than 2 per cent the conductivity of blood obtained in such quantities as are easily available for clinical examinations (four to five drops of blood). The time necessary for the determination of the conduc- tivity of serum (including the defibrination and centrifugalization of the blood) is from ten to thirteen minutes, of which about seven min- utes are required for the conductivity measurement; seven minutes more, making twenty minutes in all, are required for obtaining the conductivity of the blood, and thus for determining the relative volume of corpuscles and plasma. The resistance tube can be used as a hzematocrite tube. 4. In thirteen healthy students the conductivity of the serum varied from 84-111, and that of the blood from 35.3 to 52.6. 5. In one and the same healthy individual the conductivity of the serum, determined at intervals over a period of more than five weeks, and at various times after taking food and water, varied from 86.5— 111, and that of the blood from 40.2 to 46.7. 6. In a few cases of disease the conductivity of the serum varied from 91.6 to 105.6, while that of the blood varied from 37.7 to 82.6 the latter in a case of pernicious anemia. In conclusion, I have to thank Professor G. N. Stewart for his many kindnesses in suggesting this investigation, and in assisting me with it. THE EFFECT OF ‘SALT-SOLUTIONS (ON CilLiaks ACEIVAYE BY Ss. 5. MAXWETEE. [From the Laboratory of Physiology in the Harvard Medical School.| HE most extensive studies upon the effect of inorganic salts on cilia have been made by Lillie! and by Weinland.? Lillie inves- tigated the action of certain chlorides and mixtures of chlorides upon the muscular and ciliary activity of the larvze of Arenicola and Poly- gordius. Both these forms are marine and the cilia are exposed nor- mally to sea-water with its high content of inorganic salts. Most of the parallel studies upon developing eggs and embryos have been made upon marine forms. It would seem desirable that similar studies should be carried on at least as extensively upon tissues of fresh water or terrestrial animals before too sweeping generalizations are formulated. Weinland’s experiments were made on the ciliated epithelium of the frog’s mouth and cesophagus; but in order to study the stimulating action he used strong solutions, 3 mol. in most instances, and for this reason his results do not at all correspond to the effect of solutions of the same substances in strengths more nearly approximating the osmotic pressure of the cells. The experi- ments described in this paper were undertaken for the purpose of - investigating the effects of pure solutions of single salts upon the ciliated epithelium of the frog’s mouth and cesophagus in order to discover what general relation, if any, exists between the properties of the salts and their physiological action upon the tissue. In order to simplify the problem a limited series of salts having a common anion was employed; namely, the chlorides of lithium, ammonium, sodium, potassium, magnesium, calcium, strontium, and barium. These were employed in } mol. solutions. The salts used were the purest obtainable. Most of them were recrystallized, some of them several times. It was found to be especially important to avoid any trace of acid or alkali, since even a minute quantity has 1 LILLIE: This journal, 1901, v, p. 56; 1902, vii, p. 25; 1904, X, Pp. 419. * WEINLAND: Archiv fiir die gesammte Physiologie, 1894, lviii, p. 105. 154 The Effect of Salt-Solutions on Ciliary Activity. 155 a marked influence on the life of the cilia. The solutions were made up from water distilled in glass. All the solutions were prepared by exactly similar methods, and contained, presumably, fairly equal quantities of dissolved oxygen. The action of these solutions was investigated by two methods: 1. The effect which more or less prolonged immersion of the epithe- lium in the solution produces upon the work of the cilia. 2. The duration of life of the ciliated cells while immersed in the solutions. PRELIMINARY EXPERIMENTS ON WorK OF CILIA. The pioneer investigations on the work of the cilia were made by Wyman,! Engelmann,? and Bowditch.? The method employed by Bowditch was the only one which gave results in measurable terms of mechanical work. He placed the epithelium upon an inclined plane of known gradient so that the cilia by their action moved a load of known surface and weight against gravity. The movement was observed with 1 microscope containing a micrometer-ocular, and the rate determined with a stop-watch. From the known factors and the observed rate, the work was calculated in gram-millimetres per minute for a square centimetre of surface. In my experiments prac- tically the same method was employed. Experience proved that the cilia were very sensitive to changes of temperature. Ajir-currents were especially disturbing probably on account of changes in amount of evaporation from the surface of the epithelium. For this reason the inclined plane was placed in a galvanized iron tank 46 cm. long, 21 cm. broad, and 15 cm. deep, through the glass cover of which ob- servations were made. The tank was kept half full of water. This arrangement insured a good air-supply and a relatively constant temperature. The inclined plane used was of glass and had a gra- dient of 10 per cent. A rubber band around it formed a stop against which rested, during the observations, the paraffin block described below. In this way it was made certain that exactly the same por- tion of the preparation was used in all the observations. An eyepiece- micrometer was used one hundred divisions of which equalled a distance of 88 mm. and therefore a lift against gravity of eight- ninths of a millimetre. The time was measured with a stop-watch. 1 Wyman: American naturalist, 1871, v, p. O11. 2 ENGELMANN: Archiv fiir die gesammte Physiologie, 1877, xv, p- 493- 8 BowpitcH: Boston medical and surgical journal, 1876, xv, p. 159. 156 SS. Maxwell. The carriages employed were circular pieces of hard rubber 8.6 mm. in diameter, and weighing, one of them, 415 mgm., the other, 500 mgm. In connection with each of these was a set of lead weights which, placed on top of the rubber carriage, increased the load by even multiples of the original weight. In the description of experi- ments following it will be understood that the weight of the carriage employed will be considered the unit load. For use in these experiments the epithelium of the roof of the mouth was dissected out in connection with the cesophagus, which was cut open longitudinally on its ventral side. The preparation was spread out flat upon a paraffin block and pinned down with glass pins, care being taken to avoid wrinkles by sufficient stretching. The paraffin blocks were about 50 mm. long, 25 mm. broad, and 8 mm. thick, and were formed upon pieces of glass of the same rectangular area to weight them down when in the salt-solutions. During the time occupied in dissecting and pinning out the prepara- tions, considerable drying occurred, and observations made without moistening were found to be somewhat variable, the amount of vari- ability being much greater in some preparations than in others. Thus successive trials.in one preparation gave the following times in seconds occupied in carrying the 415 mgm. weight over 100 divisions of the micrometer-scale: 21, 25, 22, 24, 26, 25, 30, 32;024meae Another preparation gave 16, 15, 15, 14, 25, 20, 20, 20, 22.1 72uuue most constant results were obtained from specimens, moistened with frog-blood serum. The advantage of this is shown by the following rates from one of the preceding specimens, after moistening with serum: 17, 17, 16, 16, 17, 17, 17, 16, 17, 17. It was usuallysmecess sary, however, to reject the first few observations, because of the in- creasing irritability of the preparation, discussed below under the heading “ Staircase phenomenon.” In all these experiments one must take into account the great in- dividual differences between the specimens used even when the frogs have been just freshly taken, and these differences become much more pronounced when the animals have been kept for some time under the unnatural and relatively unhealthful conditions due to con- finement in the laboratory tanks. For this reason it was decided to throw out of consideration as abnormal all specimens which, on the first six or eight trials, when moistened with blood-serum, required longer than fifteen seconds to move 0.5 gm. weight a distance fifty divisions of the micrometer-scale; experience with such specimens The Effect of Satt-Soluteons on Ciliary Activity. 157 showed that their cilia became inactive in a very much shorter time than those of normal preparations kept under identical conditions. EFFECT OF LOAD. Before entering upon the investigation of the effects of the various salts, a study was made of the effect of load and of fatigue upon the power of doing work. These experiments were made upon preparations moistened with the ordinary laboratory physiological salt-solution, made up from commercial salt and tap-water, and also with specimens moistened with frog-blood serum. Although these experiments were merely preliminary, the results appear to have some interest in themselves, and are here reported in fuller detail than their relation to the subject would require. Under the conditions described above, if one begins with the unit- weight of say 0.5 gm., and then uses successively two, three, four, five, etc., times this unit, until a weight is reached which the cilia do not move, it will be found that the rate of movement has declined rather steadily; but that on employing the unit-weight again, the movement is slower than at first. The effect of fatigue has length- ened the time-intervals for the heavier weights. On this account, it is necessary to use larger multiples of the unit. In practice it was found convenient to use a series of loads in the proportion of 1, 4, 7, 10, 13, 16, 19, and 22. Even with such a series the results are far from being as regular as could besdesired. It is, therefore, necessary to make a very large number of observations with frequent controls. After trying various procedures, the plan finally adopted was to make a control with the unit-load, after each experiment with a heavier load. If the control at any time varied more than one second from the initial rate, a short period of rest was allowed to in- tervene, and the rate with unit-load was taken again. For example Specimen H of October 29 gave: Rest2min. Rest 4 min eee. 5 4 gk «Fel 10 I 131 16 f FE 19 T 1 Rate inseconds 8 10 7 12 7 15 8 21 8 24 10 9 60 10 10 Whenever, as in this instance, after load 19, the rate with unit-load did not return to within one second of the initial rate, the experiment was broken off, because the figures obtained for the heavier weights would not be fairly comparable with those obtained for the lighter ones. 158 SS. Maxwell. AUNBILIG, I EFFECT OF LOAD ON MECHANICAL WORK OF THE CILIA. In each experiment the upper line gives the number of seconds required to carry the load 50 degrees of the micrometer-scale, and the lower line is the corresponding work reckoned in gram-millimetres per minute for each square centimetre of epithelium. ~ Relative wt. of load Absolute wt. of load Experiment 1 15 1.27 13 1.461 * The preparation used in Experiment 10 was placed in the laboratory normal saline solution and after twenty-four hours gave the results shown in Experiment 11. Experiments 12 and 13 were also made on preparations which had been twenty-four hours in normal saline. The Effect of Salt-Solutions on Ciliary Activity. 159 Table I shows the results of thirteen experiments. The loads are in the ratios mentioned above, namely, 1, 4, 7, 10, etc. The unit- weight was 0.415 gm. The weights of the loads used are given in grams in the second line of the table. In each experiment the figure above gives the time in seconds for fifty divisions of the millimetre- scale. The lower number in each space gives the work in gram- millimetres per minute for each square centimetre of epithelium. A dash in the position for seconds in the table shows that the experi- ment had to be ended, because the preparation did not recover from fatigue. Table II differs from Table I only in the fact that the unit-load is 0.5 gm. The ratios of loads and the arrangement is the same as in Table I. TABLE II. Data arranged as in Table I. Relative wt. of load Absolute wt. of load | 0. i ; 5.0 17 2S Experiment 1 : 9.449 | 9.179 18 22 8.922 | 10.40 21 30 MOD 2 35 ent It will be seen from inspection of these tables that the curve of work of the cilia rises at first rather rapidly with increasing load, maintains itself near the maximum through a considerable range, and then falls again rapidly to the point where the load cannot be moved at all. The maximal in most cases is higher than that obtained by Bow- ditch, 6.805 gram-millimetres per square centimetre. The difference is probably due to the different size of carriage employed, 1.437 square centimetres in his experiments, and 0.581 in mine; but this has not been sufficiently investigated. The shape of the carriage, oval in the one case and circular in the other, may also be a factor. It is evident, of course, that the figures given above represent only a fraction of the actual work of the cilia, for the force of their strokes 160 S. S. Maxwell. is applied, not directly to the load to be moved, but to the liquid with which its surface is moistened. While this is true, it seems fair to assume that the degree of regularity shown indicates that actual re- lations are exhibited, although not actual quantities. The quantity and quality of mucus-or other liquid present very probably affects the results; but this did not prove so important a factor as might be ex- pected. Except where a clotted mass of material was present, very little difference was observed in the rates before and after removal of so much of the mucus as could be taken off with a moist camel’s-hair brush. It is possible that the three cases given in Table I, Experi- ments II, 12, and 13, where the preparation had been for twenty-four hours in a normal saline solution, may indicate the necessity of mucus in the moving of heavy loads. In each of these, the curve of work falls very rapidly. These preparations, owing to the exhaustion of ~ the goblet cells, are almost free from mucus, and the slow rate of the heavy weights might be due to the lesser adhesiveness of. the liquid between the carriage and the epithelium. On the other hand, such preparations are especially susceptible to fatigue, from which they recover with extreme slowness, and this would suggest that the real cause of the characteristic form of their work-curve is an actual low- ering of their working power, and not a change in the character of the liquid with which they are moistened. STAIRCASE-PHENOMENON. Kraft! and others have called attention to the fact that cilia are markedly responsive to mechanical stimulation. There is also good reason for believing that up to a certain degree each stimulus causes the cilia to become more irritable to succeeding stimuli. I find among my notes many records, in which, at the beginning of the experiment, the rate has increased with a number of successive appli- cations of the same load. In some of them the staircase-effect is quite as well marked as in the characteristic examples of muscular work. For example, in an experiment of October 13, the following successive rates were obtained : 28, 23, 17, 16, 15, 13, 14; 12,0 3ymue In another experiment, October 10, the rates were 27, 26, 22, 23, 18, 17, 17,17. In other instances, the first trial or two are markedly slower than the succeeding ones, which then run along with a fair degree of uniformity. 1 KRAFT: Archiv fiir die gesammte Physiologie, 1890, xlvii, p. 215. Lhe Effect of Salt-Solutions on Ciliary Activity. 161 A similar phenomenon was often seen in specimens which had been kept for some time in salt-solution, for example, a preparation which had been twenty-seven and a half hours in normal saline solution gave the following : 30, 15, 14, 13, 13, 13, 13. It is possible that in these cases the solution was deficient in oxygen, and that the increase in rate was due to gradual recovery from partial asphyxiation; but the phenomcnon is so similar to that seen in fresh specimens, that such an assumption is unnecessary. FATIGUE. The occurrence of fatigue of the cilia has been already mentioned in discussing the effect of work. It might be supposed that the slow- ing of rate, which occurs after heavy loads have been moved, is not a true fatigue phenomenon, but is caused by accumulation of mucus, or by drying. That this is not true is shown by the fact that often a few minutes’ rest is sufficient, without removal of mucus or addition of moistening liquid, to restore the original rate; and also by the fact, already mentioned, that preparations which have stood for a consider- able time in normal saline solution are much more quickly fatigued than fresh specimens. In such specimens, the goblet cells are practi- cally exhausted, and the surface is almost free from mucus. THE EFFECT OF SALT-SOLUTIONS UPON WORKING POWER. Table III shows the results of the experiments with the eight solu- tions. Each preparation was moistened first with frog-blood serum, and the rate determined with the 0.5-gm. load. It was then washed with the solution and the rate determined again; a third determination was made after one hour’s immersion in the solution, a fourth after four hours, and, if still active at that time, the preparation was tested again at the end of twenty-four hours. The figures given are the seconds required to move the 0.5-gm. weight fifty divisions of the micrometer- scale. Whereacross appears in the column, it means that the weight moved perceptibly but so slowly that the determination of the exact time would have had no special significance. A zero indicates that the weight did not move. Since all the figures are for the same load, it is unnecessary to present the calculation of work done. The work is, of course, in inverse proportion to the rate. In these observations, the rate for the 2-gm. and the 5-gm. loads was also recorded, but no fact of importance was brought out by them. 162 SS. Maxwell. ALINE ON JOE Effect of salt-solutions on working power of the cilia. Seconds required to move 0.5 gm. fifty divisions of the micrometer-scale. = No. of In In Salt. exp Pee celine 1 hour. 4 hours. | 24 hours. LiCl 1 7 10 - 0 2 9 14 126 0 3 il 14 0) 0) 4 14 16 + 0) NH,Cl ] Tl 6 50 4- 0) 2 9 6 60 0) 3 13 9 90 (0) - 4 12 11 180 OF NaCl J 9 9 Wi iy 0) 2 13 7 14 15 0) 3) 13 9 20 22 _ 4 li 7 21 23 =F Kel il 11 7 85 a 0) 2, 8 9 62 0) $) AS 7 170 + 0) 4 14 11 = 0) MgCl, 1 10 10 43 0) 2 8 9 23 0) 5 1a 1] 24 0) 4 1s} ial 74 0) CaCl, il 9 7 35 + 6) 2 8 8 87 O 0) 3 12 9 S4 0) ¢) 4 13 ll 72 0) SrCl, ] 12 7 26 0) 2 15 12 30 0) 3 9 7 15 420 ¢) + 9 10 20 740 O BaCly 1 8 10 a 0) 2 15 10 — 0 3 10 7 0 0) 4 9 $ ot 0) + The weight moved perceptively, but so slowly that the determination of the exact time would have had no special significance. 0 The weight did not move. Inspection of the table will show the following suggestive facts : 1. Of the eight solutions employed, sodium chloride is much the most favorable for preserving the cilia in such condition that they can do mechanical work. All the specimens in this solution were The Effect of Salt-Solutions on Crliary Activity. 163 o able to do work at the end of four hours, and almost one-half of them could still move the weight perceptibly at the end of twenty- four hours. 2. Arranged in the order of most favorable action upon the work- ing power of the cilia, these solutions stand sodium, strontium, potas- sium, calcium, ammonium, magnesium, lithium, barium. 3. Most of the solutions exert a stimulating action when first ap- plied, as can be seen by comparing the first and second columns of figures. The most stimulating is ammonium; it is also one of the most injurious. 4. Magnesium and lithium do not stimulate. Lithium, especially, has from the first a markedly depressing effect. * DURATION OF CILIARY ACTIVITY IN THE VARIOUS SOLUTIONS. This method of comparison possesses several advantages over the preceding one. Chief among these is the fact that several specimens can be made from the throat of one animal, and these placed in differ- ent solutions can be used as checks upon one another; while in the study of work done each animal furnishes material for a single ex- periment only. Again, the observations are made by means of the microscope directly upon the cilia themselves, and there is no ques- tion as to the possibility of error through change in the quantity or quality of mucus, or through drying during observation. Then, too, as the evidence presented will show, it is possible for a solution, by acting upon the cement-substance, to loosen the cells from their attachment, and thus destroy their power of moving a load without at the same time exercising a markedly injurious action upon the cilia. On the other hand, there are certain decided disadvantages. The cilia cannot be observed zz sztw with sufficient certainty, and teasing off portions for examination leaves the specimen each time in less and less favorable condition. Besides this, not all portions of the epithelium act with equal vigor,! and, under like favorable conditions, some portions probably live longer than others. Another disadvan- tage lies in the fact that, in almost every preparation, some cells are 1 GRUTZNER: Physiologische Studien von Griitzner und Luchsinger, 1882, p. 15. (Festschrift fiir Valentin. ) 164 S.S. Maxwell. very greatly more resistant than others, and hence it is difficult to say when ciliary action has ceased to be general.! In the experiments here recorded, each specimen was cut into a number of pieces, usually either four or six, and each piece was put into a different solution. The sets of experiments were arranged so as to overlap, and by some solution common to both, usually sodium chloride, act as controls upon one another. In order to remove adherent liquid each piece was rinsed in a quan- tity of the solution in which it was to be placed, the liquid was then poured off and a supply of clean solution added. Observations were made from time to time by teasing off small portions of the ephithe- lium, and examining with a Leitz No. 7 objective. In these exami- nations it was necessary to keep in mind the fact mentioned by Engelmann,? that cilia which have come to stand still in a salt-solu- tion became active when stimulated by teasing, or by exposure to the air, but quickly return to the inactive state. It is not easy to give the results of these experiments without re- peating in tiresome detail the notes upon the various preparations, but the main results are sufficiently shown in Table IV. SAB ICE SVE Average duration of activity of the cilia of the frog’s cesophagus in pure } mol. solutions. TiGi ae ee Zoimtoo hours: MgCl: > - = 28itos5shours: INDECENT Ge 5 Ale Sky CaGlo-. Be Zone Soe NENG oe oo. 6 UO Sr@loms (oe ey OO Ome HCC arr eh ig, SL FE ah0) Ba@lo = 0 = 0s 0 4 Omen Inspection of the above table shows : 1. That the cilia live much longer in a pure sodium-chloride solu- tion than in a solution of any other single salt. 2. That the chlorides of strontium and barium stand next to sodium in power of maintaining ciliary activity. 3. That no marked difference exists in the effect of lithium chloride, magnesium chloride, and calcium chloride. 1 A similar fact was observed by BARRAT in his study of the effect of acids and alkalies on living Paramcecium. He found that in a tube containing about sixty specimens, almost all of which were dead in fifteen minutes, there were likely to be two or three individuals which would remain alive from thirty minutes to an hour or more. Zeitschrift fiir allgemeine Physiologie, 1904, iv, p. 441. 2 ENGELMANN: Jenaische Zeitschrift, 1868, iv, p. 321. The Effect of Salt-Solutions on Ciliary Activity. 165 4. That no marked difference exists in the two most unfavorable salts, potassium chloride and ammonium chloride. If these eight solutions are arranged in the order of their power to maintain ciliary activity, they would probably stand ag follows: So- dium, strontium, barium, magnesium, lithium, calcium, potassium, ammonium. COMPARISON OF THE RESULTS OF THE Two METHODS. If we compare this order with the order obtained in the experiments upon the work of the cilia, it will be seen that they are markedly dif- ferent.! Such a difference, however, is not surprising when considered in connection with certain facts in regard to rhythmical contraction of muscle. In one solution the tissue may retain its irritability for a con- siderable time without making any rhythmic contractions. In another rhythmic contractions may occur for a time, and the muscle lose its irritability. An apical strip of tortoise ventricle placed directly in serum or Ringer’s fluid usually fails to show spontaneous rhythmic contractions, but retains its irritability for many hours.?, On the other hand, if a strip is placed in a pure solution of sodium chloride, it goes for a time into a state of rhythmic activity; but its con- tractions decrease in energy until irritability is finally lost in a much shorter time than that of the strip which has remained quiescent in Ringer’s. It is certainly possible that a solution which causes a tissue to go for a time into a state of intense metabolism may at the same time act unfavorably upon the continued life of the tissue. It seems to me that in the study of rhythmic contractility there has not been so far sufficient discrimination between condi- tions which favor the retention of irritability and conditions which by reason of the irritability bring about periodic activity. In the analysis of the conditions of rhythmic action, three factors need to be looked for, namely, (1) those which maintain the irritability of the 1 In order to test the possibility that the preparations for work were in part shielded from rapid diffusion by pinning to the paraffin block, and hence that the two sets of results might not be comparable, several control experiments were made. Each preparation was cut into equal longitudinal strips. One of the strips was pinned down to a paraffin block and placed in a dish of solution as in the experiments on work; the other strip was immersed in another dish of the same solution. No essential difference was found in the duration of life in the two sets of preparations. 2 GREENE: This journal, 1899, ii, p. 125. 166 S.S. Maxwell. tissue ; (2) those which, aside from any nutritive function in the ordi- nary sense, enter into the process by which mechanical energy (con- traction) is produced; and (3) those which act as the stimulus and cause the discharge of the contraction-process. It is possible that in the application of a solution of a single salt to the ciliated cell these factors would be unequally influenced, and that different salts would act most markedly upon different factors. We should then expect to find that one salt, while maintaining the cell in a living condition for a considerable length of time, might very materially lessen the amount of work done, by not furnishing the necessary condition of stimulation. Another salt could cause vigor- ous action until the store of available energy was exhausted. Or, to state the matter in general terms, the difference between the actions of the various salts is qualitative as well as quantitative. And even when there is little quantitative there may be considerable qualitative difference. As regards effect upon the duration of action and the working power of the frog’s ciliated epithelium, lithium and ammo- nium both stand very low; but it is evident that they do not both act in the same way, for lithium depresses, while ammonium stimulates. Lithium is not quite so injurious as ammonium, but we cannot con- clude that this is because it does not stimulate, for sodium and stron- tium, which are much more favorable, also stimulate. It is fair to conclude then that lithium and ammonium owe their characteristic action to effects upon different factors in the life of the cell. In comparing the order of salts as to favorable action upon work- ing power and duration of life of the cilia, the most striking differ- ence is shown in the position of barium, which is among the most favorable to continued activity, and the most unfavorable of all in its influence upon the working power. In this case the explanation is simple. All the specimens which stand for a time in a salt-solution undergo a certain amount of maceration, by which single cells or cell- groups become loosened from their connection with other cells. In some of the solutions this is insignificant in amount, but in barium chloride the effect is extreme. Within a few hours multitudes of the ciliated cells are loosened from one another and float free in the liquid, the cilia remaining vigorously active. Under the microscope one often sees these cells swimming actively about, or spinning round and round by the action of the cilia, and exhibiting a marked resemblance to free-swimming protozoa. It is evident that as soon as this process of separation of the cells begins, the working power of the epithelium The Effect of Salt-Soluteons on Crliary Activity. 167 as a whole must be exceedingly reduced, while the life of the indi- vidual cells is not necessarily affected. This action of barium is par- ticularly instructive, for it shows an instance of marked action upon one part of the tissue—the cement-substance between the cells — without any very great effect upon the cilia. The comparative position of potassium in the two orders referred to is, next to barium, the most noticeable. In this case no external cause for the differ- ence was discovered, but it is not improbable that there is here as sharp a differentiation in the action upon the internal constituents of the cell as that which occurs more externally in the case of barium. INTERPRETATION OF RESULTS. I wish now to consider whether the results obtained can be ex- pressed in terms of any general relation between the properties of the solutions employed in the foregoing experiments. 1. Molecular weights. — Weinland! concluded from experiments with a considerable number of solutions, including strong solutions of six of the eight salts which I have employed, that, with few exceptions, the stimulating action upon the cilia of the epithelium of the frog’s cesophagus has a direct relation to the molecular weight ; the higher the molecular weight, the more stimulating and also the more injurious the substance. With } mol. solutions, however, this is certainly not the case. In Tables III and IV the salts are arranged in the order of their molecular weights, and this is very far from the order of their physiological effect. Ammonium chloride, the most injurious salt in the series, stands next to sodium chloride, the most favorable. Barium chloride, with its very high molecular weight, is more favorable to continued life of the cilia than lithium chloride, with its low molecular weight. Strontium chloride is more favorable than magnesium chloride. In fact, so far as these eight salts are concerned, absolutely no relation between molecular weight and physiological action can be found. 2. Valence. — The question of the possibility of a relation between toxic action and valence of the cation has been discussed by Loeb,’ Mathews,? and others. 1 WEINLAND: Archiv fiir die gesammte Physiologie, 1894, lviii, p. 131. 2 Logs: Archiv fiir die gesammte Physiologie, 1got, Ixxxviii, p. 68; This jour- nal, 1902, vi, p. 429; Lors and Girs: Archiv fiir die gesammte Physiologie, 1902, xcili, p. 246. 8 MATHEWS: This journal, 1905, xii, pp. 424 and 438. 168 S. S. Maxwell. Inspection of Tables III and IV shows that in the action of these salts upon the frog’s epithelium no direct relation exists between valence and the physiological action upon the ciliated cells. Stron- tium chloride stands next to sodium chloride. Magnesium chloride and lithium chloride have nearly equal effects. 3. Solution-tension. — Evidence of a relation between physiological action and solution-tension has been presented by Mathews.! Ac- cording to this, we should expect that, as shown by Table IV, sodium, barium, and strontium would be among the more favorable, and cal- cium, lithium, and magnesium among the more injurious. But potas- sium should be the most favorable of all. Barium should be more favorable than strontium, and magnesium more injurious than lithium. If we apply the correction which Mathews? suggests, we do not yet bring them into the order of the experiments, and are led to the con- clusion that the formula is not applicable to the ciliated epithelium of the frog’s throat. 4. Probability of the existence of a direct relation. —If we could regard living protoplasm as a definite compound possessed of clearly marked diagnostic properties, we might well seek to find a direct relation between this compound and a given series of inorganic salts, and we might have reason to expect that the salts constituting the series would be active in proportion to certain definite properties. We should also expect to find that results obtained for the cilia of one organism would hold good for another, and that principles which were found applicable to these would hold for other forms of proto- plasm, as nerve, muscle, or developing eggs. If, on the other hand, living protoplasm is a mixture rather than a definite compound, a mixture containing carbohydrates, lipoids, and proteids of more than one kind, and if the proteid-mixtures are different in different organ- isms and different tissues, we cannot expect to find a general for- mula which will express the physiological action of such a series of salts. A comparison of the effect of the chlorides of sodium, potassium, and magnesium upon cilia will throw some light on this question. According to Lillie’s observations a pure sodium-chloride solution stops ciliary action almost instantly in the larve of Arenicola® and in a short time the cilia become completely dissolved. In Polygor- 1 MATHEWS: This journal, 1904, xi, p. 481. * MATHEWS: This journal, 1904, xi, pp. 485-488. ® LILLIE: This journal, 1901, v, p. 62. The Effect of Salt-Solutions on Ciliary Activity. 169 ‘dius! the same kind of action occurs, but not quite so promptly. In both of these, magnesium chloride? and potassium chloride act much more favorably, maintaining ciliary motion for a considerable time. In the frog’s epithelium we have seen that a very different relation exists; and hence we have the right to infer that the protoplasm of the cilia in the two instances is of dissimilar composition. It may be objected to the above comparison that the two kinds of cilia men- tioned are adapted to life in very different media, but this objection is an admission of the existence of different properties with a prob- able difference of composition, and hence of reaction to inorganic salts. Comparing those cilia, however, which are adapted to the same media, we still find suggestive differences. For Parker® has found that the cilia of Metridiam, which, like Arenicola, live normally in sea-water, may be retained in good condition for many hours in a pure solution of sodium chloride. It is evident that the line of reasoning just presented may well be extended to other structures than cilia. Solutions which allow ciliary action to continue render muscular contractions impossible* and muscular contractions can occur in solutions which destroy ciliary activity.° In each case the reaction depends upon the composition of the tissue, as well as upon the inorganic reagent. The properties of the series of protoplasmic structures must be considered, as well as those of the inorganic salts employed. If this is true, no general formula can be constructed from series on either side without taking into consideration the individual properties of the entire series on the other side. SUMMARY. The results of the experiments with the chlorides of lithium, am- monium, sodium, potassium, magnesium, calcium, strontium, and barium in i mol. solutions upon the ciliated epithelium of the frog’s cesophagus may be briefly summarized as follows: 1. Of the eight salts examined, sodium chloride is the most favor- able to the prolonged life of the cells and to the preservation of their power to do mechanical work. 1 LILLIE: This journal, 1902, vii, p. 41. 2 LILLIE: This journal, 1904, x, p. 422. 8 PARKER: This journal, 1905, xiii, p. 6. 4 Logs: This journal, 19009, iii, p. 336. BLIGE + Loc. cit. 170 S. S. Maxwell. 2. Arranged according to their effects upon the working power of the cilia, their order is not the same as when arranged according to effects upon duration of cell-life. 3. The physiological action of these salts upon the frog’s ciliated epithelium bears no direct relation to valence of the cations or to molecular weight. 4. In general, salts of higher solution-tension are more favorable than those of lower; but the order of favorable action of the individ- ual salts does not agree with the order of their solution-tensions. FURTHER OBSERVATIONS ON THE CATAICY LIC DECOMPOSITION OF HYDROGEN PEROXIDE. By A. S: LOEVENHART. [From the Laboratory of Physiological Chemistry of the Johns Hopkins University.] HE recent work of Cohnheim! on the glycolytic action of the ex- pressed juices of the pancreas and muscle has aroused great in- terest in consequence of the light it seems to throw on the nature of diabetes and the rdle of the pancreas in carbohydrate metabolism. Cohnheim showed that the expressed juices of the pancreas and muscle have individually very little glycolytic activity, but mixtures of the two in the proper proportion possess very marked power to destroy glucose. Rahel Hirsch,? working independently, found that a similar relation obtains between the liver and pancreas. Arnheim and Rosen- baum ? (using different methods) found that the pancreas, liver, and muscle possess individually distinct glycolytic activity, but that com- bining pancreas with the other tissues causes marked acceleration. Combining liver with muscle causes no acceleration. Various ob- servers have found that many organs, namely, pancreas, muscle, liver, ovary, kidney, etc., possess some glycolytic activity. All this work seems to indicate that the various tissues possess some slight power to destroy glucose, and that the pancreas furnishes an internal secretion which greatly increases the glycolytic action of the tissues. The great importance and practical bearing of this work make it important to check the results in every way. The nature of the glycolytic process has not been cleared up in the least by these recent contributions, and we are at present entirely ignorant of the fate of the sugar in the gly- colytic process. It occurred to the writer that it would be interesting in this connection to study the effect of mixing extracts of different organs on some other better understood chemical process which all 1 CoHNHEIM: Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 336; 1904, xlii, p. 401. * RAHEL HirscH: HOFMEISTER’S Beitrage, 1904, iv, p- 535- 8 ARNHEIM and ROSENBAUM: Zeitschrift fiir physiologische Chemie, 1903, xl, Pp. 220. 171 L72 A. S. Loevenhare. the tissues are capable of bringing about to some degree. The catal- ysis of hydrogen peroxide was selected for this purpose, because the determinations are rapid and fairly accurate, and this reaction was being studied in other connections. The organs used were obtained from dogs and pigs; the extracts which were at first employed being prepared as follows: 10 gm. of the freshly removed tissue was thoroughly ground up with sand, and extracted with water. This was SERTES LIVER AND PANCREAS EXTRACT (10 PER CENT). 1. | 2. 3. 4. 1 ‘cc. liver. 4 c.c. pancreas. | 1 c.c. liver. S of 1 and 2. 4 c.c. water. |1 c.c. water. eine 4 c.c. pancreas. Time in seconds. 5.4 1.0 8.0 14 10.1 17 ED 2.1 13.5 2.4 15.0 2.8 16.3 3:32 17.4 3.5 then strained through cloth and diluted to 100 c.c. and toluene added. These extracts were sometimes employed for several days without much change being noted. They are always faintly acid. The method of following the reaction was the same as that used by Kastle and the writer.’ 5 c.c. of hydrogen peroxide were placed in a small bottle, which was in turn placed in a larger one containing the extract. The larger. bottle was then connected with a gas burette. The ex- periment was started by overturning the smaller bottle without open- ing the apparatus, thus bringing the hydrogen peroxide in contact with the extract. At intervals of fifteen seconds, the volume of the oxygen liberated was read, and is recorded in cubic centimetres in the * LOEVENHART and KasTLE: American chemical journal, 1903, xxix, pp. 397, 563. Catalytic Decomposition of Hydrogen Peroxide. 173 tables which follow. The duration of each experiment was two min- utes. The gas volumes are not corrected for temperature and pressure, as only relative results were required, and all of the experiments of each series were performed within a few hours. Moreover, no attempt was made to draw conclusions from small differences, only striking results being considered. During the experiments, all of which were performed at room-temperature, the bottle was shaken contin- uously in order to get uniform results. The hydrogen peroxide em- ployed was the commercial Oakland “ Dioxogen,” which has an acidity varying from 74 — #% in different specimens. 5 c.c. of this hydrogen peroxide solution were used in each experiment, and this amount yields from 55~—74 c.c. of oxygen on complete decomposition. The first experiments (Series 1) showed that by mixing liver and pancreas extracts the reaction is accelerated very remarkably. By comparing Columns 3 and 4 it will be seen that the reaction pro- ceeded from 230 per cent to 330 per cent faster when the liver and pancreas acted together than when they acted separately. From Series 2 we see that the combination muscle and liver is practi- cally as active as pancreas and liver. Another series of experiments SERIES 2. LIVER AND MUSCLE. ie 2. 3. 4. 1 c.C. liver. 4 c.c. muscle. San or lencaiel esc liver. 4 c.c. water. 1 c.c. water. 4 c.c. muscle. Time in seconds. 15 Ses 1.3 : 27.8 30 8.0 1.9 38.7 45 10.1 2.6 48.5 11.9 3.2 13.5 3.8 15.0 4.3 16.3 4.9 174 5.3 carried out with 1 c.c. liver extract, 2 c.c. pancreas and 2 c.c. muscle extracts acting simultaneously give practically the same result as that 174 A. S. Loevenhart. recorded above for the liver and pancreas and for the liver and muscle. It was found that combining pancreas and muscle caused only a very slight acceleration. This result is quite contrary to what has been observed in the work on glycolysis. Arnheim and Rosenbaum found that mixtures of liver and muscle show no acceleration of the glycolytic process, whereas mixtures of pancreas and muscle showed marked acceleration. Cohnheim' reports in his last communication that the activating substance of the pancreas in the glycolytic process is not destroyed by boiling. Some of the pancreatic extract used in the above experiments was boiled and filtered from the coagulated masses. It was found that this boiled extract which, to be sure, had no activity in itself, was capable of accelerating the liver to the same degree as that shown above by the unboiled extract, as is seen in.Series 3. SERIES 3. ; : _ 1 c.c. liver. Hen ta I c.c. liver. 4 c.c. boiled seconds. 4 c.c. water. pancreas extract. 222 34.4 45.6 55.0 61.2 65.7 68.1 69.2 The reaction under consideration resembles the glycolytic process in that the accelerating effect of the pancreas is not destroyed by boil- ing. Cohnheim ? found that increasing the amount of the expressed juice of pancreas with which a given amount of muscle juice was mixed, causes an increase in the rate of glycolysis up to a certain point, beyond which a further increase in the amount of pancreas juice markedly diminishes the velocity of the reaction. No such optimum for the amount of pancreas acting has been found for 1 COHNHEIM: Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 401. 2 COHNHEIM: Loc. cit. Catalytic Decomposition of Hydrogen Peroxide. 175 the reaction here studied. Within ordinary limits, an increase in the amount of pancreas acting causes an increase in the rate of the catalysis. In order to see if any peculiar relationship exists between any of the organs with reference to this reaction, extracts of the liver, adrenal, lung, muscle, brain, spleen, kidney, and pancreas of a dog were prepared, and the action of each tested on the other. The only noteworthy result was that in every case where two extracts were mixed together, some acceleration was noted, though in several in- stances it was very slight. We next determined to find, if possible, _ the cause of the acceleration, and with this in view a clear solution of the enzyme was prepared. This was made by using a modifica- tion of Jacoby’s uranyl acetate method similar to that recommended by Rosell. 100 c.c. of a 10 per cent extract of the liver or pancreas was treated with 20 c.c. of a saturated solution of uranyl acetate, and the mixture neutralized by adding a few drops of a saturated solu- tion of sodium carbonate and sodium phosphate. After standing a few moments, 5 c.c. of a saturated solution of sodium phosphate were added, and the mixture centrifugalized and then filtered. The water- clear filtrate thus obtained was always found to possess strong cata- lytic activity whenever the original extract was active. Boiling causes it to lose all activity. This solution can be kept several days by adding a few drops of toluene, and only requires to be filtered from the precipitate, which slowly appears. These clear extracts contained only a very small amount of proteid. They were always amphoteric or very slightly alkaline in reaction; whereas the turbid extracts were always slightly acid. The results obtained with the turbid extracts were confirmed with the clear extracts. It was found that not only can boiled pancreas accelerate the decomposition by liver, but the boiled liver itself can accelerate the action of fresh liver. These points are brought out in Series 4. In all experiments which follow, the clear liver extract was diluted with 14 volumes of distilled water in order to reduce its activity sufficiently to show the acceleration well. In the experiments in which the boiled liver extract was used, the undiluted 10 per cent extract was employed in order to have the conditions exactly comparable to the experiments with the pancreas. By comparing Columns 3 and 4, it will be seen that the reaction is 1 RosELL: Inaugural dissertation, Strassburg, 1go1 (cited by MAGNus: Zeit- schrift fiir physiologische Chemie, 1904, xlii, p. 149). 176 A. S. Loevenhart. accelerated from 100 to 300 per cent by allowing the liver and pan- creas to act simultaneously. Columns 5 and 6 show that boiled liver and boiled pancreas have about the same power to accelerate the catalysis by fresh liver that fresh pancreas has. In these ex- periments the action of the liver was increased about seven times in the earlier readings. By comparing Columns 2, 7, and 8, it will be observed that neither the boiled liver nor the boiled pancreas is capable of accelerating the catalysis by fresh pancreas to any SERIES 4. Time in seconds. 2. ec pancreas. cic: water. 4. Irexer liver. Ik eyes pancreas. 218 31.0 38.0 43.2 48.1 51.6 5. li ele: liver. il! eke: boiled pancreas, 18.8 26.0 6. lee. liver. cic: boiled liver. 209 28.8 33.4 38.8 429 46.2 ibe were: pancreas. 1 e:e: boiled liver. ast Se) 6.3 6.8 les: 8.0 8. ec pancreas. ice: boiled pancreas. 5 : 48.8 8.6 5: : 50.9 ON: extent. Thus we see that the action of boiled liver and boiled pancreas is the same, namely, they both cause a very great accel- eration in the action of liver, but neither is capable of acclerating the action of pancreas to any extent. This indicated that liver and pancreatic catalysis may be different, and some experiments were made in order to further test this point. With this in view the action of several substances on catalysis by liver and pancreas extracts was investigated in the hope of finding that certain substances would accelerate the one and retard the other, and thus confirm the non- identity of the two catalases. Solutions of sodium sulphate, sodium thiosulphate, thiourea, ammonium sulphocyanate, and ammonium nitrate were employed, the results being given in Series 5 and 6. In each experiment 1 c.c. of the extract and 1 c.c. of the solution Catalytic Decomposition of Hydrogen Peroxide. 177 to be tested were used. Experiments (1 and 7) with 1 c.c. water is given for comparison. A comparison of the corresponding columns shows that qualitatively the effect of these substances on the catalysis by liver and pancreas extracts is the same, but certain quantitative differences are very SERIES 5. LIVER EXTRACT (4 PER CENT) Time in 1. 2. 3. 4. 5. 6. seconds.| Water. | 7 Na,S Og. | 3 Na,SO4.| « NHyNOsg. | 2 NH,CNS.| § thiourea. a 15 1.0 9.7 08 — 0.1 0.1 1.4 30 1.6 11.9 08 0.2 0.1 3.4 45 2.0 140 2.0 0.3 0.2 5.3 60 2.6 16.0 2.4 0.3 0.2 6.5 75 3.0 18.1 2.9 0.3 0.2 7.6 90 3.5 20.1 3.3 0.3 0.3 8.5 105 4.0 22.4 BY 0.4 0.3 9.3 120 4.4 24.2 41 0.4 0.3 9.9 SERIES 6. PANCREAS EXTRACT (10 PER CENT). Time in Up 8. | 9. 10. at 12. seconds.| Water. | 2 Na,S.Oxg.| # # Na,SO,. | # NH,NOs3. | # NH,CNS. | § thiourea. 15 1.6 7 Mey 0.2 0.2 Zit 30 2nf 6.9 Dall 0.2 0.3 357 45 Shi 8.1 3.5 0.3 0.4 49 60 4.5 93 4.3 0.4 0.4 5.8 75 5.4 10.6 Sel 0.4 04 ded 90 6.2 11.8 5.8 0.4 0.4 7-3 105 7.0 13.2 6.6 0.5 0.5 7.9 120 7.7 14.5 td 0.6 0.5 8.4 178 A. S. Loevenhart. marked. Thus the action of the liver extract is accelerated from 400 per cent to 800 per cent by sodium thiosulphate, whereas pan- creas extract is only accelerated from 100 per cent to 250 per cent. The thiourea also accelerates the action of the liver more than it does the pancreas extract. No conclusion as to the identity or non- identity of hepatic and pancreatic catalase could be drawn from these experiments. The effect of temperature on the activity of the two ex- tracts was studied. It was found that the activity of liver extract at I’ is 41.5 per cent of its activity at 20°, and the activity of the pan- SERIES 7. aE 2. 3 4 5 6. Timean Pears PO. 2.0 ce. + Cc. 40 ce. Slee anarriee iver. liver. iver. lver. iver. liver: ces) CHES 4: Oleic: 3.0'c:e: 2.0 c.c. 1Olcre: water. water. water. creas at I° is 48.5 per cent of its activity at 20°. Hence temperature affects the activity of liver and pancreas to about the same extent, and the writer concluded that the enzymes are probably identical, and that the differences observed are due to differences in the environment in the liver and pancreas. The fact that boiled liver can accelerate the catalysis by liver, as has been already pointed out, proved that the liver must contain be- sides the catalase a substance which is not destroyed by heat, and which is capable of accelerating the action of the liver catalase. As the liver contains a substance which can accelerate its own catalase, it was thought possible that by increasing the amount of liver acting, more than a corresponding increase in activity might be noted. Ac- Catalytic Decomposition of Hydrogen Peroxide. 179 cordingly experiments were made in which the concentration of the liver extract was varied, while the concentration of the hydrogen peroxide remained constant. The results are seen in Series 7 (4 per cent liver extract used). Columns 1 and 2 show that for small quantities of the extract the rate of the oxygen liberation is proportional to the amount of extract acting. When larger amounts of extract are used, however, an in- crease in the amount of extract acting causes far more than a propor- tional increase in the rate of oxygen evolution. Thus by comparing the first readings in Columns 1 and 6, it will be seen that the reading in Column 6 is ten times as large as it would be if the liver acted in proportion to the amount present. A similar series was carried out. with a 10 per cent pancreatic extract, the results of which follow : SERIES 8. I 2. 3. 4. 5. Time in 1 c.c. pan- 2 c.c. pan- 3 c.c. pan- 4 c.c. pan- 5 c.c. pan- seconds. | creas extract. | creas extract. | creas extract. | creas extract. | creas extract. 4 c.c. water. 3 c.c. water. 2 c.c. water. 1 c.c. water. No water. 0.5 13 4.1 5.9 8.3 0.7 Wee) 5.7 8.0 0.8 2.5 1.2 o9 0.9 Sell 8.5 1.0 S| 9.8 1.1 4.3 11.0 12 4.9 12.2 1.3 55 1325 Thus an increase in the amount of pancreas acting alone also causes more than a corresponding increase in the rate of the catalysis, but the figures are not so striking as in the case of the liver. This was expected from the fact that the action of pancreas is not stimu- lated by boiled pancreas to anything like the extent that the action of liver is accelerated by boiled liver. G. Senter! has shown that the action of the catalase of the blood on hydrogen peroxide at o° follows the equation of a reaction of 1 SENTER : Zeitschrift fiir physikalische Chemie, 1903, xliv, p. 257- 180 A. S. Loevenhart. the first order, but at ordinary temperature the value 0.4343 K con- stantly diminishes as the reaction proceeds. This he attributed very properly to the oxidation or at least destruction of the enzyme. Issa- jew! has recently found that the action of yeast catalase is also a reac- tion of the first order. The results of the experiments reported above could not be expected to yield a constant when applied to the equa- tion of a reaction of the first order, because they were performed at room-temperature, and the reaction proceeded too rapidly under the conditions of the experiment. The value 0.4343 K in the above exper- iments often diminishes 50 per cent as the reaction proceeds. As the reaction has been previously shown to be one of the first order, however, the value 0.4343 K was calculated from the above experi- ments and by averaging the mean velocity for a given: experiment determined. The average values of 0.4343 K in the series, with in- creasing amounts of liver (p. 178), are as follows: c.c. liver acting. 0.4343 K. c.c. liver acting. 0.4343 K. 0.5 0.000109 3.0 0.0081 1.0 0.00021 4.0 0.017 2.0 0.0015 5.0 0.037 If the coefficient of velocity were proportional to the amount of liver acting, the value for 5 c.c. liver extract should be ten times as great as for 0.5 c.c., whereas we find it to be 339.5 times as great. Hence the reaction proceeds 33.95 times as fast when 5 c.c. of liver extract are employed as it would have done if the coefficient of velocity were proportional to the amount of liver acting. The results here obtained offer an explanation of the results of Issajew. Issajew found that increasing the amount of enzyme acting did not cause a proportional increase in the coefficient of velocity, but after dialysing the enzyme solution, an increase in the amount of enzyme acting causes more than a proportional increase in the value of K. Issajew considered, among other explanations, the possibility that during dialysis sub- stances were removed which affected in some way the reaction. This is undoubtedly the correct explanation. Pawlow and Parastschuk 2 have recently shown that by varying the conditions, the velocity of the action of rennin may be proportional to the amount of the enzyme acting, or to its square, or its squareroot. It is, therefore, important, 1 IssAJEW: Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 102. ? PAwLow and PARASTSCHUK: Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 415. Catalytic Decomposition of Flydrogen Peroxide. 181 in studying the velocity of enzyme action, to remember that we never have present merely the enzyme and the substance acted upon, but also other substances which often greatly influence the velocity of the reaction.! In the hope of finding the cause of these remarkable accelerations, it was decided to vary the conditions of the experiments. This was done in the first place by neutralizing the slight acidity of the hydro- gen peroxide. 5 c.c. of the hydrogen peroxide required for neutraliza- tion from I c.c. to 1.5 c.c. of 4 NaOH, and, in the experiments which follow, 5 c.c. of the hydrogen peroxide were placed in the small bottle, previously described, together with the amount of 4’, NaOH required to neutralize it. The experiments previously described with the acid hydrogen peroxide were now repeated with the neutralized solution under the same conditions. The results were as follows: SERIES 9.— Effect of increasing the amount of liver acting on the neutralized hydrogen peroxide. Time in 1. 2. 3. 4. 5. Beconds 1 c.c. liver. 2 c.c. liver. 3 c.c. liver 4 c.c. liver. 5 c.c. liver. : 4 c.c. water. 3 c.c. water. 2 c.c. water. 1 c.c. water. No water. Uc 15:3 21.1 27.0 33.1 10.0 20.5 30.7 37.8 43.7 12.0 24.7 Sil 43.9 48.2 13.8 28.7 42.4 47.4 49.4 15.5 32.9 45.9 48.9 49.5 17-3 36.4 48.2 49.2 49.5 19.1 39.9 49.2 49.2 20.7 | 42.7 49.5 49.2 For purposes of comparison, only the early readings can be used in a reaction proceeding as rapidly as this one, and by examining the first readings in the above tables it will be seen that the rate of oxygen evolution is proportional to the amount of liver acting, whereas with the acid peroxide the velocity of the reaction increased far more - rapidly than the increase in the amount of extract acting would indicate. 1 Since this article was written, BACH has expressed the same opinion (Berichte der deutschen chemischen Gesellschaft, 1904, xxxvii, p. 3785). 182 A. S. Lovenhart. This table shows that with the neutralized hydrogen peroxide an increase in the amount of pancreatic extract acting does not cause SERIES 10. — Effect of increasing the amount of pancreatic extract acting on the neutralized hydrogen peroxide. A 2. 3. Time in 1 c.c. pancreas | 2 c.c. pancreas | 3 c.c. pancreas seconds. extract. extract. extract. 4 c.c. water. 3 c.c. water. 2 c.c. water. 2.0 3.6 ye 2.6 4.8 6.8 Se eS) 8.2 SH7 6.9 9.5 4.2 TES) 12 48 9.0 12.8 5:3 10.1 14.4 5.8 et So: quite a corresponding increase in the rate of oxygen evolution. It was next determined whether the pancreas could accelerate the cataly- sis by liver when the peroxide had been previously neutralized. The results are seen in Series II. A comparison of Columns 3 and 4, and of 7 and 8, shows that whereas the combined action of pancreas and liver acting on acid peroxide causes an acceleration of about 200 per cent, practically no acceleration is noted if the peroxide is neutral. It is perfectly evident from all the facts that the power of the pancreatic extract to acceler- ate the catalysis of acid hydrogen peroxide by liver extract is merely due to the power of the pancreatic extract to neutralize the retarding effect of the acid contained in the commercial peroxide. It is to be noted in this connection that the pancreatic extracts employed were often slightly acid toward litmus, especially the turbid extracts originally employed, and hence it is the more interesting that they can prevent the retarding action of the acid contained in the peroxide. It will be noted on comparing Columns 1 and 5, and also 2 and 6 of the last tables, that while the action of the liver extract was acceler- ated 475 per cent by merely neutralizing the hydrogen peroxide, that of the pancreas was only accelerated from 100 per cent to 200 per cent. The fact can be expressed in another way, namely, the action Catalytic Decomposition of Hydrogen Peroxide. 183 of liver extract is inhibited relatively much more by acids than the action of pancreas extract is. Since both boiled and fresh liver and SERIES 11. AcID HYDROGEN PEROXIDE. 1 2. 3. 4 1 c.c. liver. 1 c.c. pancreas. 1 c.c. liver extract. 1 c.c. water. 1 c.c. water. SOL 2isIe 1 c.c. pancreas. Time in seconds. SERIES 12. NEUTRALIZED HYDROGEN PEROXIDE. 5 6 UP 8 1 c.c. liver. 1 c.c. pancreas. Cumiot Sand Ge l c.c. liver extract. Ie:c. water. 1 c.c. water. 1 c.c. pancreas. Time in seconds. 15 4.1 0.6 4.7 5.6 30 4.8 0.8 5.6 6.2 5.3 0.8 6.1 6.7 EW) 0.9 6.6 (2 6.2 1.0 7.2 dell 6.6 1.0 7.6 8.2 7.0 1.0 8.0 8.7 (as ae 8.6 92 pancreas extracts are able to neutralize the retarding effect of the acid in the commercial peroxide, it is perfectly clear why the action of the 184 A. S. Loevenhart. liver extract on acid hydrogen peroxide: is accelerated more by the simultaneous presence of these extracts than is the pancreas extract. In those experiments in which an increasing amount of liver was used, each increase in the quantity of extract not only increased the amount of enzyme acting, but also improved the conditions for the reaction by neutralizing the effect of the acid present, and thereby causing the great acceleration noted. As already stated, this work was undertaken to check the work on the glycolytic action of mixed tissue extracts. The absence of any specificity and the absence of Cohnheim’s “ Ablenkung ” phenomenon prove that the reaction under consideration is not at all comparable to the glycolytic process. The results which have been reported in the above indicate, however, that great caution must be observed in drawing conclusions from experiments with mixtures of the extracts of different organs. In several cases of enzymic action, the velocity has been found to be greatly increased by mixing extracts of certain tissues with the solution containing the enzyme. It is well known that certain conditions favor the action of certain enzymes, and to conclude from such results that the inactive solution contains a peculiar “ kinase” seems unjustifiable. After treating the enzyme with the kinase, no attempt has been made, so far as I am aware, to remove the kinase, and hence we can- not be sure whether it acts on the enzyme or not. If we could not remove the hydrochloric acid from a pepsin hydrochloric acid mixture, it might be called a kinase. It should be remembered that many of these processes are as sen- sitive to a change of reaction as the indicators which are used to test the reaction, and which are notoriously uncertain in these solu- tions. In the case of the coagulation of the blood, for instance, the accelerating effect of mixing muscle extract with serum has recently given rise to a complicated theory. The whole ‘ kinase” concep- tion, in the opinion of the writer, rests on a very slender and uncer- tain basis of fact. How simple it would be, in the reaction reported in this paper, to conclude from a few experiments that the pancreas contains a ‘‘ catalokinase” or “ coferment.” There is, of course, no denying certain facts which have been brought out as a result of the work on the “ kinases,” but the view that the increased activity is due to a specific enzyme which renders kinetic the dormant enzyme, as the name kinase indicates, is an unwarranted and speculative con- ception resting on but few facts. Catalytic Decomposition of Hydrogen Peroxide. 185 SUMMARY. When unneutralized commercial hydrogen peroxide is used, it is found that: (1) Pancreas extract markedly accelerates the decomposition of the peroxide by liver extract. Muscle extract has the same acceler- ating action. (2) The accelerating property of the pancreas is not destroyed by boiling. (3) Boiled and fresh liver extract greatly accelerate the decom- position of hydrogen peroxide by fresh liver extract. (4) When neutralized hydrogen peroxide is employed, none of these accelerations are noted, and the accelerating action of these extracts is due to the fact that they neutralize the retarding action of the acid contained in the commercial peroxide. Acids inhibit the action of the pancreas extract relatively much less than they do that of liver extract, and hence the action of pancreas extract on the acid hydrogen peroxide is not so greatly accelerated by boiled or fresh liver or pancreas extract. On the other hand, sodium thiosulphate and thiourea accelerate the action of liver extract more than they do that of the pancreas extract. Temperature, however, affects the activity of liver and pancreas alike, and the differences observed are probably due to differences in the environment of the catalase in the liver and pancreas, the enzymes being probably identical. TWITCHINGS OF SKELETAL MUSCLES PRODUCEDREY SALT-SOLUTIONS WITH SPECIAL’ REFERENCE @ TWITCHINGS OF MAMMALIAN MUSCLES.! By WALTER E. GARREY. [From the Physiological Laboratory, Cooper Medical College, San Francisco.] « ROFESSOR J. LOEB and the author tested upon frog’s skeletal muscles, the action of a series of sodium salts, which precipitate calcium. This work was done as the natural sequence of Loeb’s previous work? which demonstrated that the rhythmic twitchings which these muscles exhibit in isotonic sodium-chloride % may be inhibited by very small amounts of calcium chloride. Loeb at that time advanced the view that calcium salts in the blood and tissues normally inhibit muscular twitchings, and that these salts diffuse out when the muscle zs placed in sodium chloride. The normal “ balance” of the inorganic constituents (ions) is thus disturbed. If this pro- cess could be accelerated, ¢.¢., by precipitating or inactivating the calcium salts (or other inhibiting factors), the initiation of twitches should be accelerated and the twitches become more powerful than when simply immersed in sodium chloride. We found,’® as was anticipated, that frog’s skeletal muscles im- mersed in sodium oxalate, phosphate, carbonate, sulphate, fluoride, etc., exhibited powerful contractions which began at once and were decidedly rhythmic in character.* Loeb’s more recent work shows that a number of other salts — notably barium chloride — have an effect similar to calcium precipi- tants in the production of muscular twitchings. The citrate, acetate, and succinate of sodium also have an effect similar to the salts which we employed in the earlier investigations. The author has continued 1 Details of this investigation, the results of which are summarized in this article, will be published later. 2 LoeEB, J.: Festschrift fiir Fick, 1899, p. 101. 8 A complete report of this work, done in 1899, has not yet been published ; cf. LOEB, J.: Decennial publications, University of Chicago, 1go2. 4 See BIEDERMANN: Electrophysiology (WELBy), 1896, p. 104. 186 Twrtchings of Skeletal Muscles. 187 these investigations independently, anda summary of the recent results follows. Muscles of cold-blooded animals. — The twitches of frog’s muscles which have been started in isotonic sodium chloride are markedly increased in force by the addition of small amounts of all the salts mentioned above as producing twitches, and twitches started in the above salts continue longer, with increased force, when the muscles are subsequently transferred to pure sodium-chloride solutions, — as most of the salts employed (e. ¢., sodium oxalate) are violent poisons in the isotonic concentration. The injection of calcium precipitants and “inactivators” in the form of sodium salts, into the lymph spaces, peritoneal cavity, or directly into the circulation of the frog, sets up fibrillary twitchings of the whole musculature. (These experiments confirm the work of Friedenthal.!) These twitchings may be inhibited by the simulta- neous or subsequent injection of calcium salts, and to a lesser degree, by magnesium salts. Magnesium sulphate, although a calcium pre- ‘cipitant, has this inhibiting effect. The action of magnesium is there- fore not unlike that of calcium. ; The twitchings of frog’s muscles are produced by these injections after destruction of the central nervous system, after section of the motor nerves, and after the effective injections of curare. That the twitches are purely myogenic is also shown by the fact that they may be produced after section of the nerves with degeneration of the nerve- fibres and motor end-plates. Mathews? has furthermore shown that the immersion of nerve-fibres in many of the above solutions does not set up contractions of the muscle until after a considerable time; muscles, however, begin twitching at once or after a very short latent period.® The author extended these results obtained on the frog by testing many other muscles without finding an exception to the rule that a skeletal (striated) muscle may be made to twitch and usually with a decided rhythm in the salts mentioned above. Among the muscles 1 FRIEDENTHAL: Archiv fiir Physiologie, got, p. 145. 2? MaTHEws, A. P.: This journal, 1904, xi, p. 455. MATHEWS also employed curare as above. See also BIEDERMANN: Loc. cit. 8 Neurogenic twitchings of frog’s muscles may be produced by exposing the spinal cord and immersing it in certain isotonic salt-solutions, ¢.g., sodium phos- phate and sulphate. These experiments are still in progress, and will be published at a later date. 188 Walter E. Garrey. tested were those from Nereis, Limulus, lobster, crayfish, earthworm, cricket, salamander, lizard, and turtle. Mammalian muscles. — Success also attended the efforts to pro- duce twitchings of mammalian muscles by subjecting them to isotonic solutions of the same salts as were used in the production of twitches of frog’s muscles; special conditions were, however, necessary. The results are the same irrespective of the mammal from which the mus- cles were taken; positive results were obtained with muscles of cat, dog, guinea-pig, rabbit, and with human muscle taken in two cases from amputated limbs and in one case from an excised tumor to which the muscle was adherent. The muscles of pigeons were also found to twitch in certain salt-solutions under the conditions noted below for mammalian muscle. In experiments with the mammalian muscle the tissue was excised from the narcotized animal, and either immersed in the solutions, or the solutions were perfused under varying pressures through blood- vessels of the muscles. The results of the two methods were iden- tical. It was found advantageous in the immersion method to employ thin muscles, ¢. g., platysma myoides and abdominal muscles in which all fibres are more quickly and uniformly acted on by the solutions. None of the solutions tested produced twitches in a mammalian muscle when the temperature was lower than about 30°C. At this temperature twitches may or may not be produced, the results being variable with different animals and with different muscles from the same animal. The optimum temperature for the production of the twitches is about the body-temperature, say between 35 and 4o C. At this temperature simple immersion suffices to induce twitches in isotonic solutions! of the following sodium salts: phosphate, disodium phos- phate, carbonate, bicarbonate, sulphate, oxalate, citrate, succinate, acetate, and tartrate, also in lithium carbonate (saturated) and in lithium sulphate. Many of these solutions are alkaline owing to hydrolysis, and the possibility existed of the hydroxyl ions being the causal factor in the production of the twitches of mammalian muscle. This is not the case,” however, for the twitches occur in neutral sodium sulphate 1 The solutions used depressed the freezing point to —0.6° C. Equimolecular solutions (7) proved equally effective. 2 MATHEWS, A. P.: This journal, 1904, xi, p. 478, shows that the stimulation of nerves by a similar series of salts is not due to OH-ions. Twitchings of Skeletal Muscles. 189 and even in sodium sulphate after the addition of a trace of sulphuric acid; furthermore the twitches do of occur in. sodium chloride to which sodium hydrate has been added, —no matter what the con- centration of the alkali may be. It is true, however, that a small degree of alkalinity facilitates the genesis of the twitches in these solutions while acid has the opposite effect. In pure barium-chloride solutions of all strengths, the muscles promptly die without twitching. Neither did the muscles twitch in magnesium chloride or in magnesium sulphate. In isotonic sodium chloride twitches could not be induced under ordinary conditions, either by changes in temperature or by the addi- tion of alkali, as noted above. The negative results with isotonic sodium chloride were not antici- pated, for my other results had been quite in harmony with those previously obtained with frog’s muscles. I decided therefore to fol- low out suggestions to be found in the literature relating to contrac- tions of isolated strips of mammalian ventricular muscle, and to subject the skeletal muscles in sodium chloride solutions to an atmos- phere of pure oxygen; but the experiments did not succeed; simple saturation of isotonic sodium chloride, and an atmosphere of pure oxygen, did not cause a mammalian skeletal muscle to twitch spon- taneously. Neither did I succeed after oxygenation of the sodium chloride with dioxide of hydrogen, as in Lingle’s experiments,’ with strips of tortoise heart; nor did these dioxide of hydrogen-experi- ments succeed after neutralizing the solutions or making them alkaline. Porter, in his heart experiments,” used pure oxygen under pressure, and following this lead it was easily demonstrated that intermittent twitches are instituted in isotonic sodium chloride when subjected to oxygen ata pressure of two to three atmospheres, when the temperature is approximately 40°C. These twitches continue from one half to three fourths of an hour; they cease, however, long before the muscle loses its irritability to induction-shocks. These experiments do not succeed at 20° C. The twitches in sodium chloride under oxygen pressure ‘begin sooner, are stronger, and last longer, if a very small amount of sodium hydrate is present. Similarly the twitches are more pronounced if from 1-5 c.c. of isotonic barium chloride be added to each 100c.c. of sodium chloride ; 1 LINGLE: This journal, 1902, viii, p. 79. 2 PorTER: This journal, 1898, i, p. 516. 190 Walter E. Garrey. the poisonous action of barium chloride becomes evident in the rapid death of the tissues. when larger amounts are added. The addition of small amounts of any of those salts which, under ordinary conditions induce muscular twitchings, starts the twitch- ing sooner, or increases the force of those already started in sodium chloride under oxygen pressure. Under these conditions the addition of very small amounts of cal- cium chloride or nitrate inhibits the twitching at once, and the same effect is produced, although in a lesser degree, by magnesium chloride and sulphate. Oxygen pressure will not elicit contractions of a mammalian muscle immersed in its own blood-serum or defibrinated blood. Neither would the muscles twitch in Ringer’s solution, nor in diluted sea-water isotonic with the blood. No twitches in isotonic non-electrolytes were ever noticed. Dex- trose, levulose, cane sugar, lactose, urea, and glycerin were tried. An increase in the osmotic pressure of these non-electrolytes did not induce the twitchings; ! the muscles simply contracted tonically, and quickly lost their irritability. Increasing the osmotic pressure of pure sodium chloride to 2 7 occasionally called forth a few initiatory, intermittent twitches which lacked any semblance of rhythm, and the result was not at all certain. This increase in the osmotic pressure of the sodium chloride caused a rapid death of the mammalian muscle. More forceful contractions follow an increase in the osmotic pressure of pure sodium chloride when there is a high oxygen pressure, but the contractions continue for a short time only. The mammalian muscle lives longer in isotonic sea-water than in any other inorganic solution tested, ¢..g., after an immersion of one hour the muscles will often twitch in calcium precipitants or in sodium chloride under oxygen pressure, — it is therefore an excellent solution with which to test the effects of increased osmotic pressure; it was found, however, that no change in the osmotic pressure of sea-water will suffice to induce the twitchings of mammalian muscles immersed in it. In pure sea-water (A = — 1.9 C.) mammalian muscles die within thirty minutes. , At best the contractions of mammalian skeletal muscles, produced by the means employed in this investigation, are feeble and lack much 1 An occasional exception to this result was noted, ¢. g., in one instance, a strip of muscle beat for ten minutes in 2.5 mol. solution of cane sugar. Twitchings of the Skeletal Muscles. 191 of the decided power, regularity, and rhythmicity shown by the con- tractions of frog’s muscles under similar conditions. This fact, and the extreme susceptibility of the mammalian tissue to the poisonous action of the salts employed, made it impossible to determine with any certainty the comparative stimulating power of the various salts. THE ROLE OF CERTAIN IONS IN RHYTHMIC HEART ACTIVA: By STANLEY KK. BENEDICH- Contribution from the Biological Laboratory of the University of Cincinnati. 8 yy ry S| eee question of the rhythmicality of heart-muscle is one which has justly received much attention. Ever since the demon- stration of a relationship between heart-beat and the inorganic salts of the blood, this problem has been an especially prominent one. In view of the fact that this paper is preliminary in nature, it does not seem desirable to give a résumé here of what has already been done upon this subject. For an account of such work I refer the reader to the papers of Green! and Howell,? where will be found a résumé of results obtained, as well as a very complete bibliography of this subject. Results heretofore obtained having a bearing upon our present discussion will be considered individually. The most recent aspect of the problem is one developed by Howell? and Lingle. Howell has vigorously supported the theory that calcium ions are of prime importance in heart-beat; while Lingle, influenced, no doubt, by Loeb’s results with striped muscle, has as strongly championed the greater importance of the sodium ion. Martin,‘ after two years’ work upon the subject, comes to the conclusion that both calcium and sodium ions are necessary, if rhythmic activity is to be developed or maintained, —a position already taken by Howell.® The present work was originally commenced with the idea of supporting Howell’s calcium theory, since some of Lingle’s® con- clusions from his experiments seemed to be unwarranted. 1 GREEN: This journal, 1898, ii, p. 82. 2 HowELL: This journal, 1898, ii, p. 47; /ézd., 1901, vi. p. 181. 3 LINGLE: This journal, 1900, iv, p. 265; /dzd., 1902, vill, p. 75- 4 MarTIN: This journal, 1904, xi, No. 2. 5 HOWELL: This journal, 1go1, vi. § LINGLE: Loc. cit. 192 Role of Certain Lons in Rhythmic Heart Activity 193 The strips used were prepared from the turtle’s ventricle, as described by Green.1. Three or four strips were usually prepared from one heart, in order to obtain controls for each experiment. The salts used were of the highest purity obtainable, being Kahlbaum’s best manufacture. The line of investigation determined upon was to use different salts (z. ¢., other than the chlorides) of the metals found in the blood. Before many experiments had been performed, the writer was forced to a theory differing widely from either Howell’s or Lingle’s, in order to explain the facts observed. The purpose of the present paper is the proposal and support of this theory, which may be formulated as follows: the direct pro- duction of rhythmic activity by means of a salt’s action upon heart- muscle is due to the anion of that salt, while the chief function of the cation is apparently to maintain such a tone of the heart-muscle that it will respond to the stimulus furnished by the anion. This theory had partially suggested itself to the writer prior to the commencement of experimental work, upon reading the following statement of Lingle’s: ? “If sodium chloride is so closely associated with the real stimulus, the sodium and chlorine ions of which it is composed must now be considered. Loeb, working on striped muscle and Gonionemus tissue, found that the sodium ion was the active agent. In heart-tissue the same seems to be true; for solutions containing sodium, but no chlorine, are able to do the work of sodium chloride, and indeed can do its work even better. Among such solutions is sodium bromide. A sodium-bromide solution equi- molecular with a 0.7 per cent sodium-chloride solution, if properly used, will start rhythmic beats in heart-strips, the beats so started are generally stronger than those in sodium chloride, and moreover the rhythm lasts longer.” To the writer this statement seemed to furnish evidence in an exactly opposite direction from that for which Lingle offered it. Had the action obtained with the two salts been practically identical, Lingle’s statement might indicate the importance of the sodium ion. If, however, another salt of sodium, namely, the bromide, can do the work of sodium chloride, and do this work distinctly detfer than can the chloride, we are certainly forced to believe that the rdle of the anion is not entirely a passive one, as Lingle ® would assume, but 1 GREEN: Loc. cit. 2 LINGLE: This journal, 1900, iv, p. 272. 8 LINGLE: Loc. cit 194 Stanley R. Benedict. that either the chlorine ion is hurtful, and the bromine ion less so (which we need not consider), or we must assume that the chlorine ion has an active rdle which the bromine ion can perform even better. As above mentioned, the results obtained with different salt solutions (the first used was sodium-carbonate solution, with an idea of precipitating the calcium ions) soon forced the importance of the anion into prominence. Upon applying this theory to results here- tofore obtained by others, it seemed not only warranted but demanded by many of them. These will be considered below. The experimental work done to test the validity of the statement that it is the anion which furnishes the stimulus for heart-beat was as follows. First, the correctness of Lingle’s statement that sodium bromide produces a better rhythm than does sodium chloride was ascertained by making comparative experiments with equimolecular solutions of these two salts. The results most amply bore out his assertion. Strips from the same heart immersed in sodium-bromide solution, equinormal with this, almost invariably showed a latent period shorter by ten or fifteen minutes in the latter solution, while the series of beats obtained in this solution usually lasted longer than that obtained in the sodium chloride, and the rhythm was better sus- tained. Clearly the anions chlorine and bromine are not entirely inactive, but, on the contrary, play some positive rdle in the phe- nomena of heart-beats. An interesting result obtained in some instances (when the sodium- chloride series was unusually short) was a short renewal of the series of beats in the sodium-bromide solution after exhaustion in sodium chloride. This result also points strongly to the importance of the anion. The action of equimolecular solution of calcium chloride and calcium bromide was next compared. The calcium-chloride solution was 0.026 percent. If bromine ions are more stimulating than chlo- rine ions, we might expect to find a difference here. The results obtained showed a decided difference. When immersed in pure calcium-chloride solution (0.026 per cent), ventricular strips usually pass almost immediately into a greatly increased tone, from which recovery will usually not take place upon subsequent immersion in sodium-chloride solution. This change of tone in calcium-chloride solution is usually (not always) accompanied by a short series of irregular, powerful contractions. Role of Certain lons in Rhythmic Heart Activity. 195 In calcium-bromide solution, equimolecular with 0.026 per cent so- lution of calcium chloride, the strip also shows the change of tone; but this change is practically always accompanied by a series of beats last- ing from ten minutes to half an hour, — when, as we should expect, the muscle passes into the so-called calcium rigor. The beats obtained in the bromide solution begin after a shorter latent period, are more regu- lar, and last longer than those obtained in the calcium-chloride solution. In a few instances a series of beats somewhat resembling a sodium- chloride series has been obtained in the calcium-bromide solution, 7. ¢., the beats began at a minimum, and gradually increased in intensity, the series occupying over forty-five minutes. The controls in cal- cium chloride did not show this long, graduated series, nor have I ever been able to obtain it in this latter solution. The difference in the action of the chlorides and bromides of so- dium and calcium can be accounted for only by bestowing upon the anion an active rdle hitherto not ascribed to it. That the function of direct stimulation to rhythmic beats belongs to the anion and not to the cation, will, it is believed, be conclusively shown by considering the behavior of ventricular strips in sodium-carbonate solution (equi- normal with 0.7 per cent sodium chloride). When the action of sodium-carbonate solution was first determined, and the results to be described were obtained, it was not known to the writer that this solution had been previously tried, even in com- bination, upon heart-strips. I have since found, however, that Gaule? in an article published in 1878, states that a ventricular strip immersed in sodium-chloride solution made alkaline with sodium carbonate, after exhaustion in sodium-chloride solution, will give a long series of beats. Martius? also obtained this result, and offered the explanation that it was due to the fact that the alkaline solution absorbed the car- bon dioxide given off by the heart. Since neither Gaule! nor Martius further interpreted this fact, nor investigated the action of sodium-carbonate solution alone fully, I shall describe results obtained with this solution, and conclusions from these at some length, in order to make plain its bearing upon the theory advocated in this paper. It may be mentioned here that neither Howell? nor Lingle,* in summing up the ways in which a strip may be made to beat after ex- 1 Gauce: Archiv fiir Physiologie, 1878, p. 291. * Martius: Archiv fiir Physiologie, 1882, p. 543. PELOWELL : Loc. ct. 4 LINGLE: Loe. cit. 196 Stanley R. Benedict. haustion in sodium chloride, mentions the immersion in an alkaline solution of sodium chloride. Yet this fact has undoubtedly a most important bearing upon their respective theories, which bearing we | shall discuss below. Indeed, the importance of the action of sodium- carbonate solutions will be recognized when it is seen that the behavior of a strip in this one solution would make necessary a recon- struction of all our theories as to the action of salts upon heart-muscle, and further give us a most important clue as to the cause of the sodium-chloride arrest. Upon immersing a fresh strip of turtle’s ventricle in a solution of sodium carbonate (equinormal with 0.7 per cent sodium-chloride so- lution) no series of beats develops. Consider Lingle’s theory that sodium ions are the one essential in the origination of heart-beats. Here are sodium ions, yet no beats develop. Lingle’s only reply could be that the sodium-carbonate solution is somewhat toxic, and therefore the sodium ions cannot have their normal effect. Howell might say that beats are prevented in the so- dium-carbonate solution because the carbonate ions present precipitate the dissociable calcium as calcium carbonate. Yet both of these pos- sible objections are entirely overcome when we consider the following fact. If after exhaustion in sodium-chloride solution the strip be im- mersed in sodium-carbonate solution, a series of beats always begins, usually within five minutes, which lasts often for a great length of time, from ten to twenty hours being not infrequent for the duration of this series. I may give here one extract from my note-book to exemplify this interesting result. May 4, 1904.— At 2.55 P.M. ventricular strips (of the same heart) were immersed in solution of RESULT. (1) Sodium chloride Beats in fifty minutes. (2) Sodium bromide Beats in thirty-five minutes. (3) Sodium carbonate No beats after three hours’ immersion. After leaving strip (2) in the sodium bromide over night and until 10.30 A. M., May 5, it was immersed in sodium-carbonate solution. Result. — Beats began in three minutes, and lasted over twenty hours. Let us consider some of the conclusions demanded by the above facts, keeping in mind that the sodium-carbonate solution does not fail to produce beats because toxic, or it would not produce and main- Role of Certain Llons in Rhythmic Heart Activity. 197 tain them after exhaustion in sodium chloride, for the same reason it does not fail to produce them because it precipitates calcium. We come now to face the following facts: (1) Sodium-chloride solution produces beats in a fresh strip, but cannot maintain them longer than a given period. (2) Sodium carbonate cannot produce beats in a fresh strip, but can produce and maintain them after previous exhaustion in a sodium- chloride (or bromide) solution. The action of these two salts is diametrically opposite, yet the ca- tion is the same in both. The only difference between them lies in the different anions, hence the opposite effects can be accounted for only by assuming that the anion plays a most essential part in the action of the salt. Wecan see no escape from this conclusion. These facts have a further implication. They must affect our the- ories as to the cause of the sodium-chloride arrest. The problem of this arrest is one which has not as yet been satisfactorily solved. Loeb has suggested it as due to the poisonous effect of pure sodium solu- tions. Howell! has justly pointed out that there is not sufficient evi- dence to justify this position. The renewal of beats in a pure sodium- carbonate solution makes this position completely untenable, unless we are to suppose that another anion can neutralize this poisonous effect, which would seem very improbable, and, further, very diff- cult of demonstration. It might be remarked here that the term poisonous as applied to the action of pure sodium solutions would seem not wholly justified, even if beats could only take place in pres- ence of other cations. Because an animal dies when fed only carbohy- drates, we should not be justified in saying that pure carbohydrates are poisonous, and that proteids are needed to neutralize this poison- ous effect. Howell has suggested that the sodium-chloride arrest takes place because nearly all the diffusible calcium has left the strip, yet we know that renewal takes place in solution of lithium chloride, dex- trose, hydrogen peroxide, and Ringer’s mixture. Three of these contain no calcium ions. Further, as above stated, renewal also takes place in pure sodium-carbonate solutions; this solution not only contains no calcium ions, but, on the contrary, would precipitate these ions as the carbonate, were any present. The theory that the sodium-chloride arrest is primarily due to a lack of calcium would, therefore, not seem supported by the facts. 1 HowELL: This journal, rgo!, vi, p. 195. 198 Stanley R. Benedict. Lingle,! offering another explanation of the sodium-chloride arrest, says: “In this case” (referring to the fact that beats are renewed by oxygen gas after sodium-chloride exhaustion) ‘recovery occurs without any diffusion of salts, which indicates clearly that the ordinary sodium-chloride arrest is largely due to a lack of oxygen.” Later in the same article, Lingle? apparently returns to Loeb’s position that the poisonous effect of pure sodium solutions is chiefly responsible for the sodium-chloride arrest, and again, in his summary, says that this arrest is due to a lack of oxygen. The reasoning by which Lingle arrives at this conclusion scarcely seems justifiable, for, by exactly analogous reasoning, we should say that this arrest is due to a lack of dextrose, to a lack of calcium chloride, or of any other substance nor- mally found in the blood which will renew beats after the sodium- chloride arrest. We may be able to come to some satisfactory conclusion regarding this arrest, if we examine it from another point of view. First, we no- tice that none of the agents which will cause beats after the sodium- chloride exhaustion will cause a series of contractions before this exhaustion. Clearly the muscle must be in a different condition after sodium chloride has acted upon it than before. What difference, we now ask, is most pronounced? The answer is, the tone of the muscle has undergone a distinct change; the recording lever invari- ably shows a slow continuous loss of tone after immersion in this solution. We come now to examine the agents which will cause a renewal of the beats after the sodium-chloride exhaustion. Howell? sums them up as follows: I. Immersion in a Ringer’s mixture. 2. Immersion in a mixture of sodium chloride and calcium chloride. 3. Immersion in a solution of dextrose or cane-sugar. 4. Immersion in lithium chloride. To these Lingle adds: 5. Immersion in oxygen gas or addition of hydrogen peroxide to the solution. We might add here as auxiliary to 5, the method of immersion in moist air. This method was mentioned by Lingle,* and corroborated at some length by Martin ® in 1903. 1 LINGLE: This journal, 1902, viii, p. 83. 2 LINGLE: /did., p. 97. 3 HOWELL: Loc. cit. 4 LINGLE: This journal, 1902, viii, p. 83. 5 MARTIN: Loc. cit., p. 124. « Réle of Certain Lons in Rhythmic Heart Activity. 199 Since these agents differ so widely chemically and physically, some electrolytes, one a non-electrolyte and one a gas, we cannot find the direct cause in the chemical or physical nature of the substance itself. Yet we are able to find a common feature in the action of all these substances upon the heart-strip. This common feature is to cause a more or less marked increase in tone. The rapidity with which a strip recovers in these various agents is more or less directly pro- portional to the strength and rapidity of its action as a tone-increaser. Sodium carbonate is a sixth agent by which beats may be renewed after exhaustion in sodium chloride. Does this solution also cause an increase of tone in heart-muscle? We answer that a recording lever shows that it does so very rapidly just before the series of con- traction begins. Whether the cause of this increase in tone is due to the carbonate ions or the hydroxyl ions, both of which are of course present, has not been determined. In view of Zoethout’s! work upon skeletal muscle, it seems very likely that the hydroxy] ion is the active agent, since this causes a marked increase of tone in skeletal-muscle. The carbonates have practically the same effect; so that it is very possible that the hydroxyl ion is the cause of the change of tone in both solutions. When we remember that a loss of tone always occurs before the sodium-chloride exhaustion, and that the various agents which cause a renewal of beats are tone-increasers, it hardly seems possible that this theory of the cause of the sodium-chloride exhaustion is incorrect. It may be worth while to consider the action of solutions of lithium chloride and dextrose for a moment, since these solutions almost invariably permit of a slight loss of tone before any increase in tone takes place. What might appear at first glance as a weakness in our theory, will, upon consideration, prove to be an additional support to it. Green? says that a short irregular series of beats is sometimes obtained in dextrose. This takes place before the dextrose begins to have its effect of increasing the tone of the strip. The anion chlorine is present in large amount in the tissue. Lingle * says that a series of weak beats lasting a short perieds is sometimes obtained in mixtures of lithium chloride and oxalate, Again this is what we should expect: at first, loss of tone takes place, due possibly to the precipitation of calcium by oxalate ions, 1 ZOETHOUT: This journal, x, No. viii, p. 373. 2 GREEN: Loc. cit. 8 LINGLE: Loc. cit. 200 Stanley R. Benedict. and beats begin. Further, in substantiation of our position, we may state that lithium solutions, which are weakest as tone-increasers, are the most difficult in which to get a renewal after sodium-chloride exhaustion; dextrose, which is a little more powerful in this respect, comes next; whereas a mixture of sodium chloride and calcium chlo- ride, which is most powerful as an increaser of tone, causes a most rapid recovery after the sodium-chloride arrest. Consider whether this theory of the sodium-chloride exhaustion agrees with the history of a strip. In the first place, the operative procedure throws the strip into an increased tone; now only those solutions originate beats in such a strip which permit of a loss of tone (or those which, like calcium, make the muscle exceedingly irritable). Sodium-chloride solution, for example, permits, indeed causes, a loss of tone, and this solution also causes a series of beats. But the action of sodium-chloride solutions is very mild in this respect; hence the long latent period, followed by the long series of beats, which gradually decrease until the tone becomes, so to speak, just below the beating value. -If the strip is now changed to a solu- tion which is not toxic, and if the tone be increased by this solution, beats again commence, since the anion chlorine is present in the tissue in large amount. The cation may (as in sodium carbonate) be the same in both solutions, or it may differ (as in calcium-chloride solution). One objection which may be urged against this position is as fol- lows: calcium-chloride solution causes an increase in tone, yet wash- ing with this solution previous to immersion in sodium-chloride solution usually causes a much shorter latent period. The answer to this objection is, it seems to me, as follows: calcium ions, without doubt, greatly increase the irritability of heart-tissue. Whether they do this simply by causing an increase in tone is yet to be determined. From the fact that calcium will cause an increase in irritability, it is only natural to expect that the strip first washed in it will respond more quickly to whatever stimuli it comes in contact with. It must be remembered that if the calcium be not washed off, or if more than a very small per cent of it be present in the subsequent bathing solu- tion, beats will never develop. The writer’s idea is not that the only function of the cation is to maintain tone; it seems very probable indeed that it also has some other functions which will be made the subject of future investigation. A second objection which may be urged against the above position Role of Certain Lons tn Rhythmic Heart Activity. 201 is as follows: it sometimes happens that strips after exhaustion in sodium chloride will renew the series of beats in sugar solutions or moist air without visible alteration in tone. In answer to this objection the following may be stated: First, the strip is in a tone-increasing medium, 7.¢., one which in most instances increases tone, and it is very possible that a very slight increase in tone has taken place, enough to permit of a series of beats. In the second place, we should remember that the above-cited in- ‘stances are the exception rather than the rule, and exceptions to the normal behavior of heart-strips are frequent and marked. Thus I have seen a fresh heart-strip give an excellent series of beats (very shortly after immersion) in a solution of potassium iodide equinormal with a 0.7 per cent solution of sodium chloride. This result is very rarely obtained. Further, against this objection, we may state that the dest series of beats after the sodium-chloride arrest is almost invariably obtained only after an zzcrease in tone has taken place. We have seen thus far that different sodium solutions have dif- ferent powers of stimulating heart-muscle. It would seem only natural to try solutions containing still other anions to note the effects of each. Among such solutions tried were the following. The solu- tions used were all equinormal with 0.7 per cent sodium-chloride solution. The results from the fifteen experiments made with sodium iodide have been very contradictory; a longer series of experiments will be necessary to determine the effect of this sait. There can be no doubt that the iodine ion has an effect produced by neither the chlorine or bromine ion. Thus in about seven of the experiments no beats at all developed in the sodium-iodide solution, though beats did develop normally in controls in sodium chloride or sodium bromide. Further, a good series of beats was obtained upon changing the strips from the iodide solution to a sodium-chloride solution (after two and one-half hours immersion in the former solution) with a very short latent period (again the prime impor- tance of the anion is exhibited). In three experiments made with the sodium-iodide solution a fine series of powerful, regular beats was obtained with a latent period of only eight to ten minutes. The series lasted about one and one-half to two hours. The controls in sodium chloride showed an unusually short latent period also, so that it is probable that these strips were unusually irritable. In the re- maining five experiments a series of beats was obtained practically 202 Stanley R. Benedcct. the same as an ordinary sodium-chloride series. Thus it will be seen that the results obtained so far with sodium iodide are by no means conclusive. Sodium-nitrate solution usually produces a series very similar to that obtained in sodium chloride. The latent period in the nitrate solution is usually somewhat longer, and the series of beats somewhat shorter than in the sodium-chloride controls. In many cases a short series of beats may be obtained by immersing a strip which has ceased to beat in sodium nitrate into a sodium-chloride solution. Among other sodium salts tried were the sulphate, acetate, tartrate, and chlorate. The sulphate gives a good series of beats. The acetate and tartrate series resemble closely the one obtained with sodium nitrate. In most instances a renewal of beats can be obtained for a longer or shorter period by immersing the strip in sodium chloride or sodium bromide after it has ceased to beat in the other sodium solutions. But a few experiments have been made with sodium-chlorate solution. The results obtained were, however, very interesting. The latent period was much longer (an hour or more) than in the controls in sodium chloride, while the series of beats which was obtained was very weak and: lasted only from one to two hours. These different results with different solutions can only be ex- plained by the supposition that the anion is of prime importance in the production of heart-rhythm. The results with sodium carbonate alone would seem to the writer conclusive. Another result of Lingle’s! may be mentioned which strongly supports the theory herein advocated. He finds that sodium-oxalate solution alone is unable to start beats in ventricular strips, yet a mixture of sodium chloride and sodium oxalate will originate a series of beats which, while not as strong as the series obtained with sodium chloride alone, may continue two hours or more. Lingle appears to have entirely overlooked the importance of his own statement. It strongly supports the theory that it is the anion which is really responsible for the rhythmic activity. The rhythmic activity of the cesophagus has been said to be analogous in some respects to that of the heart. It may therefore be of interest here to mention some results obtained by Stiles* in his 1 LINGLE: This journal, 1902, viii, p. 88. 2 STILES: This journal, 1901, v, p. 338. Role of Certain Lons in Rhythmic Heart Activity. 203 work with various salt solutions in their action upon the rhythmic activity of the cesophagus. He found that the cesophagus will not exhibit rhythmic activity in pure sodium-chloride solutions. An excellent rhythmic series is, how- ever, always obtained by treatment with a Ringer’s mixture. Among other things he found that there were four salts capable of replacing sodium chloride completely in the Ringer’s mixture. These salts were sodium nitrate, sodium bromide, sodium iodide, and sodium chlorate. He also found that some other sodium salts, among which are the tartrate, acetate, butyrate, and lactate, are capable of replacing a cer- tain per cent of sodium chloride in the Ringer’s mixture, yet if this substitution be pushed further, no rhythmic contractions were pro- duced by the solution. These results are directly in line with the theory that it is the anion which stimulates, since different sodium solutions produce very different results. An objection which may be raised here is that a much greater difference in effects produced by solutions is obtained by changing the cation and keeping the anion the same. We answer that this is exactly what we should expect. Many cations, especially potassium and calcium have marked and immediate effects upon the tone of the muscle. It is a familiar fact that the irritability of a muscle is greatly influenced by its condition of tone; consider, for instance, the example of ordinary skeletal muscle passing into rigor caloris, when stimuli often produce five or six contractions which are entirely unable, under normal conditions, to produce even one. What then is more natural than that the alteration of the tone of the muscle, by changing the cation, should greatly change, or even entirely alter, its power of reacting to the stimulus furnished by the anion. The fact that some sodium solutions can produce beats, while others, which do not precipitate any cation and are not toxic, are not able to do so, can, we believe, be explained only on the assumption that it is the anion which stimulates. As to the means by which the anion stimulates, no suggestion can yet be offered. It will be noticed that the theory herein presented coincides (with the exception of the function of the cation) with the one worked out by Mathews! concerning the action of salts upon nerve-tissue. The writer desires to state that he had formulated his theory, and done much of the work upon it, some time prior to the publication of 1 MATHEWS: This journal, 1904, xi, No. v. 204 Stanley R. Benedict. Dr. Mathews’ paper. The fact that very similar conclusions were arrived at in two different fields, by investigators unacquainted with each other’s work, would seem to lend additional strength to the theory itself. In conclusion, it may be said that this paper is intended in the nature of a preliminary communication. While the author regards the evidence offered as necessitating the theory herein advocated, there is much work still to be done upon this problem. Work is being continued to determine the action of other salts in various strengths upon heart-strips, and to test whether a modification of this theory may be applied to ordinary skeletal muscle. My sincere thanks are due to Professor Michael F. Guyer, whose continued encouragement and suggestions during this work have been invaluable. CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E. L. MARK, Director. No. 161. THE EFFECT OF PIGMENT-MIGRATION ON THE PHOTOTROPISM OF GAMMARUS ANNULADUS* Sol SMITH: By ‘GRANT SMITE: I. INTRODUCTION. HE migration of the retinal pigment in crustaceans has been known for some time to be a process by which the amount of light reaching the recipient elements can be controlled. Since a number of animals change the sense of their phototropism with a change of the intensity of the light — becoming negative in bright light and positive in dim light —it is conceivable that, in a crusta- cean, a change in phototropism may accompany the migration of the retinal pigment. In testing this question I undertook experiments on Gammarus annulatus S. I. Smith, a common salt-water amphipod of the New England coast. The animals were collected at Lynn Beach, Mass., and were kept in the laboratory at Cambridge in sea-water at a temperature of 12° C. In their natural habitat, their actions do not Seem to be as rigidly controlled by light as those of some other animals are. As is well known, they are highly thigmotropic and hence may be collected in large numbers from the under sides of loose stones or pieces of board along the seashore at low tide. To the casual observer they would thus appear to be negatively photo- tropic. But they can also be found abundantly either swimming back and forth across little pools or in gently running water. At such times, when they come to rest, they do not always hide under- neath protecting objects, but are very often to be found resting in the bright sunshine. They are in no sense nocturnal; they feed and move about freely in the bright sunlight as though it controlled them no more than it does the higher animals. It should also be recorded, 205 206 Grant Smtth. as having a possible bearing on the question of their responses, that the animals were collected during the spring when they were actively breeding. All the females used in these experiments were either | copulating or carrying eggs. II. RETINAL PIGMENT-MIGRATION. The retinal pigment-migration in Gammarus annulatus was found to be so like that in G. ornatus as described by Parker (99, pp. 144- 146) that the following statement quoted from the account by him applies equally well to the species at hand. LIGHT. con. cl. tied. ek ee rhe. eons _ q ele él. ace. a ql. acc. mb. ba. nd. rind. 1. for.n Fa oem efor as ‘a Ne clorin: ay: <2 ‘ ! 4 QE 2. a, Sit. 4 a OMMATIDIA OF GAMMARUS ORNATUS AFTER PARKER (99). FIGURE 1.— Longitudinal section of an ommatidium, showing the arrangement of pigment due to exposure to bright light. FiGuRE 2.— Transverse section of a retinula from such a preparation as that shown in Fig. 1, and taken at the level marked 7d. in that figure. FicurE 3,— Longitudinal section of an ommatidium, showing the arrangement of pigment due to the absence of light. Ficure 4.— Transverse section of a retinula from such a preparation as that shown in Fig. 3, and taken near the level marked rd. in Fig. 1. Abbreviations: c/. crm., corneal hypodermis; c/. rtn/., retinular cell; c/. acc., accessory pigment-cell ; coz, cone; crz., corneal cuticula; for. #., retinal nerve-fibre; 2. d2., basement mem- brane; #/. rtn/., nucleus of retinular cell; #/. acc., nucleus of accessory pigment-cell ; rhb., rnhabdome. “Tn an eye of G. ornatus that had been subjected to light for some six hours (Fig. 1), a considerable amount of black pigment Pigment-Migration on the Phototropism of Gammarus. 207 was found uniformly distributed through the distal and middle portions of the retinular cells, thus sheathing the cone and rhab- dome laterally (Fig. 2). The proximal portion of each cell contained a few irregularly scattered pigment-granules except near the nucleus, where the pigment was more abundant. “In an eye from an animal kept some six hours in the dark (Fig. 3), the pigment in the distal portions of the retinular cells presented the same conditions as in the eyes exposed to light. The middle portions, however, were entirely devoid of pigment, while the proximal portions were as densely filled with pigment as the distal portions. “Obviously the changes induced by the presence or absence of light affect the pigment of only the middle and proximal parts. When an animal that has been kept in the dark is exposed to light, the pigment that is massed in the proximal parts of the retinular cells (Fig. 3) migrates distally and fills the middle portions, with- out, however, entirely abandoning the proximal parts, especially around the nucleus (Fig. 1). When an animal that has been kept in the light is placed in the dark, the pigment in the middle portions (Fig. 1) migrates into the proximal parts till almost no pigment is left in the middle portions. In other words, the presence of light induces a distal migration of much of the pigment from the proximal parts, and the absence of light brings about a proximal migration of almost all of the pigment in the middle parts.” The rate at which this migration took place was not determined by Parker, and I therefore undertook to ascertain by experiment what it is. Accordingly, live specimens of G. annulatus were placed in the dark-box over night. The next day several of them were poured into a small net and killed immediately by dipping them for a moment into water at 80° C., after which preparations of their eyes were made by further treating, embedding, and cutting them in paraffin. The remaining animals in the dark-box were exposed to the light under the same conditions as those which were subsequently used in the experiments. After they had been subjected to light for fifteen minutes several were removed and killed. Others were treated similarly after thirty minutes, forty-five minutes, and sixty minutes respectively. There were thus at hand for comparison, “dark” eyes, “light” eyes, and eyes in which the pigment was presumably in an intermediate position. After an exposure to light for fifteen minutes the parts of the 208 Grant Smith. retinular cells free from pigment in the “dark” eye were occupied by many scattered granules, which increased in numbers in the later intervals until the “light ” condition of the eye was reached in about an hour. It is noteworthy that the migration of the pigment is very considerable in the first fifteen minutes. No effort was made to learn how soon pigment-migration began, but it would probably be possible to recognize intermediate stages at five and ten minutes after exposure to light. The rate of migration as determined for the eye of G. annulatus agrees fairly well with what Parker (97) found for the proximal pigment-cells of the more complex eye of Palamon- etes vulgaris. In this species the migration was accomplished in one-half to three-quarters of an hour; while in G. annulatus a full hour is necessary. The way in which the facts of migration bear on the question of the stimulation of the retina can be appreciated from the following quotation from Parker (’99, p. 146): “The axial light which . . . finds its way into the rhabdome must have a certain degree of intensity in order to stimulate that organ. Ordinary daylight is presumably more than sufficient to call forth this stimulation, and such superfluous light as may pass to the edges of the rhabdome or through it is probably absorbed by the black pig- ment that in bright light (Figs. 1, 2) surrounds that body. In dim light, however, there must be times when the light which enters the rhabdome is scarcely intense enough to stimulate that organ. Under such circumstances the more oblique rays, which ordinarily would be absorbed by the black pigment on the sides of the rhabdome, would materially aid in stimulating it if they were turned back into the rhabdome. ' That these rays are probably thus turned back is shown by the fact that in dim light the black pigment is removed from the rhabdome and the surrounding whitish reflecting pigment of the accessory pigment-cells is exposed (compare Figs. 2, 4).” If this view of the action of the parts of the eye is correct, an animal taken directly from the dark, and with the reflecting layer of the retina exposed, ought to react to light differently from one which had been in the light some time and in which the reflecting layer had consequently become covered. Following a general rule of phototropism, namely, that animals are negative in strong light and positive in weak, we should expect a specimen of G. annulatus taken directly from the dark to be negative, or at least less positive than one which had been some time in the light. To ascertain whether this supposition is true, the following experiments were tried. Pigment-Migration on the Phototropism of Gammarus. 209 III. PHoTorropism. I. Apparatus. — All the observations on phototropism were car- ried on in a large dark-room in the Zodlogical Laboratory of Harvard University. The source of light throughout was a large, so-called one-hundred-candle-power, incandescent, electric light. With the exception of an occasional, brief change in the intensity, due to the shifting of machinery at the power-house, this light was steady and reliable. Moreover its intensity could be accurately measured, a matter of great importance, in studies in phototropism which aim to be accurate. By the use of a Lummer-Brodhun photometer and a standardized candle it was ascertained that the intensity of this light after a heat-screen had been placed in front of it was 52.7 candle- power. The lamp was enclosed in a black sheet-iron chamber open on one side. The heat-screen consisted of a flat, rectangular glass jar filled with water, which was changed from time to time a& it became warmed. The aquarium in which the animals were tested was about four inches deep, twelve inches wide, and twenty-four inches long. Inside it was painted a dead-black, in order to prevent internal reflection. The two ends were made of glass. On the inside of the end away from the light, which may be called the negative end, was fitted a black-painted slate to absorb the light. An aquarium of relatively large size was used, because G. annulatus is very active and not pronouncedly phototropic; hence the larger the aquarium, the better the chance of discovering a tendency for the animals to remain near or away from the lighted end. Furthermore, thigmotropic responses to the sides of the aquarium being less frequent if the tank is large, one of the most disturbing influences in such experiments may be reduced by this means. In order to cut off reflections from the walls of the room, the aquarium was placed in a frame-work which was hung round with black cloth, within which light was admitted only through the water of the aquarium. The central point of the aqua- rium was placed 1 metre from the source of light. Under these con- ditions the maximum intensity of light at the positive end of the aquarium was about 110 candle-metres. The intensity of light at the negative end varied from 30 candle-metres, when very clear water was used, to something below this intensity, if, as sometimes happened, the water was slightly turbid. The animals chosen for a given day’s experiments were placed the night before in a dark-box 210 Grant Smtth. in the dark-room where the water had a temperature of 15° C., the same as the water in the aquarium. The water was aerated artifi- cially by passing bubbles through it. 2. Experiments. — Many preliminary observations were made upon single individuals in order to discover the general tendency of their responses, and greater accuracy was obtained as errors in the method of treatment were eliminated. It is especially important that the water in which the animals are swimming be free from débris, other- wise thigmotropism may overbalance the responses to light. The animals should be transferred to the middle of the aquarium in the dark and allowed to remain quiet for two or three minutes before the light is turned on. It was not noticed that the phototropic response was changed by handling the animals, as Towle (:00, p. 351) found for Cypridopsis. But it was observed that the animals would sometimes scurry to the bottom upon being transferred to the aquarium, and lie there quietly for a time, just as they do when one attempts to capture them at the shore. Now and then animals were found which were not typically active, and such were discarded. Occasionally individuals were found which gave responses opposite to those of the majority. This lack of homogeneity was probably due to the fact that the specimens were not brought to the laboratory as a single colony, but in several small catches, and consequently under somewhat different physiological conditions. Inasmuch as Parker (:02, p. 110) found that the responses of male and female copepods are not alike, the amphipods were separated on the basis of sex. Most of the observations were made on females, though as a matter of fact there proved to be little, if any, difference in the phototropic reactions of the two sexes. My method was to use*several animals at a time, and to record at half-minute intervals the number to be found in the positive half of the aquarium (7. ¢., the half toward the light) during an hour. To use more individuals than one can count at a glance renders accurate observations impossible, and hence six animals were chosen as the largest number that could be easily and instantly counted. In Table I, which illustrates the method of keeping the records, and is the record for Set 4 in Table II, each number in the horizontal rows in the body of the table represents the number of individuals, out of six, which were found in the positive half of the aquarium at the end of given half-minute intervals. For convenience, these are grouped into ten-minute periods, each one of which is represented by a horizontal Pigment-Migration on the Phototropism of Gammarus. 211 row of figures, and at the end of each of these rows a total is given, which represents the number of records for the positive half in a possible total of 120. The table contains records of six ten-minute intervals of the first hour, and a final record (seventh ten-minute period) taken at the beginning of the second hour. TABLE, I. Records of the number of individuals, out of six, found in the positive half of the aquarium at the end of half-minute intervals during an hour and ten minutes. Differences between + and — Number of individuals in positive half of aquarium : halves of at end of half-minute intervals in each aquarium in 10-minute period. : 10-minute periods. Per cent. —03.3+ 43.3+ 48.3+ 46.6+ 66.6+ 60.0 66.6+ From an inspection of Table 1 it is to be seen that the positive half of the aquarium was not entirely free from animals at any time during the entire seventy minutes, at the instants when the records were taken. There were only two occasions when the number in the positive end was reduced to one, and these occurred during the first ten minutes. On the other hand, the negative half was found entirely deserted twenty-two times, but all these observations were made some time after the first ten minutes. Moreover, during the first ten minutes the animals were so distributed that the total of positive records, 58, was very near to the indifferent point, 60; in the second ten minutes this total mounted quickly to 86, after which it showed a fluctuating increase but never reached the possible maximum of 120. It must not be supposed that the totals recorded in Table I repre- sent the real influence of the light, nor is it easy to get an exact expression for this. But it seems probable, since the animals while in 212 Grant Smith. the dark coursed more or less continuously from the positive to the negative end of the aquarium and back again, that some of this locomotion takes place in the light. In the case ofa positive animal, a measure of this activity can be found in the number of times animals are present in the negative half of the aquarium; and if this number is subtracted from the number of records for the positive half, the remainder should be a measure of the true light-reaction. This computation has been carried out for Table I, and is given in the last two vertical columns in that table. From this, as well as from other aspects of Table I already alluded to, it appears that the animals were indifferent or slightly negative in their responses at first, and that they afterward became quickly and decidedly, though not com- pletely positive. ’ Seven other sets of animals were dealt with in the way just described, and a summarized statement of the results, including those from Table I, is given in Table II. ABU Ele Summary of the records on eight sets of observations each similar to the one given in detail in Table I. Sets of observations. erase etween + and — halves of aquarium in 10-min. periods. Averages. Per cent. . | Percent. 76.5 | 63.7 275 89.8 | 74.8 49.8— 90.3 | 75.2 50 6+ 95 | 92.5 | 75.4] 5 54.2— 100 92.6)| 75.5/| & 54.4— 106 96 | 7 | 100 | 108 93.3 76.1 55.6+ i | | i | Retested after “i Mocs 100" Sanae Caos 104 OL” 96.5 80.4 In comparing this table with Table I we see in Sets 5 and 6, as in Set 4 just described, that the responses during the first ten-minute period are slightly below the indifferent point, z. ¢., the animals are Pigment-Migration on the Phototropism of Gammarus. 213 slightly negative. This uniformity is probably due to the fact that the animals for these three experiments were brought from the shore in one lot, and thus probably were in a uniform physiological con- dition. In the five other sets (Table II) the response was not negative at any time, but zt was much lower at first than subsequently. There is the possibility that the use of a stronger light would have rendered even these five remaining sets of animals negative at first; and thus the table would have shown a generally negative condition at the outset. However this may be, it is certain that the light- responses in three of my sets out of eight were for a few minutes slightly negative, and in all the sets the responses were not strongly positive at first. Indeed, if we examine Table I more closely we shall find that the animals were more strongly negative in the first five minutes, than the total would indicate. During that period the possible maximum number of responses was sixty, and of this number only twenty-four were positive, while thirty-six were negative. Hence the difference, twelve negative responses, is a measure of the true light-influence for the first five minutes. The results were similar in Sets 5 and 6 (Table II). Thus we probably have in these records evidence of the well-known phenomenon of reversal of phototropism. But it differs from all other cases, I believe, in the fact that the reversal was quickly made and without any increase of light at its source. If the responses for the first ten minutes in Sets 4, 5, and 6 be left out of consideration, the remaining responses are uniformly positive. This was quite the reverse of what was to be expected from the general statement made by Holmes (:01, p. 212), that the aquatic amphipods are negative. The specimens of G. annulatus (= G. locusto Linn.!) which Holmes used were from a fresh-water pond at Falmouth, Mass., and the light and other conditions under which he experimented were so different from those in the present case that a basis for the accurate comparison of results is wanting. Holmes has shown that the phototropism of amphipods was at times reversed by foul water. It did not seem possible that I had used foul water in my experiments, but to exclude this point with certainty, I made tests with animals and water brought fresh from the shore. These animals also were not negative, though they were by no means as 1 Dr. Hotmes informs me that since the publication of his paper he has de- termined that the species with which he worked was G. annulatus rather than G. locusto Linn. 214 Grant Smith. active as the previous lots. I think there is no doubt that under the conditions of my experiments G. annulatus, though indifferent or slightly negative at the outset, is generally positive. This is, of course, particularly true after the animals have remained for some time in the light. In Table I it has been shown that the animals were strongly positive after they had remained in the light for an hour, giving a total of 100 positive reactions out of a possible 120, and with a light-response of 80(67 per cent). The average of the last (seventh) series of records in Sets 3 to 8 (Table II), though the observations were not all made at exactly the end of an hour, also shows essentially the same condition, in that there are 579 positive reactions out of a possible 720, and therefore a light-response of 438 (60 per cent). Though usually positively phototropic, G. annulatus is certainly not so remarkable an example of this as Talorchestia longicornis, which, according to Holmes (:01, p. 213), follows a light carried about in a darkened room, and remains positive even to light so intense as to cause its death. In the present case, however, it is significant that not one of the nine hundred and sixty half-minute records for the eight sets showed the positive half of the aquarium entirely deserted ; and an average of 4.5 out of a possible six animals were in the posi- tive half all the time. On the other hand, the negative half was frequently free from animals. Connected with the reversal in the direction of the response already mentioned, is the interesting fact that the positive response is not constant, but gradually increased for an hour. In Table II, it may be seen that the response starts low in every set, even if it is not negative, and that in every set there is a sharp increase of the positiveness during the second ten minutes as compared with the first, after which the rise is more gradual to the end of the hour and somewhat beyond. There are two points in Table II at which the responses decreased markedly: in the fifth period of Set 5 and in the sixth period of Set 6. Since, however, the responses rise immediately after each of these to a higher level, it is probable that these depres- sions are purely accidental. Notwithstanding the fluctuations, it must be evident that when the animals first emerge from the darkness they are individually indifferent, slightly negative, or more frequently slightly positive to light; that within ten minutes they change and become decidedly positive, — a condition that is gradually intensified for an hour or more. Pigment-Migration on the Phototropism of Gammarus. 215 IV. Discussion OF RESULTS. From what has been said in the preceding sections it is plain that the distal migration of the retinal pigment in G. annulatus is accom- panied with a change in the animal’s phototropism. During the first fifteen minutes of exposure to light, there is a decided movement of retinal pigment distad, and parallel with this there is a very rapid rise in the number of positive responses. Then follows a period of some three-quarters of an hour, during which a more gradual pig- ment-migration distad is accompanied by a gradual increase in positive responses. Thus it appears that with an increased protection of the rhabdome from light by the migration of pigment there goes an increase in positive phototropism. The effect of retinal pigment-migration on phototropism is so evi- dent from these observations that clearly it must be taken into account in the study of the photic reactions of any animal whose eyes show this phenomenon. It has been shown that an animal may respond to a sudden change in the intensity of illumination by a gradual alteration in its reactions, and that this alteration is depend- ent upon the adjustment of the retinal pigment to the new conditions. In studying the phototropism of such an animal, it is therefore necessary to know what was the position of the pigment when the experiment began, what length of time it requires to reach the extreme conditions, what the previous illumination had been, and what intensity of light is being used. To institute just com- parisons it is desirable that investigators should have some com- mon starting-point and in work of this kind it is obvious that the “dark” condition is the only basis from which experiments may safely proceed. It is my hope soon to contribute something more in reply to some of the questions not answered by these observations. My warmest thanks are due to Professor G. H. Parker, whose supervision I have had in these experiments. V. SUMMARY. 1. In Gammarus annulatus, as had previously been shown by Parker in G. ornatus, the retinal pigment in the “dark” eye is accumulated in the distal and the proximal ends of the retinular cells and leaves the rhabdome exposed to the reflecting action of the accessory pigment. In the “light” eye the rhabdome is ensheathed 216 Grant Smith. by the retinal pigment, and the accessory pigment can no longer act as a reflector. 2. The migration of the pigment from the condition in the “dark” eye to that in the “light” eye is rapid during the first fifteen minutes of exposure to light, and then slower till its completion, which occurs in about an hour. 3. G. annulatus when taken from the dark and exposed to a con- stant light of 30 to 110 candle-metres in intensity is, during the first ten minutes, either indifferent, slightly negatively or slightly posi- tively phototropic. This condition rapidly gives way to a strongly positive phototropism, which gradually increases, in the course of an hour or so, to a maximum. 4. The retinal pigment-migration, since it controls the amount of light reaching the rhabdomes, is in all probability the means of induc- ing this change in phototropism. BIBLIOGRAPHY. HouMES, S. J. ‘ol. Phototaxis in the Amphipoda. Amer. Jour. Physiol., vol. 5, no. 4, pp- 211-234. PARKER, G. H. ? 97. Photomechanical Changes in the Retinal Pigment Cells of Palemonetes, and their Relation to the Central Nervous System. Bull. Mus. Comp. Zodl. Harvard Coll., vol. 30, no. 6, pp. 275-299, I pl. PARKER, G. H. ’99. The Photomechanical Changes in the Retinal Pigment of Gammarus. Bull. Mus. Comp. Zod]. Harvard Coll., vol. 35, no. 6, pp. 141-148, 1 pl. PARKER, G. H. :02. The Reactions of Copepods to Various Stimuli, and the Bearing of this on Daily Depth Migrations. Bull. U. S. Fish Comm., vol. 21, for Igo!, pp. 103-123. TOWLE, E. W. 700. A Study in the Heliotropism of Cypridopsis. Amer. Jour. Physiol., vol. 3, no. 8, pp. 345-365. Sa NATURE OF CARDIAC INHIBITION WITH SPECIAL mEPERENCE TO: THE) HEART OF LIMUEUS. BY vA. |; CARLSON: [From the Hull Physiological Laboratory, University of Chicago. | 1. THE INTERPRETATIONS OF THE ACTION OF THE VAGUS ON THE HEART. BE! for the fact that the unique anatomical relations between the nervous and the muscular tissues in the heart of Limulus render it possible to determine whether the inhibitory nerves act on the ganglion or directly on the heart-muscle, or on both, the sub- ject of inhibition of the Limulus heart might be dismissed with the statement that the heart is connected with the brain by two pairs of inhibitory nerves, the stimulation of which produces the identical __ phenomena observed in the vertebrate heart on stimulation of the vagi. The view originally advanced by Weber that the vagi inhibit the automatic activity of the ganglion cells in the heart in a manner analogous to the processes of inhibition in the central nervous sys- tem has been replaced by the theory that the cardio-inhibitory nerves act directly on the heart-muscle. The latter theory, based mainly on the researches of Gaskell and Engelmann and their students, is very generally accepted. It is a necessary deduction from the myogenic theory of the heart-beat and the nature of co-ordination or conduc- tion in the heart. If the heart-rhythm is an expression of the metab- olism of the heart-muscle, or due to chemical stimulation from the blood, and not the result of nervous impulses reaching the muscle from automatic or reflex nerve-centres, it is self-evident that any force that shall stop that rhythm must act directly on the muscle. The two theories can thus not part company. When conclusive proofs of the myogenic nature of the rhythm and the conduction are forthcoming, the other follows as a matter of course. Apart from the observations which appear to count in favor of the myo- genic theory, such as the rhythm of the embryonic heart, the view that the inhibitory nerves act directly on the heart-muscle seems to 217 218 A. J. Carlson. receive support in the nature of the changes produced in the heart by the stimulation of the nerves. The inhibitory nerves diminish the rate and the strength of the beats, the power of conduction, and the excitability to direct stimulation. Weber himself observed that the strength of the heart-beat may be diminished by stimulation of the vagi. Now, in as much as the heart-muscle contracts accord- ing to the “all-or-none” law, this diminution of the amplitude of the beats is, according to Gaskell’s reasoning, a conclusive proof that the vagi act directly on the heart-muscle in a manner to diminish the power of contraction. If the premise were sound, this conclusion would be valid. That the excitability of the heart to direct stimulation is diminished — some have even claimed abolished — during stimu- lation of the vagi was first pointed out by Schiff. The influence of the inhibitory nerves on the conductivity in the heart has been par- ticularly elucidated by the researches of Gaskell and Engelmann. While the above interpretations of the nature of the action of the vagi on the heart are corollaries of the myogenic theory of the con- traction and the conduction, and must be accepted as proved as soon as conclusive demonstration of the myogenesis of the normal beat and the normal co-ordination is forthcoming, it is evident that all of these conclusions do not necessarily fall to the ground with the myogenic theory. How can the influence of the inhibitory nerves on the heart be reconciled with the neurogenic theory? The sum-total of the numerous researches on the nature of cardiac inhibition since the time of Weber is this, that the inhibitory nerves decrease the rate and the strength of the beats, the power of conduction, and the ex- citability to direct stimulation ; but whether this action of the nerves is on the intrinsic nervous tissue, or on the muscular tissue, is, it seems to me, just as much an open question to-day as it was fifty years ago. The view that the inhibitory nerves diminish conduc- tivity in the heart by direct action on the muscle is incompatible with the neurogenic theory, for if the conduction and co-ordination take place in the nervous elements, as is the case in the Limulus heart, the inhibitory nerves can diminish the rate and the power of conduction only by acting on these nervous elements. But if the contractions are caused by nervous impulses from the ganglion-cells, the inhibitory nerves may retard and diminish or stop these contrac- tions, either by acting on the ganglia in a way to stop or diminish their activity, or by acting on the heart-muscle in a way to render it less excitable to the nervous impulses reaching it. But does not the The Nature of Cardiac Inhibition. 219 fact that the inhibitory nerves diminish the excitability of the heart to direct stimulation, and diminish the amplitude of the beats, prove conclusively that they act directly on the muscle? That the heart of some of the higher vertebrates responds to direct stimulation in accordance with the “all-or-none”’ principle, appears to be well-estab- lished; but that the heart-muscle, apart from the nervous complex, responds to stimuli in this manner is an assumption with scant ex- perimental evidence in its support. We know as yet of no drug or solution that abolishes the action of the accelerator or intrinsic cardio- motor nerves on the heart without injury to the muscle. Hering has lately shown that the augmentor nervous mechanism in mammals (rabbits, dogs, cats, monkeys) lives on for many hours after the death of the animal, and may be restored by Ringer’s solution, even after freezing the heart and the heart-nerves for hours.? Since we have not yet succeeded in throwing out the augmentor or motor nervous mech- anism in the heart, we do not yet know the properties of the heart- muscle apart from this nervous mechanism, and it is furthermore evident that the diminished excitability of the heart during stimula- tion of the vagi is no proof that the vagus fibres act on the heart- muscle. At least to some kinds of stimuli ganglion-cells have a higher excitability than nerve-fibres, and nerve-fibres a higher excita- bility than muscle-tissue. A diminished excitability of the nervous elements in the heart, unattended by any change in the excitability of the muscle itself, would, therefore, in all likelihood diminish the excit- ability of the heart to direct stimulation. This action of the inhibi- tory nerves is, therefore, as readily explained on the neurogenic as on the myogenic theory. The changes in the electrical tension of the heart (ventricle) of the toad on stimulation of the vagi is also ascribed by Gaskell, who discovered the fact, to a direct action of the inhibitory fibres on the muscle. The point of peculiar interest is this, that the change in electrical tension on stimulating the vagi is the opposite of that caused by stimulation of the augmentor-nerves. But there is noth- ing in Gaskell’s experiments which would warrant us ascribing these changes in the electrical tension to changes in the muscle rather than to changes in the nervous tissue. If the inhibitory and the augmen- 1 HERING: Archiv fiir die gesammte Physiologie, 1903, xcix, p. 250; CARLSON: Science, 1904, xx, p. 684. 2 HERING: Archiv fiir die gesammte Physiologie, 1903, xcix, p. 245. 8 GASKELL: Journal of physiology, 1887, viii, p. 404. 220 A. /. Carlson. tor nerves induce changes of opposite electrical sign in the ganglion cells or the cells and their processes as well, under the conditions of Gaskell’s experiment this variation in tension would be registered by the electrometer just as readily as a corresponding change in the heart-muscle. It would, therefore, seem that the changes in the heart produced by the stimulation of the cardio-inhibitory nerves can just as readily be interpreted on the neurogenic as on the myogenic theory. And as far as the vertebrates are concerned, we have no conclusive proof in favor of either interpretation, just as we have no clear-cut demonstration of the myogenic or the neurogenic na- ture of the contraction,’ although the action of some of the drugs on the heart can with difficulty be interpreted on the Gaskell-Engel- mann hypothesis.? é It has been shown in previous papers® that in Limulus the cause of the heart-beat is to be sought in the ganglion-cells, and that con- duction in the heart takes place in the nervous and not in the mus- cular tissue. Of these two facts there can be no question, as their demonstration is direct and relatively simple. These facts alone 1 PoRTER (Journal of experimental medicine, 1897, ii, p. 391; American text- book of physiology, i, p. 151) regards his observations on the ventricle of the dog as a demonstration of the myogenesis of the heart-beat. The fact that the isolated portion of the ventricular apex continued to beat with a rhythm different from the rest of the heart seems to exclude the possibility that the contractions were caused by impulses through nerve-fibres that might have remained intact along the course of the nutrient artery. Isolated portions from any region of the ventricle of the hagfish continue to beat rhythmically for hours, even in the absence of nutrition ; but nothing can be concluded from this, as nothing is at present known of the nervous elements in the heart of this animal. It is well known that the mamma- lian ventricle possesses greater degree of automatism than does the ventricle of reptiles and amphibians; but even if it is a fact that an isolated portion of the apex of the dog’s ventricle continues to beat for some time when fed with blood through its nutrient artery, that does not prove that the heart-rhythm is myogenic, because the question of the presence of nerve-cells in the apex of the ventricle can not be considered as settled in the negative. SCHWARZ (Archiv fiir mikro- skopische Anatomie, 1899, liii, p. 63) and others have described a multitude of cells along the course of the nerves and the arteries over the entire ventricular myocard; but SCHWARz does not think that these cells are nervous in their nature, chiefly because of their small size and the absence of nucleated envelopes. It might be asked why all the small nerve-cells in the heart should be provided with a nucleated envelope, when this is not the case, for example, in the plexuses of AUERBACH and MEISSNER. 2 HARNACK: Archiv fiir Physiologie, 1904, p. 415. 8 CARLSON: This journal, 1904, xii, p. 67; /did., 1905, xii, p. 471. The Nature of Cardiac Inhibition. 221 make Weber’s theory of the mechanism of cardiac inhibition a pos- sibility, if not a probability. We will now proceed to show that Weber’s theory is not only a possibility, but an actual fact, at least in the Limulus heart. 2. THE ANATOMY OF THE CARDIO-REGULATIVE NERVES IN LIMULUS. For a more detailed account of the cardiac nervous system of Limulus, the reader is referred to the papers of Milne-Edwards,' and Patten and Redenbaugh.? The intrinsic nervous system of the heart consists, in short, of a ganglionated cord running the whole length of the heart in the dorso-median line, and a system of nerves passing from either side of this nerve-cord to the heart-muscle, the several nerves uniting in one common nerve-trunk at either lateral angle of the heart. This intrinsic nervous system is connected with the cen- tral nervous system by an elaborate complex of nerves shown in Fig. 1. These nerves have been worked out in great detail by Patten and FicurEe ].— Heart of Limulus, dorsal view. #. c., dorso-median nerve-cord. 7. 1. ¢. pericardial nerves. 7-8, cardiac branch from the last two hemal nerves from the brain. 9-13, cardiac branches from the hemal nerves of the abdominal ganglia. 7-8, contain inhibitory, 9-11, augmentor fibres. Redenbaugh, and the only point touching the anatomy that I have to add to the excellent figures and descriptions of these authors is the fact that the two posterior pairs of nerves from the dorsal side of the brain make connections with the nerve-cord ca the heart in the man- ner shown in the diagram. Patten and Redenbaugh were not able to trace these nerves to the heart-ganglion. It can be done, however, in the largest specimens, and stimulation of the nerves near the brain 1 MILNE-EDWARDS: Annales des sciences naturelles, 1873, sér. 5, xvii. 2 PaTTEN and REDENBAUGH: Journal of morphology, 1899, xvi, p. 9I. 222 A. J. Carlson. shows by its effects on the heart that the relations disclosed by the dissection are not due to mistaking connective tissue-fibres for nerve- fibres. From the dorsal side of the brain, corresponding to each pair of nerves to the ambulatory appendages, is given off a pair of relatively tiny nerves which take a dorsal direction to innervate the integument, the viscera, and the dorsal musculatures. These nerves are evidently homologous to the nerves occupying a similar relation with reference to the thoracic ganglion in the crayfish (Palinurus). The anterior pair of nerves from each abdominal ganglia takes a similar course and makes connections similar to those made by the nerves from the dorsal side of the brain. As these nerves go to innervate structures dorsal or hzmal to the level of the central nervous system, Patten and Redenbaugh call them ‘“hzemal” nerves in contradistinction from the nerves to the ambulatory appendages and the gills, which they designate the “neural” nerves. The hzmal nerves from the abdominal chain of ganglia take a dorsal and posterior direction, and, after giving off branches to the digestive tract and the integument, penetrate the pericardial cavity, sending small filaments in the dorsal pericardium to unite with the median nerve-cord on the heart approximately opposite the fourth to the eighth pairs of ostia. These cardiac branches, with the exception of those from the ninth and tenth pairs of nerves, are so tiny that they cannot readily be distinguished from the connective tissue-fibres in the pericardium. Before the nerves enter the dorsal pericardium to connect with the nerve-cord, each nerve sends a communicating branch to the relatively large nerve-trunk which runs parallel to the heart at either angle of the pericardial cavity. These nerves are termed pericardial nerves by Patten and Redenbaugh. Of the hemal nerves taking their origin from the brain, the only ones I was able to trace to the heart were the last two pairs (7, 8). The cardiac branches of these two nerves unite in one common trunk before reaching the pericardial cavity. The main branches of this nerve go to make up the pericardial nerve and to supply the large inter-tergal muscle, which lies dorsal to the heart in this region. The branches that pass to the epidermis dorsal to the heart connect with the ganglion on the heart in the manner represented in Fig. 1. There is considerable individual variations as to the exact place of union with the nerve-cord. In some specimens there appears to be connec- tions only at the level of the second pair of ostia, in others the con- The Nature of Cardiac Inhibition. 22/3 nections are made in the middle of the third segment, while in still others the main if not the only connections are the ones just behind the third pair of ostia. 3. THE PHYSIOLOGY OF THE CARDIO-REGULATIVE NERVES. In order to study the influence of the brain and the abdominal ganglia on the heart, it is not practicable to expose the heart from the dorsal side by removing the carapace dorsal to the heart, because this can rarely be done without some injury to the nerves in the dorsal pericardium. The dorsal carapace may be removed in the region of the first heart-segment, so as to allow the movements of the heart to be observed, or the anterior end of the heart may be exposed from the ventral side by removing the intestine and the reproductive gland. This can be accomplished without any injury to the nervous connec- tions with the heart. For more accurate work it is desirable to connect the heart with a recording apparatus. The following proved to be the only feasible method. After exposing the brain and the abdominal ganglia from the ventral side, the carapace was severed transversely in the region of the first pair of cardiac ostia, and the anterior portion of the carapace removed, care being taken in remov- ing the digestive tract, the reproductive gland, and the dorsal muscles not to injure the exposed heart-segment. The remaining part of the animal was now tilted edgewise on the supporting platform, and held by clamps in such a way that one side of the exposed heart-segment could be connected with the recording lever, the opposite side of the heart-segment being held in a fixed position by means of a clamp. The force of the contraction of the heart is in the transverse direc- tion, owing to the transverse arrangement of the muscle-fibres. In fact, when the heart is severed from the suspensory ligaments it actually elongates on contraction. The above contrivance thus allows graphic records to be obtained of the influence of the ventral nervous system on the heart. But there is this difficulty to contend with, that the hzmal nerves send fibres not only to the heart and the integu- ment (sensory ), but also to the inter-tergal muscles and other muscles about the pericardial cavity. Stimulation of the brain, the abdominal ganglia, or the hemal nerves directly produces contractions of these muscles, which in turn alters the tension on the heart and displaces it to some extent, thus obscuring the tracings. The errors from this source can be eliminated by stimulating the cardiac branches of the hzemal nerves in the pericardium. 224 A. J. Carlson. ¢ As long as the hemal nerves are intact stzm/ation of the brain with the weak interrupted current inhibits the cardiac rhythm. ‘The tracing in Fig. 2 is typical of these inhibitory effects. Complete inhibition | eM OWOUNON CC UU FIGURE 2.— Record from anterior end of heart on stimulation of the brain, the hemal nerves being intact. Incomplete inhibition. Weak interrupted current. of the heart is not maintained in any case for more than a few min- utes, but on continuing the stimulation, a diminution of the rate and the amplitude of the beats may be observed for as long as twenty minutes. Mechanical stimulation of the brain also produces com- plete or incomplete standstill of the heart. And these inhibitory effects are furthermore produced by stimulation of any of the sensory nerves connected with the brain. Stimulation of the central end of any of the nerves to the ambulatory appendages with the weak inter- rupted current diminishes the rate and the strength of the beats, at least at the beginning of the stimulation. This appears to be a true reflex inhibition, and any of the sensory nerves-to the brain appears to be able to act as a “‘ depressor” to the heart-rhythm. The inhibi- tion of the heart on stimulation of the afferent nerves to the brain is neither as marked nor as long maintained as that produced by stimu- lating the brain directly. A typical tracing showing the reflex inhibition is reproduced in Fig. 3. KRONE AAA FicuRE 3.— Reflex inhibition of the heart by stimulating one of the ambulacral nerves with a weak interrupted current. Hzemal nerves from the brain intact. It can be shown by direct experiment, as well as by the process of elimination, that the inhibitory fibres to the heart leave the brain in the last two pairs of hemal nerves. After cross-section of the abdominal commissures near the brain, direct stimulation of the brain or stimulation of the ambulatory nerves still causes inhibition. Nor does lesion of the hamal nerves 1 to 6, that is, all the hamal nerves save the last two pairs, abolish the inhibitory effects of stimulation The Nature of Cardiac Inhibition. 225 of the brain, but after lesion of the last two pairs (7, 8), leaving all the other hzemal nerves intact, stimulation of the brain has no effect on the heart, that is, in case the abdominal commissures are also severed. When the haemal nerves 7 and 8 are isolated and stimulated near the brain, typical cardiac inhibition is produced. There appears to be no difference in the nerves on the right and those on the left side in their efficacy to inhibit the heart, but the inhibition is usually more complete, and may be maintained for a longer time on stimula- tion of the seventh than on stimulating the eighth pair of nerves. It is, therefore, probable that the majority of the inhibitory fibres pass to the heart in the seventh pair of haemal nerves. The tracing repro- duced in Fig. 4 is typical of the effects on the heart of stimulating the hemal nerves 7 and 8 near their point of union with the brain. IYI UU