tv Waals yt Seale Meth * a PK it HN t nie tet Pa tht, RNa ne pian bane 4 i De ta ty : : ) By 1) Z sat hae ay ih ay ’ Waar fr iN ye teh nes Ge frie! ue wey ai aN F ( Avi y on Pt ae Na Heat }) ie ri oe ss Hote ea are aye Wi Mi: Neb iy ey a ut a? i Ate AHA Me v glalet , te ) ty K in MARA 2 Walsh hts Had 9 ay Hy ‘Un as = = ete : Wit i Bate if bitte (hs ¢ eee Oo Dane H se Lo faa bean arate Le es 4 i fn ee Were rai es Halter ; ¥, eh tiie iss ae Aid mi ‘ fh) cK a ALG, Fy 9 ‘¥, ht ye an ate iy nett x] Cesare es yt FRY ae » 4 jae aie Cran # yy e, tig dha ‘ i Pom 8 Nye eri « p ou 1 2G bp : TET ? Me yas ii uy : A bay a } VE 5 aes i ae a Pane rT > Ly ne 3 THE AMERICAN JOURNAL OF PiySlOLOGY. EDITED FOR The American Physiological Society BY H. P. BOWDITCH, M.D., BOSTON FREDERIC S. LEE, PH.D., NEW YORK R. H. CHITTENDEN, PH.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 Se THE AMERICAN JOURNAL Rl Y¥olOj1.O G ¥- VOLUME XXIV. ag. my \0 Son \ y: ca\ BOSTON. U.S: A. 1909. Copyright, 1909- BY Wok PORTER: Aniversity {ress Joun WiLson AND Son, CAMBRIDGE, U.S.A. i Oe i wee wy % CONTENTS. No; Sly APRIL 15 rgq00; HEAT COAGULATION IN SMooTH MuscLE: A COMPARISON OF THE EFFECTS OF HEAT ON SMOOTH AND STRIATED MuscLe. By Edward B. Meigs ON THE CONNECTION BETWEEN CHANGES OF PERMEABILITY AND STIMULA- TION AND ON THE SIGNIFICANCE OF CHANGES IN PERMEABILITY TO Carson Dioxipe. By R. S. Lillie. FACTORS REGULATING THE CREATININ OuTPUT IN MAN. By P. A. Levene and L. Kristeller . ACAPNIA AND SHOCK.—III. SHocK AFTER LAPAROTOMY: ITS PREVENTION, PRODUCTION, AND RELIEF. By Yandell Henderson THE ROLE oF THE ASH CONSTITUENTS OF WHEAT BRAN IN THE METAB- OLISM OF HerpivoraA. By E. B. Hart, E. V. McCollum, and G. C. Humphrey . . THE EFFECT OF SMOKING UPON THE BLOOD PRESSURES AND UPON THE VoLUME OF THE HAND. By James W. Bruce, James R. Miller, and Donald R. Hooker ... , THE PRODUCTION BY HYDROGEN PEROXIDE OF RHYTHMICAL CONTRACTIONS IN THE MARGINLESS BELL OF GONIONEMUS. By O. P. Terry STUDIES IN THE PHYSIOLOGY OF THE CENTRAL NERVOUS SYSTEM. — I. THE GENERAL PHENOMENA OF SPINAL SHOCK. By F. H. Ptke HyYDROLYsIS OF VITELLIN FROM THE HEN’s Ecc. By Thomas B. Osborne and D. Breese Jones Hyprotysis OF THE MuscLte oF SCALLOP (PECTENS IRRADIANS). By Thomas B. Osborne and D. Breese Jones EXPERIMENTAL STUDIES ON THE PHYSIOLOGY OF THE MOoLLuscs.— FouRTH Paper. By Lafayette B. Mendel and H. Gideon Wells CONCERNING THE SUPPOSED CONNECTION BETWEEN PROTEIN COAGULATION AND THE HEAT SHORTENING OF ANIMAL TissuES. By Edward B. Meigs PAGE 66 86 104 18 9/ 124 153 161 170 177 Vi Contents. No. II, May 1, 1909. PAGE MERCURIAL POISONING OF MEN IN A RESPIRATION CHAMBER. By Thorne M: Carpenter and Francis G. Benedict: . 2 i025. 2-2 2s 2 = > HOF PRELIMINARY OBSERVATIONS ON METABOLISM DURING FEVER. By Thorne M. Carpenter and Francis G. Benedict . . 4 = --2 = = = 2 > = = = 208 THE VARIATIONS IN THE ENZYME CONCENTRATION WITH THE VARIATION IN THE BLoop SUPPLY TO THE SECRETING GLAND. By J. G. Ryan 234 On RHEOTROPISM. — II. RHEOTROPISM OF FISH BLIND IN ONE EYE. By J tig) eae 8 ae ee ee Peer Ate A nc tea Seo) 2YU! HypROLysIs OF CRYSTALLIZED ALBUMIN FROM HEN’s Ecc. By Thomas B. Osborne, D. Breese Jones, and Charles S. Leavenworth ..... - 252 THE EFFects OF CHLORIDE, SULPHATE, NITRATE, AND NITRITE RADI- CLES OF SOME ComMMON BASES ON THE FRroc’s Heart. By F. C. COORE ee i Shs. cE Pah ec ee a ee A MerHop oF DETERMINING THE POSITION OF THE CENTRE OF GRAVITY IN 1TS RELATION TO CERTAIN BONY LANDMARKS IN THE ERECT Position. By Edward Reynolds and Robert W. Lovett ...... .- 286 A Notre ON THE ABSORPTION OF Fat. By R. H. Whitehead . . . ~~ - 294 No, TIL unt: a; 2900. THe EFFECTS OF BONE ASH IN THE DIET ON THE GASTRO-INTESTINAL Conpitions oF Docs. By Alfred Peirce Lothrop ....----- 207 AN Improvep METHOD OF DESICCATION, WITH SOME APPLICATIONS TO BLOLOGICAL “PROBLEMS: By Ly FShackell) eau 2 ene = pe eee THE INFLUENCE OF THE TEMPERATURE OF THE HEART ON THE ACTIVITY OF THE VAGUS IN THE TorTOISE. By G. M. Stewart. . ....- - - 341 AN APPARATUS FOR STUDYING THE RESPIRATORY EXCHANGE. By Francis G. Benedvcl SS Sa Ss ee nen ee oes No: IV, JULY.-1} 900: CAN FUNCTIONAL UNION BE RE-ESTABLISHED BETWEEN THE MAMMALIAN AURICLES AND VENTRICLES AFTER DESTRUCTION OF A SEGMENT OF THE AURICULO-VENTRICULAR BUNDLE? By Joseph Erlanger. . . . - 375 PsEUDO-FATIGUE OF THE SPINAL Corp. By Frederic S. Lee and Sumner Bverin gna. oo Saad gel ae gene ed ee I aa ee THE INNERVATION OF THE CORONARY VESSELS. By Carl J. Wiggers. . 391 THE COAGULATION OF Buoop, By EL. J. Relivey: . . 2: : 22 2 5 = eee Contents. No. V, AUGUST I, 1909. Hyprotysis oF Ox Muscite. By Thomas B. Osborne and D. Breese Jones THE RATE OF DIGESTION IN COLD-BLOODED VERTEBRATES. — THE INFLU- ENCE OF SEASON AND TEMPERATURE. By Oscar Riddle ...... THE RELATION OF IONS TO CONTRACTILE PrOcEsSES.— IV. THE In- FLUENCE OF VARIOUS ELECTROLYTES IN RESTORING MuscULAR CoN- TRACTILITY AFTER ITS Loss IN SOLUTIONS OF SUGAR AND OF Macnesium Cuaiormpe. By Ralph S. Lilie ........---- THE ABSORPTION OF FATS STAINED WITH SuDAN III. By Lafayette B. NYA GELATIN AL Ge ogee ER a PAD ery eS Pa en a ee vil ~ 7 > ° . +e" ers ve THE American Journal of Physiology. VOL. XXIV. APRIL 1, 1909. NO. I. HEAT COAGULATION IN SMOOTH MUSCLE; A COM- PoaeisON On THE EFRECTS:OF HEAT ON SMOOTH AND STRIATED MUSCLE. By EDWARD B. MEIGS. [From the Laboratory of Physiology in the Harvard Medical School.] [~HE effects of heat on striated muscle have received a great deal of attention and are described in detail in many of the text-books of physiology. The effects of heat on smooth muscle, on the other hand, have been little studied; Vernon? has, however, given a more or less detailed account of the changes in length which take place in this tissue at various temperatures. Fig. I is a copy of two of Vernon’s curves. The upper curve was obtained from the frog’s gastrocnemius, and the lower one from a ring of the smooth muscle of the cesophagus of the same animal. The curves represent the changes in length which take place when the tissues are heated gradually from 20° C. to go° C. in 0.7 per cent sodium chloride solution; the time required for the whole process was about thirty minutes. As is seen from the curves, the two tissues behave oppositely between about 40° and 50°, the striated muscle shorten- ing in this interval and the smooth muscle lengthening almost as markedly. At about 53° both tissues shorten. Vernon has studied the changes in length which occur under the influence of heat in the striated and smooth muscle of a number of different animals. He finds that the temperatures at which these 1 VERNON: The journal of physiology, 1899, xxiv, p. 239. I 2 Edward B. Meigs. changes make their appearance vary in different species, but the general result of his experiments is that shortening under the influ- ence of heat begins at a considerably higher temperature in smooth muscle than in striated muscle. He evidently considers a preliminary lengthening characteristic of smooth muscle, though he states that it does not always occur. It is of course evident that such a length- ening could occur only in muscle which was in a state of more or less tone at the beginning of the experiment; a piece of muscle already in a state of extreme extension could not be made to lengthen any farther. It will appear later that there are other factors which may have occasionally masked the preliminary lengthening in Ver- non’s experiments. I have repeated some of Vernon’s experiments with the smooth muscle of the frog’s stomach and have obtained in general the same results. Fig. 2 represents one of my curves. As may be seen by a comparison of Fig. 2 with the lower curve of Fig. 1, the chief dif- ference between Vernon’s results and mine is that he has obtained a larger and more rapid shortening from 53° onward. The probable cause of this difference will be discussed in the proper place. It is an interesting question whether the smooth and striated muscles of the frog undergo the same chemical and coagulative changes between 40° and 50°, or whether the changes which occur in the striated muscle at these temperatures do not occur in the smooth muscle until a temperature of 53° or higher is reached. To answer this question [| have compared the behavior of the two kinds of muscle between 40° and 50° in four respects: first, in regard to the changes in irritability; second, in regard to the coagulative changes; third, in regard to the production of acid; and lastly, in regard to the changes in weight. The importance of the last con- sideration will be seen later. Vernon? states that both smooth and striated muscle lose their irritability at about 40°. In testing the irritability of the two tissues while they are being gradually heated, I have found it to persist to a somewhat higher temperature. In one experiment irritability dis- appeared in the frog’s sartorius at 44°, and in the smooth muscle of the frog’s stomach at 46°. No doubt the temperature at which irritability disappears depends on the nature and strength of the stimulus used and on the rapidity with which the temperature is raised; the same changes which occur almost instantaneously in the 7 Vernon: Loc. cit., pp. 242 and 254. Heat Coagulation in Smooth Muscle. 3 two kinds of muscle at 50° occur more slowly at 40°. It may be stated positively that irritability is permanently destroyed in both kinds of muscle by keeping them for five minutes at a temperature of 50°. It will be noted that this temperature is well below that at which the heat contraction begins in smooth muscle. The question of protein coagu- : lation is a more complicated one. A large amount of work has been done on the temperatures at which the proteins in the extracts of stri- ated muscle are precipitated, and it has been found that these ex- tracts contain proteins which are precipitated at various tempera- tures between 40° and 70°. Vin- cent and Lewis? have compared Bievicbantiotorexttactsot stated 2° 220° Me Be G0, ae muscle with that of similar ex- FicuRE 1.— Curves representing the ef- fects of heat on the length of striated tracts of smooth muscle. They and smooth muscle; after Vernon. The found that the temperatures at which the proteins of the extract were precipitated depended on the nature of the salt used in the upper curve is the heat curve of the frog’s gastrocnemius; the lower one, that of the cesophagus of the same animal. In both cases the muscle was stimulated at in- tervals by electric shocks; the effects of extraction and on the reaction of the extract, whether neutral, acid, or alkaline. They state (p. 452) that in extracts made from both kinds of muscle with 0.9 per cent NaCl solution the proteins undergo spontaneous precipitation at the temperature of the laboratory. Extracts made with 5 per cent MgSO, solution, however, behave differently, according as they are prepared from smooth or striated muscle. Some of the proteins in the striated muscle extracts are precipitated at about 47°, while no precipitation occurs in the smooth muscle extracts until a temperature of 55° or 60° is reached. But the authors adduce evidence to show that this difference depends partly on the reaction of the extracts. The extracts of smooth muscle, as ordinarily prepared, are neutral or faintly alkaline, while those of the striated muscle are distinctly acid. Vincent and Lewis quote this stimulation and their cessation at about 40° are represented in the curves. The heating from 20° to 90° occupied about thirty minutes. original size. Four sevenths the 3 VincENT and Lewis: The journal of physiology, r9g00-1901, xxvi, p. 445. 4 . Edward B. Meigs. Demant,* who states that the proteins are not precipitated in an alkaline extract of striated muscle even after several hours at a temperature of 47°. On the other hand, the addition of 0.02 per cent lactic acid to the smooth muscle extract causes one of its pro- teins to be precipitated at 40°.° It may be thought that the work of Vincent and Lewis does little toward answering the question whether any connection exists be- tween protein coagulation and the heat shortening of muscle, but their experiments leave little doubt of the fact that the temperature at which the proteins in muscle extracts are precipitated depends on the reaction of the extract and on the salt used for the extraction. There is every reason to believe that the reaction of the fresh muscle fluids is neutral or very faintly alkaline. Pieces of living striated and smooth muscle pressed against litmus paper turn it blue to about the same extent, and it is improbable that the reaction of any tissue is constantly much different from that of the blood. Finally, it is perfectly easy to prepare a neutral or faintly alkaline extract of striated muscle if the precaution be taken to crush the living muscle under ice-cold alcohol according to the method of Fletcher and Hopkins.® Extracts of striated and smooth muscle prepared in this manner do not differ appreciably in their reaction, which is about that of the blood [(E1) — he oon |e It will be shown later that, when either striated or smooth muscle is heated to a temperature between 40° and 50°, very considerable amounts of lactic acid are formed within the tissue, — much more in both cases than Vincent and Lewis found necessary to markedly alter the coagulation temperatures of the muscle proteins. It will be seen, therefore, that it is impossible to determine the temperatures at which the proteins coagulate in a gradually heated living muscle from the temperatures at which they coagulate in the muscle ex- tracts. The rapidity with which the proteins coagulate at any given temperature in the muscle will depend on the rapidity with which the acid is formed at that temperature, and this is still an unknown quantity. A rough idea of the temperature at which the proteins coagulate in the living muscle may, however, be gained by watching pieces of muscle as they are heated at the rate used by Vernon in * Vincent and Lewis: Loe. cit., p. 450. ® Vincent and Lewis: Loc. cit., p. 451. ® FretcHer and Hopxtns: The journal of physiology, 1907, xxxv, p. 252. Heat Coagulation in Smooth Muscle. 5 his experiments. In both the smooth and striated muscle of the frog there is little change in appearance under this treatment until the temperature rises above 45°; between 45° and 50° there usually occur a marked whitening and opacity in both cases. It is quite certain, therefore, that a considerable protein coagulation occurs in both striated and smooth muscle if the tissues be kept for a few seconds at 50°. The question of acid formation must next be considered. Fletcher and Hopkins‘ have shown that one of the most striking character- istics of heat rigor in striated muscle is the production of a large amount of lactic acid. These authors have made quantitative de- terminations of the amount of lactic acid in fresh muscle and in muscle which has been kept for an hour at between 40° and 50°. They find that the fresh muscle contains only from 0.015 per cent to 0.02 per cent of lactic acid, while that which has been heated contains about 0.4 per cent, —some twenty times as much. My experiments have been directed toward answering the question whether a similar production of lactic acid occurs in smooth muscle similarly heated. As material for this study I have used the mus- cular coat of the frog’s stomach and in one case that of the cat. The stomach is dissected out, freed as much as possible from ex- ternal connective tissue, and cut open along the line of the lesser curvature. The main muscular coat is then separated from the mucosa and sub-mucosa. Such a preparation of smooth muscle from an average frog weighs about 0.25 gm. In making my ex- tracts I have uniformly employed the method described by Fletcher and Hopkins. The muscle is crushed under 96 per cent alcohol at 0°, then ground as fine as possible and allowed to stand under the alcohol for twenty-four hours. The mixture of alcohol and muscle is then filtered, the filtrate evaporated to dryness, and the residue rubbed up with distilled water at 100° and a little animal charcoal. This mixture is again filtered and the filtrate tested in the manner to be described later. As a rule, the muscle from the stomachs of 10 frogs, weighing about 2.5 gm., was used in each experiment. Enough water was used-in dissolving the last residue to give a final filtrate of about 5 c.c. Extracts made in this manner from fresh muscle were compared in various ways with extracts similarly made from muscle which had been kept for an hour at between 40° and 50°. 7 FLETCHER and Hopkins: Loc. cit., pp. 266 et seq. 6 Edward B. Meigs. The extracts were tested with regard to their action on litmus paper and rosolic acid, and also by Ueffelmann’s test, and by a new color test for lactic acid recently devised by Fletcher and Hopkins.® All these tests showed that the extract of heated smooth muscle is distinctly more acid than that of the fresh muscle; and the two specific tests used indicate that the acid formed during the heating is lactic acid. The effect of the muscle extracts on the color of rosolic acid was compared with that of a series of standard mixtures of monosodium phosphate and disodium phosphate. The results indicate that the extract of fresh muscle has an alkalinity equal to that of a solution _ in which (H) = 4 X 10°, while the extract of heated muscle has an acidity equal to that of a solution in which (Hi) = (12 < Tous That is, the extract of fresh muscle is, in the physico-chemical sense, about as alkaline as an m/8,333,333 solution of NaOH in distilled water, while the extract of heated muscle is about as acid as an m/15,000,000 solution of HCl. I wish here to offer my thanks to Dr. L. J. Henderson for his kindness in suggesting this test for the reaction of the extracts and in providing me with the standard mixtures of monosodium and disodium phosphate. Dr. Henderson calculates that the results indicate a probable lactic acid production during the heating of from 0.1 per cent to 0.2 per cent of the weight of the muscle. These results are of course extremely rough, but they are in full agreement with the others, which in- dicate that considerable amounts of lactic acid are formed in smooth muscle when it is heated, though much less than under similar cir- cumstances in striated muscle. With Ueffelmann’s test the extracts of heated muscle give a positive reaction, while those of the fresh muscle give either a negative reaction or a positive reaction very much fainter than that obtained with the extract of heated muscle. In one case I tested the extract of heated muscle by the new color test for lactic acid recently devised by Fletcher and Hopkins. These authors believe that this test is in general more specific for lactic acid than Ueffelmann’s test and quite specific in the case of physiological material. The extract of heated muscle gives a strong positive reaction with this test. Besides testing the extracts in the manner described above, I have compared the reaction of fresh smooth muscle with that of 8 FLETCHER and Hopkins: Loc. cit., p. 308. Heat Coagulation in Smooth Muscle. 7 muscle which has been heated to between 40° and 50°, by simply pressing such pieces of muscle against litmus paper. The heated muscle is always distinctly the more acid. It might be objected that in this case the increased acidity is due to CO., but this is un- likely, as Fletcher ® has shown that a surviving muscle is continu- ally giving off CO,, which indicates of course that its fluids are always saturated with this substance. There is every reason to believe, then, that smooth muscle heated to between 40° and 50° produces lactic acid, though in somewhat less quantity than striated muscle. There is little tendency toward a change in weight in either smooth or striated muscle heated to 50° or less. Both kinds of muscle may be kept for half an hour in 0.7 per cent salt solution at 50° without undergoing any considerable changes in weight. Except therefore in regard to length, smooth muscle and striated muscle undergo similar changes when heated to between 40° and 50°. Both lose their irritability, in both the proteins are coagulated, probably at about 50°; in both considerable amounts of lactic acid are formed, and neither exhibits any marked tendency to change in weight. The shortenings which occur in both tissues at 52° and above must next be considered. It must first be pointed out that the changes in length which occur between 40° and 50° are highly characteristic of muscle, while those which occur at 52° and above are by no means so. There are very few substances which undergo a change in length of from 30 per cent to 60 per cent when heated to between 40° and 50°, but there are a great many substances which shorten quite as rapidly and quite as strongly as muscle does when heated to 52° and above. Among such substances nerve,’® tendon,'? and the cat- eut of violin strings are conspicuous. The last-named substance is, of course, prepared connective tissue, but the fact that it 1s still capable of undergoing such shortening after its preparation for commercial purposes indicates that the shortening in question is not a manifestation of any property peculiar to the state in which the tissues are found in the living body. It is a fact well known to histologists that tissues “ shrink ” when ® FLETCHER: The journal of physiology, 1898, xxiii, p. Io. 10 HALLIBURTON: Biochemistry of muscle and nerve, Philadelphia, 1904, pp. 105 and 106. 11 JENSEN: Zeitschrift fiir allgemeine Physiologie, 1908, viii, pp. 309 and 310. 8 Edward B. Meigs. heated to a temperature much above 50° in the paraffin oven. This shrinkage takes place in all dimensions, though chiefly in the longi- tudinal direction in such tissues as nerve and tendon. It may amount to nearly half the original length of the tissue. Such shrinkage is of course accompanied by a passage of the melted paraffin out of the tissue interstices, and it seems to be a very general property of organic tissues. The characteristic peculiarities of “ shrinkage ” in the histological sense*are that it takes place in tissues saturated with fluid at tem- peratures above 50°, and that it is accompanied by a passage of fluid out of the tissue interstices and by a consequent loss of weight. The later heat shortenings of muscle resemble shrinkage in the temperature at which they make their appearance and in the fact that they occur in a tissue saturated with fluid. It is an interesting question whether they are accompanied by a passage of fluid out of the tissue. This question may be answered at once in the affirma- tive. Both striated and smooth muscle exhibit a marked tendency to lose water when heated above 50°. The following experiments may serve as evidence for this statement. A frog’s sartorius was dried rapidly on filter paper and found to weigh 0.150 gm. It was then immersed for twenty-four minutes in 0.7 per cent sodium chloride solution at 50°, after which it was again dried and weighed; it still weighed 0.150 gm. It was now immersed for two minutes in 0.7 per cent sodium chloride solution at 65°, and, on being again dried and weighed, was found to have decreased in weight to 0.120 gm. An exactly similar experiment carried out on a piece of the smooth muscle from a frog’s stomach showed that this tissue suf- fered no considerable change in weight during twenty-four minutes’ immersion in 0.7 per cent sodium chloride solution at 50°. During two minutes’ immersion in the same solution at 65°, however, it de- creased in weight from 0.050 gm. to 0.035 gm. The above experiments were carried out to show that the loss of weight takes place rapidly and at a comparatively low tempera- ture. If the tissue be kept at a higher temperature for a longer time, the loss of weight may be very much more marked. In one case a piece of smooth muscle from the frog’s stomach lost one half of its original weight during about five minutes’ immersion in, physiological salt solution at 85°. The shortening of catgut at about 55° is in many respects similar Heat Coagulation in Smooth Muscle. 9 to that of muscle at the same temperature. So much stress has been laid by Engelmann '” on this subject that it seems worth while to consider it in some detail. Engelmann soaks commercial catgut in water for an hour, then heats it for two or three minutes to above 80°, after which it is ready for use. Such a piece of catgut will shorten whenever heated above 55° and lengthen on being allowed to cool again provided it be kept always saturated with water. | have followed the changes in length and the changes in weight which occur during all these processes. During the preliminary soaking the gut of course gains in weight and at the same time shortens very slightly. During the two minutes’ heating above 80°, it gains still further in weight and shortens from 40 per cent to 50 per cent of its original length. At the same time it is partly gelatinized, and the previously twisted fibrils become untwisted. If it be now immersed in water at room temperature, it gains still further in weight and lengthens markedly. The absorption of water and lengthening go on for some time, — perhaps an hour, — after which an equilibrium is reached. The gut is now in the state in which Engelmann uses it for his experiments; it will shorten and lose water whenever heated to above 55°, and lengthen and absorb water on being allowed to cool again. The loss of water on heating takes place very quickly and is quite large, often more than 10 per cent of the original weight of the catgut. The experiments which show this were carried out as follows: A piece of gut prepared for Engelmann’s experiment was dried on filter paper and weighed. It was then immersed for two or three minutes in water at about 65° and again dried and weighed. Finally it was immersed for a few minutes in water at room temperature and dried and weighed as before. Sometimes these processes were repeated once or twice with the same piece of gut. The figures in a typical experiment were as follows: A piece of catgut prepared for Engelmann’s experiments weighed 0.345 gm. The same piece after 2 minutes’ immersion in water at 67° weighed 0.310 gm. The same piece after 8 minutes’ immersion in water at 20° weighed 0.350 gm. The same piece after 2 minutes’ immersion in water at 65° weighed 0.320 gm. The same piece after 8 minutes’ immersion in water at 20° weighed 0.370 gm. It will be noticed that the catgut on being allowed to cool after the first heating gains more in weight than it lost during the heat- 12 ENGELMANN: Ueber den Ursprung der Muskelkraft, Leipsic, 1893. 10 Edward B. Meigs. ing and that this tendency continues through the subsequent heating and cooling. It is practically certain that the heating of catgut in this manner causes two different changes. One of these is an irre-. versible tendency toward gelatinization and the capacity to absorb more water. The other is the actual driving out of water already present and is reversed on subsequent cooling. That the shortening accompanying the heating is connected with the loss of water and not with the tendency toward gelatinization is shown by the fact that it is reversed on cooling when the direction of water flow is reversed. It may also be shown by an independent experiment. If a piece of catgut prepared for Engelmann’s experiment be allowed to dry, it shortens quite markedly during the drying and may be made to lengthen again by re-immersion in water. The changes which occur during the preparation of the catgut for Engelmann’s experiment seem to be of a different nature. Both the preliminary soaking and the first heating to above 80° are ac- companied by a gain in weight and a decrease in length. It seems possible that the shortening in these cases is due to the fact that the catgut is much stretched during its preparation, and fixed, as it were, in a state of extension by drying. But the nature of these phenomena has no particular bearing on the questions under dis- cussion. I only wish to show that the shortening which occurs in catgut prepared for Engelmann’s experiment is accompanied by a loss of water and that the subsequent lengthening is accompanied by an absorption of water. A remarkable peculiarity of the behavior of the catgut is the complete reversibility of the shortening and of the loss of water. These tendencies are exhibited by both striated and smooth muscle to a slight extent. Both tissues lengthen somewhat on being allowed to cool after being heated to 70° or 80°, and in both there may be demonstrated a slight tendency to reabsorb some of the water which had been lost. It must be admitted, however, that the catgut ex- hibits these tendencies much more markedly than does the muscle. The differences between Vernon's results with smooth muscle and my own must now be briefly discussed. Fig. 2 represents the changes in length undergone by a piece of smooth muscle heated gradually to 100° in a solution containing 0.67 per cent NaCl and 0.06 per cent Na,CO,. The Na,CO, was added because smooth muscle gradually loses tone in pure physiological sodium chloride solution, and it was found that this amount of alkali preserved the i meni Heat Coagulation in Smooth Muscle. II particular piece of muscle used in about the state of tone it had on immersion in the solution. I have found that pieces of smooth muscle preserve their irritability in such weak alkaline solutions for a long time. If Fig. 2 be compared with the lower curve of Fig. 1, it will be seen that in the former the preliminary lengthening begins at a 20° 30° 40° 50° 60° 70° 80° 90° 100° 20° 30? 40° 50° 60° 70° 80° 90° 100° | a a ee, = FicurE 2.— Curve showing the effects of FicurE 3.— Curve showing the effects of gradual heating on the smooth muscle of the frog’s stomach; the heating was carried out at such a rate that it required twenty-four minutes forthe temperature to rise from 20° to 100°. The muscle was weighted with 0.6gm. Magnifica- tion of writing lever, 5; length of muscle between attachments at end of experi- gradual heating on a ring of the mucous , membrane of the frog’s stomach; the heating was carried out at such a rate that it required twenty-six minutes for the temperature to rise from 20° to 100°. The tissue was weighted with 0.6 gm. Magnification of writing lever, 5; length of tissue between attachments at end of ment, 6mm.; proportional shortening, experiment, 8 mm. ; proportional shorten- 36 per cent of greatest length. Two thirds the original size. ing, 54 per cent of original length. ‘Two thirds the original size. considerably lower temperature (about 25°) and that the heat short- ening begins at a higher temperature (between 55° and 60°). I have obtained these results with a good deal of constancy, and do not know how they are to be explained; possibly they are due to differences between English and American frogs, possibly to differ- ences between the tissues of the cesophagus and those of the stomach. A more striking difference between Vernon’s results and mine is the greater rapidity and height of the heat shortening which he usually obtains. It seems possible that he failed to strip the mucous membrane from his preparations. He nowhere states that he did this, and in the case of the cesophagus it is a much more difficult operation than in the case of rings from near the middle of the stomach. 12 Edward B. Meigs. If this surmise is correct, the great rapidity and height of Ver- non’s heat shortenings may be readily explained. Fig. 3 represents the heat shortening of the ring of mucous membrane torn from the muscle ring, of which the behavior is recorded in Fig. 2. The facts which have just been reported are, to say the least, un- favorable to the coagulation theory as an explanation either of the contraction of heat rigor or of contraction in general. In the case of smooth muscle it seems impossible to: suppose that there is any close connection between shortening and coagulation. A large pro- portion of the protein of this tissue is undoubtedly coagulated at or below 50°, yet the muscle lengthens markedly when heated to that temperature. There is much evidence to show that in striated muscle also no connection exists between shortening and protein coagulation, and that the shortening which occurs in so many animal tissues at about 55° is not the result of protein coagulation in those tissues. This evidence will be presented in detail in a later article. For the pres- ent it will be enough to say that a marked shortening occurs at about 55° in striated muscle of which all the protein precipitable at that temperature has been already coagulated by the long-continued action of strong alcohol. Whatever may be thought of the shortenings which occur in muscle at 55°, there is plainly no reason to believe that they con- stitute a peculiarity of muscular tissues. It is much more probable that they are of the same general nature as the shortening which occurs in catgut at the same temperature. The changes in length which occur below 50°, however, are peculiar to muscle; and for these the following explanation may be offered. Moore and Parker,’® Lillie,t* and others have demonstrated that the addition of very small quantities of acid or alkali to certain colloid solutions greatly increases their osmotic pressure. It has been known for a long time that muscle also possesses this property, — it eventually swells when immersed in weak acid solutions even though the osmotic pressure of the solution be increased by the addi- tion of crystalloids to a point much above that of the muscle. Fischer 1° has reported some interesting results on this question, and has shown that the behavior of frog’s muscle immersed in 8 Moore and Parker: This journal, 1902, vii, p. 261. “4 Linus: Ibid., 1907, xx, p. 127. *® FiscHer: Archiv fiir die gesammte Physiologie, 1908, cxxiv, p. 69. Heat Coagulation in Smooth Muscle. 13 various solutions of acids, bases, and salts, and in distilled water corresponds quite closely to the behavior of a preparation of fibrin treated in the same manner. Fletcher and Hopkins have shown that a large production of lactic acid is a constant accompaniment of heat rigor in striated muscle. My own experiments, though far less complete and satis- factory from the chemical standpoint than those of Fletcher and Hopkins, nevertheless indicate clearly that lactic acid formation is a constant accompaniment of heat coagulation in smooth muscle. It seems highly probable, from the observations of Moore and Parker, Lillie, and Fischer, that the presence within a muscle of acid in such quantities as are formed in both striated and smooth muscle at temperatures between 40° and 50° would cause a swelling of those parts of the tissue in which the colloids are most concen- trated at the expense of the interstitial spaces. The histological evidence indicates that the colloids are most concentrated in the fibrillae or sarcostyles of striated muscle and in the fibre cells of smooth muscle. In a recent article '® I have adduced evidence to show that swell- ing of the sarcostyles of striated muscle results in their shortening, while swelling of the fibre cells of smooth muscle results in their lengthening. | In this way, therefore, all these facts may be grouped together without the aid of a single gratuitous assumption. It is certain that the heating of either striated or smooth muscle causes the pro- duction of a considerable amount of lactic acid within the tissue. There is every reason to believe that the presence of acid in such quantities would cause the swelling of the sarcostyles in the one case and of the fibre cells in the other. Finally, there is much inde- pendent evidence to show that the swelling of the sarcostyles of striated muscle always results in their shortening, while the swelling of the fibre cells of smooth muscle results in their lengthening. 16 Merrcs: This journal, 1908, xxii, p. 477. ‘ ON THE CONNECTION BETWEEN CHANGES OF PER= MEABILITY AND ‘STIMULATION AND ON] Gis SIGNIFICANCE OF “CHANGES IN PERMEABILIRY TO CARBON, DIOXIDE, By R. S: LILEITE, [From the Marine Biological Laboratory, Woods Hole, and the Physiological Laboratory Zoclogical Department, University of Pennsylvania.] INTRODUCTORY. VARIETY of evidence now exists indicating that stimulation of an irritable tissue is dependent upon a temporary and readily reversible increase in the permeability of the surface layers or plasma membranes of its cells or elements. Perhaps the most unequivocal evidence of this kind is that pre- sented by the motile organs of such plants as Mimosa and Dionzea or the stamens of Cynarez. In the sensitive plant (Mimosa pudica) the normal position of the leaves is maintained by the turgidity of the pulvini or cushion-like masses of parenchyma cells at the base of each leaflet and petiole. On stimulation this turgidity undergoes a sudden diminution accompanied by an escape of fluid from the turgid cells into the intercellular spaces which communicate with the vessel system of the plant. The pulvinus thus decreases in volume and the petiole falls to the drooping position characteristic of the stimulated plant. If the latter is undisturbed, a gradual re- covery follows (in the light). Since turgor in plant cells is due to the osmotic pressure of the cell contents acting against the semi- permeable plasma membrane (Pfeffer, de Vries), and so putting the extensible cell wall on the stretch, such a sudden collapse of the cell seems open to only two possible interpretations: (1) if the per- meability of the protoplast remains unaltered, the effect can be due only to a sudden decrease in the concentration of the osmotically active substances within the ¢éell, presumably by combination to form larger and fewer molecules; or (2) if, on the other hand, the 14 . Changes of Permeability. 15 plasma membrane becomes suddenly permeable to the dissolved substances, the resistance to the collapse of the stretched cell wall must at once disappear, and the movement necessarily follows this disturbance of equilibrium. That the latter change is the main if not the sole determinative factor is indicated by several considera- tions: first, the invariable rule is that the cells regain their turgor only slowly; 7. e., time appears to be required for the production of osmotically active substances, indicating that these have been lost from the cell’ and not merely rendered less active by combination to form larger molecules; in the latter case the reverse or splitting process would presumably take place with equal velocity and we should expect the curve of restoration of turgor to correspond with that of loss. The fact that the rising and falling portions of the curve are not symmetrical —as in the case (approximately) with muscular contraction — may be interpreted as indicating a mecha- nism not depending primarily on rapid chemical change. Second, similar changes of turgor leading to similar changes in the position of the leaves occur in practically all plants when the cells die; the loss of turgor in this case is accompanied by a demonstrable increase in the permeability of the cell and is unquestionably due to this. ° This change, in fact, differs from the preceding chiefly in not being reversible. Again, heat, poisons, and other destructive influences produce simultaneously increase of permeability and loss of turgor in all plant cells, and corresponding movements result in the case of motile organs. The dependence of turgor changes on changes in permeability is in fact clearly recognized to-day by the majority of plant physiologists. The electrical change following stimulation in motile plant tisstes is similar, as shown by Burdon-Sanderson,? to that seen in animal cells; 7. ¢., a negative variation occurs, —as also in dying tissues, where the permeability evidently undergoes an increase. This agree- ment indicates a fundamental similarity in the stimulation process in both classes of organisms. It is now generally egreed that the source of these potential differences in living tissues — systems con- 1 The fluid which exudes from the pulvinus during stimulation is known to be not pure water, but a solution of considerable concentration (PFEFFER). Cf. Jost: Lectures on plant physiology, Oxford, 1907, p. 515. * A summary of BuRDON-SANDERSON’s work with references to his original papers is given in BIEDERMANN’s Electro-physiology, English translation, ii, ch. 6, pp. 1 et seq. 16 eS; Laiite: taining only electrolytic conductors —can be only some type of concentration cell; and since the potential differences are in many cases too great to be accounted for without assuming that certain ionic velocities must be markedly different in living tissues from those observed in homogeneous solutions, it is assumed that the membranes in tissues play an important réle in changing the ionic velocities, depressing some and (possibly) increasing others. The theory of Bernstein,® based on a suggestion of Ostwald, that the semi-permeable membranes of resting cells are freely permeable to the cations of the electrolyte to which the potential difference is due, but not to its anions, affords a satisfactory explanation of the changes resulting from stimulation. If the permeability to anions increases during stimulation, the observed fall of potential between exterior and interior of the cell becomes at once intelligible. Now at death, when the permeability evidently increases, there is a fall of potential similar to that seen during stimulation, but permanent in this case. The simplest and most probable inference to be drawn from this fact is that a similar increase of permeability occurs mo- mentarily during stimulation. This conclusion is confirmed and re- inforced by the phenomena observed in motile plant cells, as we have just seen. In the case of animal cells the evidence is in general of a less direct nature than in plants. Certain fundamental facts plainly have the above significance. There is the same increase in the perme- ability of the cell at death, as indicated by readier entrance of dyes, loss of susceptibility to plasmolysis, and outward diffusion of color- ing matters formerly confined within the plasma membrane. This post mortem increase of permeability is accompanied by an electrical change of the same nature as that resulting from stimulation, and in some cases, as in muscle, by contractile changes also; the basis of these latter appears to be a coagulation of the muscle proteins {Kuhne); coagulative changes in the protoplasm are in fact very generally associated with the post mortem increase of permeability and presumably result from it, and such changes in some instances result from excessive stimulation.4 Again, conditions that increase the permeability of the surface layers have in general also a stimu- lating action: heat, mechanical influences, electricity, action of vari- * J. Bernstern: Archiv fiir die gesammte Physiologie, 1902, xcii, p. 521. * Cf. the discussion in my former paper on this subject: This journal, 1908, xxii, Pp. 75, especially pp. 81-83. Changes of Permeability. 17 ous chemical reagents. The effect of these agencies in increasing permeability is most readily seen in the case of pigment-containing cells such as blood corpuscles, pigmented ova, and the cells of certain organisms like Arenicola larvee, which last show this interdependence with remarkable clearness. The cells of this organism contain large quantities of a water-soluble yellow pigment; and, as I shall describe later, conditions that produce energetic and persistent muscular con- tractions (action of various salts and fat solvents) invariably cause a visible loss of pigment from the cells. It has long been known that the electric current, the most universal stimulating agent, 1n- creases the permeability of blood corpuscles, as shown by the laking action of induction shocks and condenser discharges.® Further, the view of Overton,® that the plasma membrane owes its specific per- meability to lipoid substances, explains the stimulating action of fat solvents,’ and its nature as colloid layer accounts for the action of electrolytes and explains why such substances both stimulate and alter permeability.§ Certain other well-known facts, which, so far as I am aware, have not hitherto received attention from the present point of view, strengthen the theory that stimulation is a consequence of temporary increase in permeability. Electrical stimulation almost certainly depends on localized changes in ionic concentration within the irrit- able tissue. Several years ago Nernst propounded the view that these changes in concentration have their seat at the various semi- permeable surfaces of the tissue, 1. e., at the plasma membranes of the cells, which the researches of de Vries and Overton have shown to be impermeable ® in the resting cell to the electrolytes normally present in tissues. Such membranes must therefore offer a barrier to the movement of ions when a current is passed. As Nernst has ® Cf. Hermann: Handbuch, 1880, i, pp. 14 ef seg. for an account of these phenomena. 6 For the application of this theory to the case of muscle, cf. OVERTON: Archiv fiir die gesammte Physiologie, 1902, xcii, p. 115. 7 It appears also highly probable that cytolytic substances like toxines owe their stimulating action, as seen in fever or more obviously in the contraction of tetanus, to their influence on the permeability of nervous or muscular elements. 8 Cf. HOpErR’s interesting discussion, Physikalische Chemie der Zelle und der Gewebe, 2te Aufl., Leipzig, 1906, ch. 6, “Die Permeabilitit der Plasmahaut,” and ch. 8, ‘“‘ Wirkung reiner Elektrolytldsungen,” especially pp. 270 et seq. ® The experiments demonstrate either impermeability or slow and difficult per- meability to neutral salts of alkali or alkali earth metals. 18 ee Oe Lae, shown, the observed inverse proportionality between the stimulating action of an alternating current and its rate of alternation may be theoretically accounted for on the above assumptions.’? On the present theory these changes in ionic concentration cause stimulation by altering the colloidal consistency of the plasma membrane and thus increasing its permeability, this last effect being the im- mediate condition of the stimulating action. The essential feature of Nernst’s hypothesis — that stimulation is conditional on imper- meability (absolute or relative) to ions — is, however, fundamental to any theory of the nature of electrical stimulation. Now it is obviously a corollary of this hypothesis that if during stimulation the plasma membrane is so changed as to become freely permeable to ions, further stimulation should be impossible at such times. This is in fact the case, as is shown by the persistence, for a greater or less interval after stimulation, of a period during which the tissue is non-responsive or refractory to stimulation. This condition, as all physiologists are aware, is most easily demonstrated in heart muscle; but it exists even for highly irritable tissues like nerve.14 The refractory period, on this interpretation, is simply a period of increased permeability, during which the action of the electrical stimulus, which depends on the semi-permeability of the surface layers, is accordingly suspended. In other words, the refractory period is to be considered as evidence of a loss of semi-permeability at the height of stimulation.’? A further argument may be drawn from the striking analogies which the work of Bredig, Weinmayr, Wilke, and Antropoff 1° have shown. to exist between the phenomena of pulsatile catalysis in the case of mercury surfaces in contact with hydrogen-peroxide solu- tion, and the rhythmical process in organized tissues like cardiac muscle. The curves representing the variations in the velocity of © Nernst: Archiv fiir die gesammte Physiologie, 1908, cxxii, p. 275. " For the case of skeletal muscle cf. BAzETT: Journal of physiology, 1908, xxxvi, p. 414; of nerve, cf. Boycott: zbid. 1899, xxiv, p. 144. * There may be only a marked diminution in the normal semi-permeability, not an absolute loss; according to CARLSON the so-called refractory period is only relatively so (This journal, 1906, xvi, p. 67); 7. e., the excitability is merely Jowered, to a degree varying in different animals. ; *S BrepDiG and WernMAyr: Zeitschrift fiir physikalische Chemie, 1903, xlii, p. 601; Brepic and WILKE: Biochemische Zeitschrift, 1908, xi, p. 67; BREDIG: Tbid., 1907, vi, p. 322; ANTROPOFF: Zeitschrift fiir physikalische Chemie, 1908, Ixii, P. 513. Changes of Permeability. 19 oxygen liberation from such surfaces show a surprising similarity to the ordinary sphygmogram or cardiogram; and parallel with the rhythm of oxygen production runs a corresponding rhythmical variation in the electrical potential difference between the peroxide solution and the mercury surface. Changes of surface tension also accompany the changes in electrical potential. The analysis of the phenomena given by Antropoff makes it highly probable that the source both of the electrical rhythm and of the oxygen evolution is a periodic electrolytic dissolution of the film of mercury peroxidate formed over the surface of the mercury by the action of the per- oxide; the disappearance of this film is preceded by the appearance of a mechanically induced rupture which exposes the unaltered metallic mercury below, between which and the external surface of the film there exists a potential difference estimated at ca. 0.12 volt. At the margin of this fissure a circuit is thus formed by whose action the film undergoes electrolytic reduction to metallic mercury, and the change thus induced travels over the whole surface. The phe- nomenon is repeated in regular rhythm. The especially significant feature from the standpoint of the physiology of stimulation is that the primary occasion for this dissolution of the film is a break m its continuity. This corresponds to a localized increase in per- meability; in consequence of this change the condition for the re- duction of the film and the liberation of oxygen is propagated over the entire surface of the mercury. On the theory of stimulation advocated in the present and pre- ceding papers, the primary change in stimulation is also an increase in permeability at the point of stimulation. This results at that point in a localized depolarization which travels thence over the sur- face of the irritable element and is accompanied by an increased permeability. The conduction of this state of depolarization is an essential feature of the stimulation process, which otherwise would have an effect only at the exact point of application of the stimulus. The analogy of the mode of propagation of the surface change in the film-covered mercury drop — which is accompanied by a similar electrical disturbance — points to the possibility that in the irritable tissue the point of stimulation acts as cathode on the adjoining por- tion of plasma membrane and depolarizes this which in its turn serves as stimulating cathode for the next layer, and so on.'* On 1 In a more general form this view has long been held. Cf. Gorcn’s article in ScHAFER’s Text-book, ii, 1900, pp. 458-459. The conception that permeability 20 ROS. ihe: this view the progressive dissolution of the surface film in the mercury drop and the similarly progressive depolarization with accompanying increase of permeability in the irritable tissue are strictly analogous and similarly conditioned phenomena.’ The common feature of the above two processes, otherwise so widely disparate, thus appears to lie in an automatic and periodic alteration of the surface layer in the direction of increased per- meability, — in the one case the film being temporarily removed by the accompanying electrolytic process, in the other merely altered so as to acquire the increased permeability needed for depolarization and consequent stimulation. That periodic alterations in the ionic permeability of cardiac muscle accompany the rhythm of contraction is indicated by the action current; this must correspond to an ionic transfer between interior and exterior of the fibre, following a rhythm parallel with that of the stimulation process. The sim- ilarity in the surface process in the two phenomena under considera- tion is thus the ground of the similarity in the respective curves. If the carbon-dioxide output of the beating heart could be measured by a similar device to that used by Antropoff for the oxygen evolu- tion from the mercury surface, it would doubtless be found to ex- hibit a similar rhythmicity corresponding to the rhythm of the contraction and of the potential change. An argument from analogy or from a resemblance which a closer examination may show to be superficial may not in itself be very convincing; the present comparison may, however, serve to re- inforce the preceding considerations. In brief, the situation stands thus: the automatic rhythm of cardiac muscle, and presumably of a cilium or an automatic nerve centre, is associated with an electrical disturbance which is most readily explained on the ground of a periodic increase in surface permeability. If an inorganic phe- nomenon showing such strikingly similar characteristics as the above is actually shown to be dependent on an automatic and periodic alteration of the surface layer, the grounds for assuming a similar changes are the essential feature of the stimulation process is, however, purely an outcome of modern physico-chemical analysis. It has not, however, been shown that such an hypothesis can account for the high velocity with which the depolarization wave is conducted over the surface of many living tissues, particularly nerve; and until this is done the present suggestion has perhaps only the interest of a possible, though in my opinion also a highly prob- able, explanation of the nature of conduction in irritable tissues. Changes of Permeability. 21 condition in the case of the organized tissue are certainly strength- ened. Since the only essential distinction between the phenomena of stimulation in cardiac muscle (cilia, etc.) and in skeletal muscle or nerve is the automaticity and regular periodicity of the former under physiological conditions, any general conclusions reached from the above comparison must apply also to the phenomena of stimula- tion in general. EXPERIMENTAL. During the past summer at Woods Hole I have continued the study of the processes of stimulation and contraction in the larve of Arenicola cristata. The simple embryonic type of musculature possessed by this organism seems particularly fitted for the investi- gation of the more general and fundamental conditions of these processes. The following have been the chief questions under con- sideration: (1) the nature of the change induced by salt solutions that stimulate powerfully (isotonic sodium and potassium salts, etc.); (2) the nature of the inverse action shown by pure non- electrolyte solutions, anzesthetics like chloroform in low concentra- tion, and certain salt solutions (e. g., isotonic magnesium salts) which temporarily deprive muscles of contractility; (3) the nature and concentration of the electrolytes that restore contractility to muscles deprived of this property by solutions of non-electrolytes or magnesium salts, and the nature of this action; (4) the conditions under which lipoid solvents, like ether, chloroform, benzol, etc., produce stimulation instead of anesthetization; (5) the influence of carbon dioxide on permeability and contraction. These experiments are incomplete and will be continued. In the present paper I propose to describe in detail only those observations which appear to indicate clearly an increase in permeability during stimulation, or a decrease during inhibition, either by showing a visible transfer of substance between interior and exterior of the cells during strong stimulation, or, inversely, by showing that during anesthesia (depressed irritability) the entrance of substances from outside (e. g., dyes) or the exit of substances from the cells is less rapid than during muscular activity. In a later paper I shall describe experiments bearing more directly on the other problems cited above. 22 RS. Edhe: In an early paper 1* I have given a brief description of the appear- ance, organization, and behavior of Arenicola larve at the swarming stage. The body musculature is of a simple embryonic type, and consists at this stage exclusively of longitudinal fibres, no trace of circular muscles being visible until much later in development. The muscle cells are without definite boundaries; they consist at first of a mesenchyme which becomes applied to the ectoderm, and on the side adjoining this latter layer longitudinal fibrils early ap- pear in the mesenchyme cells immediately beneath the cell surface. The fibrils thus show the superficial position within the cell char- acteristic of primitive or embryonic muscle fibres; they are un- striated and when fully formed relatively thick in comparison with the diameter of the muscle cell, indicating a progressive transforma- tion of the sarcoplasm into the substance of the fibril.17 The absence of circular fibres at this stage indicates that both elongation and contraction are active properties of the muscle (as Sherrington has shown for Vertebrata). The state of extension of the larva depends on the tone of the musculature, and this is variable and largely de- termined by the reaction of the external medium, slight acidity pro- ducing increase and alkalinity decrease of muscular tone.'® Another peculiarity of the larva is the presence in its cells of large quantities of a yellow pigment, which appears at this stage diffused throughout the entire organism. This pigment is derived from the egg cell which is similarly colored, and disappears in later larval stages. The individual larva as seen by transmitted light is semi-opaque and light yellowish brown in appearance; when densely massed together, as in consequence of their positive phototaxis, the larvee exhibit a characteristic dark brown tint. When large numbers of larve are treated with cytolytic sub- stances like chloroform, or when they die, or are caused to contract strongly by the action of certain solutions (isotonic NaCl, KCl, NH,Cl, etc.), the pigment leaves the cells and colors the water a © R.S. Lizzie: This journal, ror, v, p. 56. For a fuller account of the structure and development of the musculature of Arenicola larve, with figures illustrating the above-described conditions, cf. my paper “On the structure and development of the Nephridia of Arenicola,’’ in Mit- theilungen aus der zoologischen Station zu Neapel, 1905, xvii, p. 341. The structure of the young swarming larva is somewhat fully illustrated in the figures on Plates 22 and 23. 'S These effects will be described in a later paper. Changes of Permeability. 23 light yellow. An increase of permeability is thus indicated. It must be assumed in the latter case that the pigment is derived from all of the pigment-containing cells and not merely from the muscles; 1. e., the increase in permeability caused by the salt is an effect ex- hibited by all the cells in common, although contraction is presum- ably confined to the muscle cells. The exit of pigment is of im- portance as indicating an increased permeability; that it 1s accom- panied by strong contractions is therefore significant. In the other cells of the larva no immediate evidence of change is seen under these conditions; but later coagulative and cytolytic alterations follow, and the entire organism swells and disintegrates: the con- nection between death coagulation and increase of permeability is thus shown. In general, any change that induces a rapid loss of pigment causes at the same time energetic muscular contraction. During the nor- mal movements of the animal there is no such loss; in this case it is to be supposed that the stimulating action is confined to the neuromuscular system of the organism and that the majority of the pigment-containing cells are unaffected. It should be added that the contraction which accompanies rapid loss of pigment (as in m/2 KCl) is ‘always far more energetic than any normal movement — the larva shortening to fully half its normal length and remain- ing thus contracted for several seconds. Conditions that produce an increase of permeability sufficient for rapid loss of pigment thus act as unusually intense stimuli. Such conditions are also highly injurious to the organism. In the following experiments the procedure has been similar to that already described in my former papers. The larve, which are strongly heliotropic, are allowed to gather in a mass on the light side of the watch glass. The sea water is then poured off, and its last traces are removed by filter paper; the solution is then added. Action of salt solutions. — Pure isotonic NaCl solutions (m/2 — ¥g m) are rapidly destructive, as I have already described.!® The effect may readily be observed under a magnification of 50 diame- ters.*° At the instant of contact with the solution the organism con- tracts suddenly and violently to about half its normal length. The Wok: S. Jorn ies! 06.» crt: 20 The larve average about one third millimetre in length. 24 RS ihe, cilia for the most part undergo rapid dissolution or liquefaction in pure solutions of Na salts, so that swarming is at once arrested. The strong initial contraction is followed by slow relaxation accom- panied by feeble bending movements, and within one or two minutes the organism regains its normal length. Feeble muscular contrac- tions may last for some minutes longer, but soon cease. The phe- nomena indicating increased permeability are as follows: in a few seconds after the addition of the solution the latter is observed, especially if numerous larve are present, to be colored by the yellow pigment from the cells; this substance leaves the cells during the initial period of strong stimulation. Under the microscope a shrink- age is seen to begin during this period, and a somewhat refractive and apparently glutinous fluid exudes from the body surface; the latter within three or four minutes exhibits a distinct separation from the thin cuticular layer which invests the entire organism. At the same time the surface consistency of the larve is altered so that they stick to the glass and to one another (agglutination effect ).21_ The action of the pure sodium chloride solution is highly injurious, and recovery of contractility on subsequent transfer to sea water is gradual and incomplete. Similar effects are produced by pure isotonic solutions of other neutral salts of sodium and by salts of the other alkali metals (K, Rb, Cs, Li, NH,). Potassium salts (e. g., m/2 KCl) produce an especially energetic initial contraction with rapid exit of pigment ; the larva shortens and thickens at first to an almost spherical form; gradual relaxation then follows, and the larvz continue to swim actively by means of the cilia, which remain unchecked in such solu- tions. Muscular contractions are thenceforth absent and clumping results.2? Similar effects differing in detail —e.g., NH, salts are more toxic than the others —are seen in solutions of the other alkali chlorides. The above effects of pure sodium chloride solutions can be pre- vented by the addition of a little calcium or magnesium chloride or 1 Tn Arenicola larve this effect almost invariably accompanies marked increase of permeability (as indicated by loss of pigment). Agglutination thus apparently de- pends on exit of adhesive substances from the cells through the altered plasma mem- brane. Hence it is very generally associated, in small cells like bacteria, spermatozoa, blood corpuscles, and ova, with the action of cytolytic substances. Cf. ZANGGER: Ergebnisse der Physiologie, 1908, vii, p. 138. * This phenomenon is described and figured in my early paper already cited. Changes of Permeability. 25 other favorable salt (salts of various other bivalent metals). Cal- cium salts are the most favorable. Thus, if a mixture consisting of 24 c.c. m/2 NaCl + 1 cc. m/2 CaCl, is added as above, the larve show no initial contraction and exhibit at first and for several minutes afterwards a typical heliotropic swarming; later the move- ments become irregular, the cilia are checked, and death results within twenty-four to forty-eight hours. The permeability of the tissues remains at first apparently unimpaired and there is no visible loss of pigment; later, after the contractions have lost their normal character, the pigment begins slowly to diffuse from the tissues. A similar checking of the permeability change, combined with pres- ervation of contractility for many hours, results from the addition of a little m/2 MgCl, though this is less favorable than CaCl,. If both of these chlorides are added to NaCl solutions in favorable proportions, normal permeability and normal contractions persist for greatly prolonged periods.?* Antitoxic action under these condi- tions evidently consists largely in preventing the abnormal increase of permeability induced by the pure sodium chloride.*4 An action apparently the exact reverse of that shown by sodium or potassium chloride is seen in solutions of certain other salts, par- ticularly those of magnesium. In these there is no initial muscular contraction with loss of pigment. Larve placed in m/2 or m/3 MgCl, (the same is true of Mg(NO;). or MgSO,) undergo, on the contrary, a gradual and progressive loss of muscular contrac- tility; during the first few seconds the body bends from side to side in an apparently normal manner and the swimming movements are heliotropic; the contractions become by degrees more and more limited and after a period of one to two minutes cease completely ; the larve thenceforward remain entirely rigid and free from mus- cular contractions during their stay in the solution; the cilia, on the contrary, continue actively vibrating, if at a somewhat slower rate than normal. Muscular contractility may, however, be restored im- mediately and perfectly by transfer to sea water or other favorable medium. The action is thus similar to that of an anzsthetic; m1x- tures of m/2 NaCl and m/2 MgCl, containing magnesium in pro- portions of 4 parts NaCl to 1 part MgCl, and higher, induce a 3 Cf. R. S. Liru1E: This journal, 1902, vii, pp. 30-31. 74 Martuews has also referred the antitoxic action of salts under certain condi- tions to their influence on the permeability of the cell: This journal, 1905, xii, pp. 439 et seq. 26 R. S. Lilhe. similar loss of muscular contractility, although more gradually. A striking difference from the action of sodium or potassium chloride solutions is also seen in the fact that there is no loss of pigment from the cells; solutions of magnesium chloride containing large num- bers of larvee remain absolutely clear and colorless for hours. The appearances suggest strongly that the only essential change pro- duced by this salt is a complete cessation of interchanges (ionic and otherwise) between medium and tissues; this would result from a decided decrease in permeability. The lack of injurious action confirms this supposition; even after lying in a state of com- plete muscular immobility for twenty-four hours or longer in m/2 MgCl, larve promptly resume contractions on transfer to sea water ; whereas even a short stay in m/2 KCl or NaCl is highly injurious.”° Magnesium chloride larve are also peculiar in exhibiting not the slightest trace of a tendency to cohere or to stick to the glass; 1. e., there is no exudation from the ectoderm cells, and the organisms lie quite loosely and freely in contact with the glass and one an- other. The entrance of dyes like methylene blue is also decidedly retarded, though not altogether prevented. These facts indicate that the primary action of pure NaCl or KCl solutions is to increase, that of MgCl, to decrease, the normal physi- ological permeability. The relations, however, appear more com- plex in the case of potassium salts. Thus the addition of consider- able quantities of m/2 KCl to otherwise favorable mixtures of m/2 NaCl and m/2 CaCl, (e. g., in the proportions 80 c.c. m/2 NaCl + 5c.c. m/2 CaCl, + 15 c.c. m/2 KCl) results in marked muscular paralysis; such inhibiting action is in fact typical of the general pharmacological action of potassium salts2® and on the present theory would indicate a general influence — in the presence of the other salts normally present —in decreasing permeability. When applied in relatively concentrated pure solution to the tissues, the effect, however, as in the case of most injurious actions, is to in- crease permeability, as just seen. This result agrees with the con- clusions of Hober ?* based on the influence of salts on the demarca- 7° R. S. Lizzie: This journal, 1902, vii, p. 45. 6 For a description of the effect on Arenicola larve of increasing the potassium- content of otherwise favorable solutions, cf. the paper just cited, pp. 28-30. *7 Hoper: Loc. cit. Potassium salts in isotonic solution also produce marked permanent contraction or increase of tone in frog’s skeletal muscle. Stimulation thus accompanies the increased permeability which H6BER’s results indicate. Changes of Permeability. 27 tion current in frogs’ muscle. Why potassium salts should show in relatively low concentrations such pronounced specific action in in- hibiting stimulation processes remains unexplained. Pure solutions of calcium chloride produce on Arenicola larvze effects somewhat resembling those of magnesium chloride, impairing muscular contractility, but without removing it completely as in the case of the latter salt. The influence of calcium on permeability is peculiar; pigment leaves the cells very slowly in pure isotonic or hypotonic CaCl,, while slight and limited muscular contractions persist for some time. In combination with sodium salts calcium is highly favorable to the preservation of muscular contractility.?* In pure m/2 SrCl, and BaCl, the larve contract strongly and lose pigment. These salts differ from MgCl, and CaCl, in produc- ing marked increase of permeability and stimulation in pure isotonic solutions. They may also, in association with sodium chloride, exert antitoxic action in low concentrations.2® Here the effect is to check the increase in permeability resulting from the action of the latter salt. An action essentially similar to that of m/2 MgCl, is seen in pure isotonic solutions of non-electrolytes (m-sugar, m-glycerine), and in sea water containing anesthetics (chloroform, ether, benzol, etc.) in appropriate (not too high) concentrations. The muscular contrac- tions become gradually less and less pronounced, without initial con- traction and without loss of pigment. These solutions, it is to be inferred, also act by decreasing the permeability of the irritable ele- ments.*° The fact that cilia preserve their activity in these solutions is of interest as indicating that the surface permeability of these structures is less readily affected than in the case of muscle. In es S. Lint: Loc. cit. 29 R. S. Litre: This journal, 1904, x, p. 419. 30 In the case of inhibition, anelectrotonus, and similar conditions it is to be as- sumed that a similar decrease of permeability takes place. I have already advanced this explanation, in attempting to account for the influence of calcium salts in fur- thering mechanical inhibition in the Ctenophore swimming plate. Cf. This journal, 1908, xxi, p. 200. The existence of a positive electrical variation accompanying vagus inhibition in heart muscle (first shown by GASKELL) indicates a permeability change in a direction the reverse of that accompanying stimulation. Cf. below, p. 494. Brinincs found that immersion of frog’s muscle in isotonic sugar solution increased the demarcation-current potential by heightening positivity at the longi- tudinal surface. This effect, on the present view, also indicates decreased per- meability: Archiv fiir die gesammte Physiologie, 1907, cxviil, p. 409. 28 R. S. Lillie. general cilia appear to be resistant and not especially irritable struc- tures; they require relatively high concentrations of indifferent nar- cotics for complete anzsthesia.*! The degree of impermeability con- ferred on the tissues of Arenicola larve by isotonic non-electrolyte solutions is, however, relatively slight as compared with that pro- duced by magnesium chloride solutions; muscular contractions dis- appear more slowly and less completely, and restoration of contrac- tility is more readily effected. The action of electrolytes in restoring contractility under such conditions will be treated in a separate paper. The following series of experiments (Table I) will illustrate the foregoing general observations: TABLE I July 13, 1908. — Nearly equal quantities of larve were collected by heliotro- pism in the seven watch glasses of the series; and after removal of the sea water ro c.c. of each solution was added. Time of addition of solutions, 11.45-11.54 A.M. The results were as follows: 1. m/2 LiCl. Strong initial contraction. At 11.55 the larve show only slight coherence, and the solution is colored a faint yellow. 2. m/2 NaCl. Strong initial contraction. At 11.56 the larvae cohere and stick to the glass more strongly than in m/2 LiCl, and the solution is colored a deeper yellow. 3. m/2 KCl. Very strong initial contraction with slow relaxation. At 11.57 the larve cohere in clumps and the solution is colored quite bright yellow. 4. m/2 NH,Cl. Strong initial contraction. At 11.58 the larve cohere as in m/2 KCl: solution a bright yellow. 5. 24 vols. m/2 NaCl + 1 vol. m/2 CaCl,: solution added 11.48. No initial contraction; active heliotropic swarming for first minute or two and solution remains colorless; by 11.51 a very faint yellow tinge is imparted to the solution. At 11.58 solution is tinged faintly yellow (much less so than in Solution 1); muscular contractions continue; larve stick quite strongly to the glass. 6. m/2 CaCl,: solution added 11.51. Larvae swim actively, soon be- coming rigid and collecting in clumps; solution remains colorless at first; by 11.59 a slight tinge of yellow is perceptible. 7- m/2 MgCl,: solution added 11.53. Active swimming; no initial contraction; muscular movements soon cease and larve collect in clumps. Solution remains quite colorless. ** OvERTON: Studien iiber die Narkose, Jena, rgor, pp. 7, 185. Changes of Permeability. 29 At 12.05 the contents of the watch glasses were transferred separately to a series of seven narrow test tubes of uniform diameter, for more ac- curate comparison of the color test. At 12.10 the appearances were as follows: Solutions 2, 3, and 4 are all quite deeply and almost equally tinged with pigment, presenting a light straw-yellow tint; Solution 1 has the same tint but fainter. In contrast to these solutions the series 5—7 appear practically colorless; a faint trace of pigment is present in So- lution 5, and somewhat less in Solution 6; 7 is quite colorless. There has thus been a decided increase in permeability in the pure solutions of the alkali chlorides, and little or none in the others; the presence of CaCl, has prevented the action of the NaCl. A comparison of the ap- pearances of the massed larve at the bottoms of the test tubes yields a confirmatory result; in Solution 7 (m/2 MgCl,) they show the char- acteristic dark brown color; in Solutions 5 and 6 they are also brown but slightly lighter (5 lighter than 6); while in the first four solutions they have become light yellow, having evidently lost a large part of their pigment. A second similar series on July 14 yielded an essentially identical re- sult. These additional solutions were also tested: (1) 24 c.c. m/2 NaCl + 1 c.c. m/2 MgCl,, (2) m/2 SrCl,, (3) m/2 BaCl,. In the last two solu- tions the effect was similar to that produced by pure m/2 NaCl: the larve showed a marked initial contraction, and the solution was instantly tinged yellow by the escaping pigment. After an hour the series of test tubes showed the following appearances: the pure m/2 solutions of ere) NaCl, KCl, NH,Cl, SrCl,, and BaCl, were all tinged light-straw yellow; m/2 CaCl, showed a very faint yellow tinge, while m/2 MgCl, was color- less. Of the two mixed solutions both were faintly colored, — the Na + Mg mixture distinctly more so than the Na + Ca, which was almost colorless; this difference corresponds to the difference in favor- ability, the Ca-solution being decidedly the superior of the two. For comparison the action of the above solutions on another form of pigment-containing cell was tested, namely, the unfertilized eggs of the sea urchin Arbacia, which are deeply laden with a bright red pigment. The eggs were found to differ from the larvee in suffering a loss of pigment in m/2 MgCl, and in certain other points of detail, but, on the whole, the effects were similar to the above. NaCl shows greater toxicity than KCI; ® a little m/2 CaCl, prevents the action 82 J. Lore found this true for the fertilized eggs of both Fundulus and Arbacia; of. This journal, 1900, iii, p. 439. 30 IOS.) Lathe: of m/2 NaCl in increasing permeability, as in Arenicola. While there is thus a general agreement, the results indicate that consider- able variety in the conditions influencing the permeability of cells exists in different organisms. Action of solutions containing lipoid solvents. — The stimulating or inhibiting action of electrolytes is to be referred to their action in changing the aggregation state of the colloids composing the plasma membrane and thus changing the permeability of the latter. An- other class of substances produce similar alterations in the perme- ability of the plasma membrane by virtue of their solvent action on its lipoid constituents; these are the organic fat solvents (ethers, esters, alcohols, normal and substituted hydrocarbons), which, as Overton has shown, act as narcotics or anzsthetics in low concen- trations. In higher concentrations, on the other hand, these sub- stances first stimulate the cell, and their further action rapidly pro- duces irreversible, partly coagulative alterations and death (cytolytic action). In the former case, therefore, such substances, on the pres- ent theory, decrease, in the latter, increase, permeability. I have accordingly begun experiments with this class of com- pounds. In general the result has appeared that such substances as - ether, chloroform, benzol, etc., inhibit muscular contractions, 1. é., act as anesthetics, in low concentrations; if, however, the concentra- tion is increased above a certain maximum, the effect is to produce a strong contraction or increase of tone accompanied by exudation of pigment and adhesion to the glass of the vessel. Thus, marked increase of permeability induced by this class of substances is also associated with strong stimulation. The action of solutions suff- ciently concentrated to produce these effects is largely irreversible and hence highly injurious. The following experiments illustrate the action of relatively con- centrated solutions of ether and chloroform in producing contrac- tions with associated loss of pigment. These solutions were made in m/2 MgCl,; larve that have lain for a few minutes in pure solu- tions of this salt exhibit absolutely no trace of muscular movement; the stimulating action of anzsthetics dissolved in this solution is thus rendered the more evident and unmistakable. It should be understood that these substances have a similar powerfully stimulat- ing and destructive action when dissolved in sea water or other in- different medium. Changes of Permeability. 31 The following table describes the results of a typical series of experiments : TABLE II. July 8, 1908. — Larve were transferred to m/2 MgCl, in the usual manner at 9.52 P.M. Loss of muscular contractility follows in a few seconds, as usual, while the cilia remain active. Larvee were transferred to the fol- lowing solutions at the times designated: 1. Saturated solution of ethyl ether in m/2 MgCl,. 10.26. Larve immediately contract to about half normal length; cilia cease at once. Gradual relaxation follows, which is almost complete intwo minutes. Lar- ve adhere to the glass, and yellow pigment soon diffuses from the cells. At 11.40 the protoplasm has an opaque and coarse (coagulated) appear- ance. Addition of fresh sea water at 11.45 produces nocontractions: larvee are.dead. 2. m/2 MgCl, two thirds saturated with ether (saturated solution + one half its volume m/2 MgCl,). 10.31. Action is less decided than in Solution 1; no immediate result is evident, but gradual and incomplete contractions soon appear in a relatively small proportion of larvae; cilia continue activity in a fair proportion. Larva show some adhesion to glass and slight separation of pigment. By 11.40 there is little change in appearance, and transfer to sea water produces contractions in a fair proportion. 3. m/2 MgCl, one half saturated with ether. 10.36. Cilia continue actively, and no muscular contractions are seen. No noticeable adhesion to glass or separation of pigment. On return to sea water at 11.50 well- marked contractions result. [In another experiment on July 6 larve showed well-marked contrac- tions with loss of pigment and adhesion in one half saturated ether solu- tion after one hour forty-eight minutes in pure m/2 MgCl,. The exact conditions of stimulation probably vary with temperature and length of stay in m/2 MgCl, etc.] 4. m/2 MgCl, saturated with chloroform. 10.43. Larve all contract immediately to half normal length by a quick, steady, uniform contrac- tion. Cilia cease at once. Gradual relaxation follows. Yellow pigment diffuses from larv, and latter adhere to glass. In an hour larve are dead and coagulated. ; 5. m/2 MgCl, two thirds saturated with chloroform. 10.49. Cilia cease, and larve all contract vigorously to half length as in Solution 4, but somewhat less promptly. Larve lose pigment and stick to glass. Relaxa- tion as before. In an hour all are dead and coagulated. 6. m/2 MgCl, one half saturated with chloroform. 11.05. Cilia cease. 32 eS Eahe: Muscular contraction begins after several seconds and is less rapid than in Soiutions 4 and 5; thecontracted state lasts longer (one minute or more) and the shortening is less extreme; pigment diffuses from larve, and latter stick’to glass. Larvae show coagulated appearance in thirty-five minutes. 7. m/2 MgCl, one third saturated with chloroform. 11.14. Cilia cease at once. Muscular contractions begin after an interval of ca. fifteen seconds and have a more normal character, are slower and less energetic than in Solutions 4-6 and last longer (four to five minutes}. Loss of pigment is more gradual; is not evident for the first few minutes. The adhesion to the glass is relatively slight, and the larve preserve normal appearance for some time. 8. m/2 MgCl, one fourth saturated with chloroform. 11.29. Cilia mostly cease. No muscular movement is seen until after two minutes, when well-marked contractions appear in a few. These contractions are less energetic than in the above solutions and cease in two or three minutes; many show no contractions. There is only slight loss of pigment and ad- hesion in this solution. Saturated solutions of xylol, benzol, and toluol in m/2 MgCl, were found to exhibit relatively slight stimulating action. Xylol produces no apparent effect on either ciliary or muscular movement; toljuol has a some- what more pronounced action, and slow bending movements appear in some few instances (not usually), while cilia are gradually checked; benzol solutions check cilia rapidly and usually cause slow muscular con- tractions to appear after an interval of a minute or more. The order of increasing action, in saturated solution in m/2 MgCl,, is thus xylol< toluol ane creatin in all these conditions is of exogenous origin. A patient (No. 9) with spastic paralysis presented similar features. We wish to call special attention to the significance of the ap- pearance in the urine of exogenous and of endogenous creatin. We also wish to call attention to the fact that the elimination of endogenous creatin, stored up in the organism, is accomplished very slowly. In our patients of the second group, whose urine patients contained creatin, the highest ratio of a Factors Regulating the Creatinin Output in Man. 51 showed the presence of creatin, it was possible to bring about a disappearance of this substance from the urine by means of a prolonged creatin-free diet. It was necessary for that purpose to continue this diet for about two weeks. However, this is not peculiar to pathological conditions only. In normal men, whose urine showed the presence of creatin, it requires the same length of time to cause its disappearance. W. Koch had noted some time ago that if the creatin output in dogs was increased by the administration of lecithin, the high output persisted several days after the last administration of this substance. The ingestion of an unusually high beef diet caused marked increase in the creatin output and a slight rise of creatinin, thus changing the ratio creatinin creatin the beef was removed from the organism, and of this about 80 per cent in form of creatin. More complicated and instructive are the results obtained on the first group of patients: we shall begin with the analysis of the re- sults of the observations on one patient with extreme emaciation, caused by fasting (No. 10). The subject, a boy fourteen years of age, for several weeks refused nourishment, and his body weight was reduced to 44 pounds. In course of the observation it was discovered that his distress was caused by intestinal parasites, and after proper treatment he began taking nourishment. The nitrogen intake at the beginning of the experiment was 2.21 gm. per day and the output 2.15 gm. The creatinin output was normal, the creatinin coefficient being about six. In a patient with extreme muscular atrophy (No. 11), without apparent lesions in the central nervous system, there was no change in the creatinin output. In this instance the reduction in the mass of the muscle, especially in the upper part of the body, was more pronounced than on any of the other patients under our observation. The muscle showed the reaction of degeneration. The creatinin output on a mixed diet remained normal. In a second patient with progressive muscular atrophy (No. 12) the creatinin and creatin output deviated considerably from the normal, on one day the ratio creatinin creatin patient the condition was complicated by disease of the kidneys. The output of creatinin and creatin by a patient with paralysis, = 2; about 7o per cent of the creatinin contained in =~ 1-15 and on another day = '1:2:5. But in this 52 P. A. Levene and L. Kristeller. caused by a tumor of the spinal cord (No. 13), varied markedly with the variation in the character of diet. On a mixed diet his urine contained a very appreciable proportion of creatin. By means of a prolonged creatin-free diet creatin was caused to disappear from the urine; it again appeared as soon as the patient was placed on a beef diet. It is noteworthy that during the first two days of the beef-diet period the creatin content of the urine was very slight, but in five days about 76 per cent of the total in- gested creatin was removed through the urine. Of the portion re- appearing in the urine 65 per cent was in form of creatin and 35 per cent in form of creatinin. In conditions of muscular imactivity, brought about by diseases of the joints, the output of creatinin on a mixed diet varied little from the normal, and the creatin content of the urine was slight (Nos. 14, 15, 16). The output of the two substances in amyotrophic lateral sclerosis (No. 17) could not be studied in detail, for the reason that the condition of the patient was too grave to permit the obtaining of the twenty-four-hour samples of the excreta accurately. On a mixed diet the urine of this patient always contained an appreciable pro- creatinin creatin ~~ The remaining two forms, namely, poliomyelitis anterior (Nos. 18, 19) and juvenile muscular dystrophy, furnish the more valuable suggestion for the interpretation of the place of creatinin and creatin in the course of protein catabolism. In poliomyelitis an- terior the following peculiarities were observed: in both patients under observations there was noted the presence of endogenous crea- tin in the urine. In patient No. 19, on a creatin-free diet, the ratio creatinin -creatin | proportion after forty-eight hours. The coefficient of creatinin was somewhat low, but the sum of creatin and creatinin coefficients corresponds to a normal creatinin output. On the other patient (No. 18) experiments were performed with the three forms of diet: it was noted that on a low protein diet the sum of eliminated creatinin and creatin was the lowest, the ratio creatinin : ns ae R ———_— = 3:2; on a cteatin-free, but protein-rich diet ‘the out- creatin put of these two substances rose, but principally that of creatinin, creatinin the ratio of — .— increasing to 2. Finally, on a mixed diet creatin portion of creatin, the ratio of fell to 2: 3, having reached an approximately constant Factors Regulating the Creatinin Output in Man. 53 containing beef, the output of both substances rose, but principally creatinin creatin | tinin output was below normal on any one of the three diets. On a low protein diet the coefficient of the sum of creatin and creatinin was below normal. On a high protein and likewise creatin-free diet it reached a low normal value. On a high protein beef-diet the coefficient did not exceed the highest normal figure. Of the five patients with muscular dystrophy, three represented typical forms, in one the symptoms of the disease appeared in very early infancy (No. 23), another patient (No. 24) showed rather late development of a mild and stationary form of the disease. In all patients of this group there was noted an output of endogenous creatin. On a cereal diet the urine of all, with the exception of patient No. 24, showed a ratio of ean = ee creatin approximately; on a high protein and creatin-free diet the ratio gradually diminished in two patients to 1:4, remaining, however, practically unaltered in one patient (No. 23). Ona beef diet there was noted a rise in the elimination of both creatin and creatinin, but principally of creatin. The ratio of creatinin to creatin in these three cases reached 1: 4, 1: 5, and 1:6 respectively. The coefficient of creatin was below normal on either one of the three diets, and the sum of the two coefficients reaching the normal coefficient for creatinin in all experiments but one. The elimination of exoge- nous creatin was very high in all experiments except one, when only 52 per cent was removed through the body; in other ex- periments the output of exogenous creatin fluctuated between 88 and 99 per cent of the intake. The output was so distributed that about 80 per cent of it was in form of creatin and about 20 per cent in form of creatinin. of creatin, the ratio of ——hOs. he cochicient ol crea- : SUMMARY. A review of the results obtained on all of the patients brings to light the following facts: in all pathological conditions involving the muscular system, the rate of catabolism of ingested creatin is lowered and part of the ingested substance is removed in form of creatinin. In forms associated with exaggerated muscular activity 54 P. A. Levene and L. Knisteller. the catabolism of endogenous creatin generally preserves its normal. course. In forms associated with dissolution of muscular tissue and with diminution of muscular activity, there were observed con- ditions where not only the exogenous but also the endogenous crea- tin followed an abnormal course of catabolism. In those conditions the output of creatinin was low and that of creatin high. In some forms the quantity of eliminated creatin and creatinin was influ- enced by the protein content of the food. In such forms a high protein content of the food caused an increase in the output of both creatin and creatinin. Finally, there occurred forms associated with an extreme degree of dissolution of muscular tissue, which preserved a normal creatinin output. All of these observations cannot be interpreted adequately on the basis of any one of the existing views on the mechanism of creatin catabolism. The hypothesis formulated by Shaffer, which postulates that the extent of creatinin output is determined by muscular eff- ciency, does not harmonize with the observations on progressive muscular atrophy, in course of which an extreme degree of dissolu- tion of muscular tissue was not associated with any marked altera- tion of the creatinin output. The theory that regards the intensity of cellular catabolism as the principal factor influencing the crea- tinin output is also not in harmony with our observation. It is likewise little in harmony with the observations made by others on exophthalmic goitre. The theory of Mellanby is not sufficient to interpret our observations. According to Mellanby creatinin is formed in the liver and transformed into creatin in the muscle. So long as the function of the liver remains normal the creatinin formation continues to be normal. On the other hand, a defi- ciency in the function of the muscle should lead to a diminution in the rate of conversion of creatinin into creatin. Hence, in dis- eases of the muscle, one should find principally a rise in the creatinin output. In the majority of our patients the creatinin output was low. This was not caused by an insufficient formation of the sub- stance in the liver, since no disturbance of the function of this organ could be detected. Thus one receives the impression that more than one factor is concerned in regulating the creatinin output, and one is bound to accept at least two: first, the formation of the substance, very prob- ably from protein, and, second, its further oxidation. Any dis- turbance of either one of the two factors may lead to an abnormal creatinin output. The deficiency in the second function may be Factors Regulating the Creatinin Output in Man. 55 partial, so that only the ingested creatin fails to be further oxi- dized. Whether or not the two functions are performed by one organ or by several still remains to be established, but there is little doubt that the muscular system takes some part in the regulation of the creatinin output. From the results of our experiments one also receives the im- pression that the formation of creatin and creatinin represents two phases in the catabolism of but one substance, as in most observa- tions a fall in the creatinin output was associated with an increased creatin elimination, and a high protein diet (creatin-free), in some patients, caused a rise in the output of both substances. The constant value of the creatinin output in normal men is condi- tioned by the high velocity of creatin combustion in health. Thus, the creatinin of the urine normally represents only a small fraction of the creatin formed in the organism. The condition might be analogous to the uric acid output in the dog, in which the power of oxidation of purin derivatives is exceedingly high. The uric acid content of the dog’s urine is minimal, and, being so, appears to be practically constant. However, as soon as the liver is excluded from circulation and the intensity of purin oxidation is diminished, the uric acid output in the dog begins to show marked variations influ- enced by the character of the food. Ina similar manner the normal creatinin output in conditions of high muscular activity may be explained by the assumption of a higher intensity in the power of the organism to oxidize creatin, although the creatin production in these conditions probably exceeds the normal limits. In harmony with this view is the observation on one of our pa- tients with continuous tremor. In the urine of this patient only 48 per cent of the ingested creatin reappeared, while in conditions of atrophy or dystrophy practically 90 per cent reappeared in the urine. On the other hand, in the condition of muscular dystrophy, both the formation of creatin and the rate of its further combus- tion are lowered. We wish to express our gratitude to Dr. Joseph Fraenkel, physi- cian of the hospital, for his interest in the work, and to the members of the staff, Drs. S. Wachsmann and D. Felberbaum, for their con- stant assistance. 56 P, A. Levene and L. Kristeller. TABE Ri Intake. Urine. Diagnosis. Date. Mt. Diet. 1908 Kg. Cal. N. Amount. pe Ni BS ee ea Locomotor ataxia. No.1. Mr. M. G. Aug. 4 Ae Mixed eee : S505 Age 48. Locomotor ataxia. No Zan MirsG-yke= Sept. 16 Bers Mixed See Sane dee Age 51. Locomotor ataxia. 24 hrs. No. 3. Mr. E. W. Sept. 19 | 52.21 Mixed ph oe aoe 485 c.c. Age 62. incomplete Locomotor ataxia. OM te No. 4. Mrs. R.M. Sept. 16 | 43.5? Mixed ee SAE 900 cc Age 51. (ae eres ataxia. 24 hrs. Pe ENT ENT July 9 60.4 Mixed Beets : 1130 Age 60. July 10 | 60.4 Mixed Std 5 Fenn 1330 Paralysis agitans. KovG. OMe HR: July 25 68 Mixed ene 4, 06 755 Age 65. July 26 68 Mixed Sah sate 840 Sept. 7 55 Mixed "| &25- Meee 1440 Tremor. Sept. 8 age Mixed seein see's 1410 No.7. Mrs. L. Sept. 9 ses Mixed aaae Fade 650 Age 2. Sept. 15- Milk Oct. 3 55 and egg 2974 14.84 1321 Avs. of June 5 Milk es aie Se eae eee 3470 18.31 1305 No. 8. Mr. R.\ June 1 Age a AEG 545 Beef 2829 18.22 | 1390 June 10 Cereal aaa 1d 54° diet 3270 9.06 | 1085 May12| 50.5 | Mixed m | 540 Spastic paralysis. May 14 aee Mixed cece 660 No. 9. Mr. W. H. Avs. of Age 35. May 26 Milk and 27 ae and egg 2342 14.70 455 Milk } June 30 | 21.07 | and egg | 426 224-2 745 | July 1 aa 2 573 S205'5) 835 Starvation on account | | July 2 ae aeACs 822 4.81 1235 of intestinal worms. July 3 seh ae gs 925 4.59 1780 No.10. Mr.M.G.} | July 4] ... ee 800 5.41 | 790 Age 12. July 26 acts Mixed Scie Se 690 July 27 aot Mixed See eee 1120 July 28 Gc Mixed sare eS 710 July 29 Bos Mixed oe cde 1440 1 Evidently incomplete. 2 Incomplete. % Two preliminary days on the same diet. * Of creatinin intake compared with the milk and egg period. Factors Regulating the Creatinin Output in Man. 57 SWAB IORY oh. Urine. Output. Coefficient. Ratio of = = creatinin Specific | Potal- Creatinin} Creati- |. Creatin | Creati- Crone to. Be hs NT and nin N. N. nin and reat | Creatin | creatin. gravity. creatin N. ane pin. gm. mg. mng. mg. 10) Yoo eae oe erate absent aes es a%eis soa sone eee eee trace sO 4 etete tera 1.017 25 82 76 6 1.6 1.5 0.1 TOUS SS |rae. 292 274 18 6.7 6.3 0.4 1.015 _ | 10.46 386 386 absent 6.4 6.4 5 1.014 | 11.16 416 416 trace 6.9 6.9 410 410 trace 6.0 6.0 354 354 trace 522, 5.2 F 384 321 63 7.0 5.8 162, 19:70:20 375 310 65 6.8 5.6 1.2 1032.0 5 347 295 52 6.3 5.4 0.9 Ose, : 287 265.5 21.5 72 4.8 0.4 TOW am iels:53 375 340 35 7.0 6.3 0.7 LOO 1.017 | 16.77 555 370 185 10.3 6.9 3.4 970250 + 48% 1.016 7.38 is 330 ae arate 6.1 “ 1.028 dees 335 275 60 6.7 5.6 it 1025 1.025 349 297 52 6.9 5.9 1.0 1 : 0.20 1.033 | 10.90 303 272 31 6.0 5.4 0.6 iL -g abt 2.15 129 123 6 2be 5.7 : 3555 123 123 wae aE y OS f 4.02 126 126 Bere ae 5.8 Ps 3.74 121 121 ee Bull 3.88 111 111 =p 5.0 ae 3.55 131 117 14 5.9 5eS 0.6 3.84 131 131 trace 5.9 5.9 aise 4.23 135 125 10 6.2 by// 0.5 4.39 119 119 trace 5.4 5.4 Bee 6 Daily 420 gm. beef (378 mg. creatin N.). Two preliminary days on the same diet. ® Three preliminary days on the same diet. 7 Table shows only the first five and the last four days of thirty days’ observation. 58 P. A. Levene and L. Kristeller. Intake. Urine. TABLE I (continued). Diagnosis. Diet. Cal. N. Amount. gm. 2 | | | —___ Atrophy. \}| Julyl | 403 Mixed S aos ee 735 peote sie da S| July6 | 40.258 tee 2346 | 11.3 730 Progressive muscular : nee atrophy. | | May 6 Biexe : Baers sires No. 12. Mr.M.M. [| May Bet| oe cot yale ek ener ote 840 Age 65. Sept. 10 Aes Mixed aoe ee eee Sept. 15 ale Mixed eee piles 1425 Sept. 16 46.3 Milk mee aes 850 Sept. 17 ie and egg ee ake 1180 Pp Sept. 18 ae i ra ee Bas 630 Paralysis due to tumor. ore x6 ee CG a Tihs aa No. 13. Mr. A. Ba. Sent 21 ae “6c ale mes 810 Age 02 \ | cén22'| oo, | Seo" Ci Syl ae Sept. 23 sisie Beef eae Sere 765 Sept. 24 oe Beef ets see 1445 Av. of Sept 25.1 46.54 | Beef | -c.cu/| .pesen aes Arthritis deformans. o No. 14. Mrs. D. O. Sept. 16 | 45.9 Mixed aie ats sent 1340 Age 32. Arthritis deformans. No: 15. Mrs. J. J. Sept. 22 | 42.6 Mixed weve apse 1065 Age 60. Chronic gout. ff No. 16. Mr. N. M.\ Sept. 18 | 53.5 Mixed Lae a5 1450 Age 51. Amyotrophic lateral Ines sclerosis. A Me 15 Weed 450 No. 17. Mr. J. L. aan ae na aa == 92) _ Age 68. o Avs. of May 14 Anterior poliomyelitis. to 16 pee) ec ou 5-36 = No.18. Mr.M.Sch. May 19 Milk Age 31. to 21 se sa; 2162 9.43 823 May 25 ee to 26 ee Beef 2273 10.60 995 | Avy. of Anterior poliomyelitis. Oct. 2 Milk No. 19. A.B. to 1l 304 and egg 2315 the wie Age ie) Cea cr)| 2047) ect || 2310.4) 12s 1440 * After four preliminary days on the same diet. ° Weight of patient could not be obtained. Table shows the slow disappearance of creatin from the urine on a milk and egg diet. 1 Daily 5 gm. beef N. and from Sept. 25-29, 170 c.c. soup. ‘? Weight of patient could not be obtained. Twenty-four-hour quantities not always complete. Factors Regulating the Creatinin Output in Man. 59 TABLE I (continued). Urine. Output. Coefficient. eee : Creatinin 1 : Creati- Z aoe Specific | Total and Creati- | Creatin | Ji. and Creati- Ceeatin ; gravity.| N. lcreatinN.| 3in N. N. aes nin. a ee ws gm. mg. mg. ms. i OUG). |i 2 3 256 256 trace 6.3 6.3 Bee 1.016 7.24 250 250 trace 6.2 6.2 rae Boss aoe 363 145 218 sto seis sexe IED sees Be 426 121 305 AAS Soc S50 Pe2:7 wisn sists 346 297 49 7.6 6.5 alate sec | LOAL 329 306 23 7.2 6.7 0.5 ee 8.62 261 250 ll 5.6 54 0.2 Bee | Lt36 355 323 32 vies 7.0 0.7 See | LONSZ 224 198 26 4.5 3.9 0.6 ee 9.63 258 246 12 5.6 5.3 0.3 10.81 299 299 Lae 6.5 6.5 aie. 8.74 200 200 4.3 4.3 Si 9.73 257 257 Lists 5.8 5.8 : 8.16 264 256 8 a one a 9.05 307 299 8 ais ee er 7.98 360 291 69 7.8 6.3 15 e POTS: hss 348) |) 331 17 7.6 7.2 0.4 cae 1) Vd ee 239 239 trace 5.6 5.6 ae PONS ass 444 423 21 8.3 7.9 0.4 ie sees 313 140 163 ferarn 50 aie 131.16 1017 4.11 170 100 70 3.8 2.2 1.6 arts 1018 6.82 240 160 80 5.5 Sell L828 n i cae 1015 8.29 480 210 270 10.7 4.8 Le irae Mpeee LOI, |)... 2 209.5 146 63.5 ~ 6.9 4.8 2.1 1 : 0.43 399 LIS |, 224 13.2 5.9 13.0 /\qhy eae os 84 % 15 13 Two preliminary days on the same diet. 14 Three preliminary days on the same diet. Patient received daily 240 gm. beef (192 mg. creatin N.) and one day 240 c.c. soup. 18 Of the creatin intake compared with the milk diet. 16 Daily 250 gm. beef (225 mg. creatin N.). 60 P. A. Levene and L. Knisteller. TABLE I (continued). Intake. | Urine. Diagnosis. Date. Wt. Diet. 1908 Cal. N. | Amount. ec | ae Muscular dystrophy. Ay. of Milk No. 20. Miss T. Y.} | April 228) 64.97 | og. 1891 1245 555 Age 21. to 23 ~ 88 Avs. of Sept. 4 Milk to 10 33 and egg | 2316 10.61 682 April 21 Milk to 23 3338 and egg | 2122 14.10 720 Muscular dystrophy. May 1 No.2l. “St ¥. to 3 32.219 | Cereal 2431 5.19 833 Age 18.* May 13 to 15 33:6” Cereal 2276 5.14 820 Mey ©) 32:57) Beck) 2851.0 len 718 May 25 | 342% | Beef | 2453 | 13.79 705 Avs. of | Septiaas Milk Feb. 24 Milk fe aeee| SUSE sell aosTanecons 790 | Mar. 5 Milk | rent ee leat ae 2074 12.80 1285 Muscular dystrophy. | | Mar. 15 7| Milk No. 22. I. Sch. to 17 ===" | endless | 72/0 || bi28 eal Age 17.| | April 14 Milk i 4016 Sek ences 2451 17.58 1227 April 30 | 2 to May 2 | 40.8 °8 | Cereal 2603 5.72 925 Mar. 24 Fa 36 4 Beef 2208 21°23 1090 Mar. 29 | to30 | -= Beef eae 21.00 ee Avs. of June 7 Milk | to8 16 nedicee 2046 10.35 720 Dystrophy. | June 7 | 16% | Cereal | 1933 | 4.52 815 No. 23, Mri ian Aide 21 Beh) | des |) Cereal | 2074 4.53 500 June? | 15.3% | Beef | 2022 | 10.62 755 Two preliminary days on the same diet. The preceding day on the same diet. | ™ Four preliminary days on the same diet. Two preliminary days on the same diet. Of the creatin intake compared with milk period Sept. 4 to 10. Only a few of the experiments performed on this patient are here recorded. Two preliminary days on same diet. Daily 14.8 gm. beef N. (396 mg. creatin N.) Of the creatin intake compared with milk period April 21 to 23. Three preliminary days on the same diet. Daily 11.4 gm. beef N. (310 mg. creatin N.) two days each 240 c.c. soup besides. Factors Regulating the Creatinin Output in Man. 61 Urine. Specific gravity. 1.028 1016 1012 1015 1013 Total | Creatinin and | } TABLE I (continued). Output. Creati- Ne |creatinN.| ™n N. gm. gm. om- 8.99 | 11.89 4.30. 4.75 | 12.11 | 12.86 | 10.04 13)-3y/ 12.91 14.53 6.22 20.09 Aus At 6.80 Sta3 3.81 6.10 | = fon) ay cose is] w 290 50 56 Creatin mg. 125 181 121 141 152 319 520 208 230 270 152 170 92 560 780 78 74 78 234 Coefficient. | Creatinin Greate and crea+ = Creatin. : nin. tin. 2.6 0.7 1.9 Hal 1.6 5a5 5.5 1.8 S\7/ 6.0 1.6 4.4 6.5 1.9 4.6 13.5 3.9 9.6 19.4 3.9 1585 Oo eZ, 4.5 ee. TES 5.6 8.4 1.7/ 6.7 52S 1.6 ehy/ 6.0 17 4.3 3.4 Mal ZS 16.1 2.6 1855 21.9 2.9 19.0 7.6 2.8 4.8 7/3 2.6 4.5 7.6 2.8 4.8 18.9 3.6 15.3 2° Two preliminary days on the same diet. 27 29 Three preliminary days on the same diet. *8 ‘Two preliminary days under cereal diet with 9 gm. N. intake daily. Of creatin intake compared with milk period Sept. 5 to 11. ® Daily 14.6 gm. N. (414 mg. creatin N.). 31 Two preliminary days on the same diet. 82 Three preliminary days on the same diet. consecutive days. Figures omitted here, since they were of the same character as those recorded. Ratio of creatinin to creatin. si eA Ws Pas i oyigt iL BAS 13.9) 1W:23:8 Y 33:8 ls 3:8 1: 2.2 LSPS e221 Be .2 1 36:63" Ge Gy 3 gs) os : 4.2 H- = S|» 26 Seven preliminary days on same diet. The patient was on a milk diet for ten 83 Three preliminary days on the same diet. 62 P. A. Levene and L. Knisteller. TABLE I (continued). Intake. Diagnosis. Date. Wt. Diet. 1908. Cal. N. Amount. (eg a2) PES FE ee | Re | aT Avs. of ‘ Sept. 14 Milk fee || (O2/0) | ae oe 2543 | 13.10 ae May 18 Milk 5 onl anes 2491 | 14.67 587 April 30 Milk rome Ps) Mantras 2110 | 12:15 613 Muscular dystrophy. June ee ies 2875 15.64 860 No. 24. Mr. E.S. Ma 12 ane SBS Age 42. ] | “014 a Cereal | 2326 5.44 660 june) gece | Cereal!) 2946) | aa 988 Move areas Beef | 3211, | 21.25 667 Maye 35 eee Bech i) 262001 | 1504 560 to 27 34 In these nineteen days no marked variation from day to day was observed. % Two preliminary days on the same diet. 3° Three preliminary days on the same diet. * ‘The preceding day on the same diet. Factors Regulating the Creatinin Output in Man. 63 TABLE I (continued). Urine. Output. | Creatinin | : Specific | Total and | Creati- gravity.| N. | creatinN.| nin N. gm. mg. | mg. Bees 248 183 1.025 9.65 229 159 1.026 9.73 223 156 1.023 | 13.60 336 196 1.023 4.90 248 207 1.010 5.28 223 182 Wa 1.028 | 14.07 100%, 263 Boa) alan), 122 230 . . 70 % 39 Coefficient. . Ratio of creatinin P Creati- . to oes nin and ae Creatin. | creatin. s creatin. P mg. 65 3.9 2.9 1.0 1: 0.36 70 3.6 2.5 ial 1:0:62% 67 3.6 Des iat 1 :0.62** 140 5:3 Sal 2.2 MEO PAL 41 4.0 333 0.7 LEO 2a 41 3.7 3.0 0.7 151022258 460 11.7 4.3 TA. Weer 263 7.9 301 4.2 iealat % Three preliminary days on the same diet. *® Of creatin intake as compared with milk and egg period Sept. 14 to Oct. 3. * Two preliminary days on the same diet. Daily 17 gm. beef N. (477 mg. creatin N.). 41 Four preliminary days on the same diet. Daily 12.5 beef N. (350 mg. creatin N.). 64 P. A. Levene and L. Knisteller. TABLE II. INFLUENCE OF PROTEIN INTAKE ON THE ELIMINATION OF CREATIN AND CREATININ IN SOME OF THE PATIENTS. Name. M. Sch. I. Sch. Diet. N. Intake. Cereal 5.6 Milk 9.4 Cereal Sey, Milk 13.0 .. 9.4 . 12.8 i 14.3 s 17.6 Cereal 4.5 be 4.5 Milk 8.4 N. Output. Coefficient. Total. Creatinin.| Creatin. 3.8 2.2 1.6 5:5) 3 1.8 3.4 dd 2.3 Dell 1.2 4.5 7.2 1:5 5.6 8.4 uy / 6.7 BES) 1.6 34 6.0 Ld 4.3 7.6 2.8 4.8 dol 2.6 4.5 8.4 3.0 5.4 Factors Regulating the Creatinin Output in Man. 65 BIBLIOGRAPHY. 1 Formn: This journal, 1905, xiii, p. 85; Fourn, Festschrift fiir Olaf Hammer- stein, Reprint, Upsala, 1906. * KLERCKER, Biochemisches Zeitschrift, 1907, iii, p. 45 (see also reference 12). 3 SHAFFER: This journal, 1908, xxiii, p. 11. * Weser: Archiv fiir experimentelle Pathologie und Pharmakologie, 1908, lviii, P. 93. 5 MELLANBY: Journal of physiology, 1907-1908, xxxvi, p. 288. 8 Yon HoaAGENHUYZE and VERPLOEGH: Zeitschrift fiir physiologische Chemie, 1908, xlvi, p. 415, and 1908, lvii, p. 161. 7 Benepict and Meyers: This journal, 1907, xviii, p. 377. 8 LEATHES: Journal of physiology, 1906-1907, Xxxv, p. 205. ® FroscHBACH: Archiv fiir experimentelle Pathologie und Pharmakologie, 1908, lviii, pp. 112-140. 10 CLosson: This journal, 1906, xvi, p. 252. 11 SHAFFER: Jbid., 1908, xxii, p. 445. 12 HOAGENHUYZE and VERPLOEGH: Zeitschrift fiir physiologische Chemie, 1905, xlvi, p. 415. 43 BeneEpDicT: Carnegie Institution of Washington, 1907, Publication No. 77. 14 BENEDICT and DiEFENDORF: This journal, 1907, xviii, p. 362. 15 SpricGs: Quarterly journal of medicine, 1907-1908, i, p. 63. 18 UNDERHILL and KLEINER: Journal of biological chemistry, 1908, iv, p. 165. 17 RicHARDs and WALLACE: Jbid., 1908, lvii, p. 179. 18 LeFFMAN: Zeitschrift fiir physiologische Chemie, 1908, lvii, p. 476. 19 GoTTLiesB and STANGASSINGER: Ibid., 1900, lii, p.1; 1908, lv, p. 295 and p. 322. 20 HOAGENHUYZE and VERPLOEGH: Ibid., 1908, lvii, p. 265. ACAPNIA AND SHOCK.! —III. SHOCK AFTER LAPA-— ROTOMY: ITS PREVENTION, PRODUCTION, AND RELIEE: By YANDELL HENDERSON (Witu THe CoLLaporaTion OF JAMES RYLE COFFEY anp FRANK ELMER JOHNSON). [From the Physiological Laboratory of the Yale Medical School.] CONTENTS. I. .Acapnia-as.an element in inhibition. -°.5 . 2.22. 2 2 202 = 2 poe 66 If. Relations, of the blood gases to peristalsis 2 ..6 5 5 <2 = 3). /o>-u sue 69 Till: "xhalation of CO; from exposed: viscera... Se 4 2) See 76 IV; @he'effects of aeration of the viscera) .: .: <2- so. =) 0+ =) 5 CM V. TDhe* production and ‘relief of acapnial’shock 5 2 2 3.5 .) 6 <0. = eee 80 VI. Conclusions I. ACAPNIA AS AN ELEMENT IN INHIBITION. HEN the abdomen is opened, peristalsis ceases. After re- _ closure the motility of the stomach and intestines does not return for a considerable period. Exposure and handling of the viscera induce also a loss of tonus,—as evidenced by the tym- panites after surgical operations upon human beings. Ether an- eesthesia even for operations other than laparotomy is frequently followed by flatulence. The investigations of Bayliss and Starling,? Eliot and Barclay- ‘ For the two preceding papers of this series, see This journal, 1908, xxi, p. 126; and 1909, xxiii, p. 345; see also the latter volume, p. xxx, for an abstract of a future paper of this series. * Cf. von Braam Hovuckceest: Archiv fiir die gesammte Physiologie, 1872, vi, p. 266. * Baytiss and StaRLING: Journal of physiology, 1901, xxvi, pp. 107 and 125. 66 Acapnia and Shock. 67 Smith,* Cannon and Murphy,°® Meltzer and Auer,* Magnus,’ and many others have made it probable that the cessation of peristalsis and loss of tonus in exposed viscera are not due to a single simple process. According to Meltzer,* “Opening of the abdomen is instrumental in bringing out a two-fold inhibition —a reflex in- hibition and an inhibition of a local peripheral mechanism.” Per- haps each of these processes will in the ultimate analysis need to be resolved further into several distinct elements. It is not, how- ever, the object of this paper to attempt a solution of this general problem, but to contribute a new point of view and some experi- mental evidence on certain phases of the topic to which relatively little attention has been paid. Briefly stated, the points to be con- sidered are: I. Stimulation of afferent nerves causes hyperpnoea. From the excessive pulmonary ventilation reflexly induced by laying open the abdomen, — or indeed by any surgical operation, — there re- sults a diminution in the CO, content of the arterial blood. Fur- thermore, ether is a respiratory stimulant. Even in operations under morphin and ether in dogs, and probably also in men,® respiration is constantly excessive, and the arterial CO, at nearly all times subnormal. This acapnia is a factor in the central, or reflex, inhibition of which the splanchnics are the efferent path. 2. When the viscera are exposed to the air, exhalation of CO, occurs, and the focal acapnia which results is a factor in the pe- ripheral inhibition. 3. General acapnia from hyperpncea and local acapnia from exposure are the initial causes of surgical shock after laparotomy. * Error and Barcray-SmitH: Journal of physiology, 1904, xxxi, p. 297. ° Cannon and Murpuy: Journal of the American Medical Association, 1907, xlix, p. 840. ® MeELTzeER and Aver: Zentralblatt fiir Physiologie, 1907, xxi, no. 3, and This journal, 1907-1908, xx, p. 259. Also AUER, This journal, 1907, xviii, p. 347; and 1909, XXlii, p. Xvii. 7 Macnus: Archiv fiir die gesammte Physiologie, 1908, cxxii, p. 210 (refs. to previous papers). ; * Met1zER: Archives of internal medicine, July, 1908, (review and discussion of literature). ® In man under ether the heart rate is nearly always more rapid and often far more rapid than in normal life. It was shown in the first paper of this series (Loc. cit.) that under such conditions the heart rate is an index which varies inversely as the arterial CO;. 68 Yandell Henderson. The starting-point of the reasoning and experimentation which have led to these conclusions was the well-known fact that the exposed and quiescent viscera of a rabbit are thrown into violent peristalsis by occlusion of the trachea. It was formerly supposed that this phenomenon and the other functional stimulations of as- phyxia were caused by lack of oxygen. The course of. recent investigations has tended to show, however, that deficiency of this element exerts a purely paralyzing influence upon respiration,'®? — at least in respect to the direct effect in mammals. The excitant ‘influences of asphyxia upon respiration and arterial pressure are really due to excess of CO,. It is probable that in hke manner the excessive peristalsis of asphyxia is caused by hypercapnia, not by anoxhzemia.. Conversely, it is a matter of common knowledge that respiration is increased by mental excitement, by pain, and by ether anzesthesia in some stages. Doubtless hyperpncea from any of these causes involves excessive pulmonary ventilation. It is not necessary that the breathing should be violent in order to be “ excessive,’ im the sense in which the word is here used. A slight increase in rate or depth brings about a diminution of the CO, content in the arterial blood, and ultimately of that i the tissues of the body. It is certain from Haldane and Priestley’s' investigations that even a slight decrease is something against which the body is nor- mally protected by nature with the utmost care. Even a small degree of acapnia induces marked alterations in the functional activity of many nerve centres ard of some peripheral mechanisms. Accordingly the failure of motility in the stomach and intestines during and subsequent to hyperpncea from anger or sorrow, from ether excitement, amd from irritation of afferent nerves, may be due in part to a greater or fess degree of arterial or of general acapmia. When a tissue is exposed to the air without the protection of the skin, marked alterations rapidly develop. As described by Crile’? the “exposure particularly affects the vaso-motor mech- ” Cf. HALDANE and Povutton: Journal of physiology, 1908, xxxvii, p. 390. For a review of the literature, see SCHENCK: Ergebnisse der Physiologie, 1908, vii, p. 65; also Henperson, Y.: This journal, 1908, xxi, p. 130. ‘\ HALDANE and PrrestLey: Journal of physiology, rgo5, xxxii, p. 225. ? Crite: Surgical shock, 1899, pp. 130, 135, 136, 147. Similar observations were described by von BrAAmM HouckcGeEEst: Loc. cit. Acapnia and Shock. 69 anism. If a bloodless field of operation, the thigh for example, be exposed, it soon becomes suffused with blood, all the vessels be- come dilated, the translucency of the tissue is lost, and further dis- section then becomes bloody.” The abdominal viscera exhibit these reactions in an especially high degree; after exposure ‘‘ even the clear transparent peritoneal spaces in the mesentery display vessels and sometimes become red.’ The intestine also undergoes a rapid loss of tonus, either as a result of the congestion or directly from the same causes as those which affect the blood vessels. Handling the intestine accelerates the development of these con- ditions. But even when mechanical irritation is avoided and the loop under observation is kept warm and moist, extreme loss of tonus and intense congestion may occur. The amount of CO, in solution in the tissues and fluids of the body is such that it must exhale readily from an .exposed surface, and a local acapnia in the immediately underlying tissues must result. It is sometimes a practice of surgeons to “ protect’ exposed viscera with cloths moistened in warm saline and frequently changed. The condi- tions thus maintained are as unfortunate as if they had been specially designed for the production of local acapnia. Our ex- periments show that, if the atonicity and congestion in an exposed loop of intestine has not gone too far, restoration of the CO, content of the tissues induces recovery. of both myenteric ‘* and vascular tonus. They show also that a degree of paralysis can be induced by warm moist aeration so complete that introduction of CO, gas into the lumen of the gut or asphyxiation of the sub- ject fails to elicit any reaction. II. RELATIONS OF THE BLOOD GASES TO PERISTALSIS. The significance of the following series of experiments arises from the fact that no one previously (so far as we can learn) has ever seen directly, in the opened abdomen of an animal with spinal cord intact, the stomach, the small and the large intestine all per- forming their normal movements. It is probable that the condi- tion which inhibited peristalsis in the investigations of previous observers was acapnia. We have data which show that a diminu- tion in the CO, content of the blood usually occurs in animals 13 Cf. observations of BoKAl, quoted on p. 80 of this paper. 70 Yandell Henderson. under operative conditions. The three experiments given below demonstrate that when this acapnia is prevented the forms of motility of the gastro-intestinal canal are practically identical with those shown by radiographs of unoperated animals. The objection cannot be made that the results of these experiments were due to asphyxia (anoxhzmia or hypercapnia), for the blood-gas analyses show that the oxygen supply was ample (note the analyses of the venous blood), and that the arterial CO, in two of the experi- ments was not considerably above that of normal life. - The experiments were performed upon three dogs, — each of about 10 kilos body weight. Each received 0.05 gm. morphin sul- phate, and was then anesthetized with chloroform with as little hyperpneca as possible. The abdomen was laid open by an incision through the entire length of the linea alba. The omentum was cut out, and the intestines moved sufficiently to bring into the visible field the greater curvature of the stomach, the lower part of the ileum, and the ascending and transverse colon. A sheet of thin, flexible, transparent celluloid (of about the width of the animal's body, 2 cm. longer than the incision in the linea alba and with rounded corners) was inserted inside the body wall. A practically air-tight window was thus formed through which the viscera were clearly exhibited. The air in the space below the celluloid was expelled by means of a stream of CO, gas. In one case pendular movements in the small intestine manifested themselves immedi- ately thereafter. No special precautions were taken against cool- ing, except that the experiments were performed during July (1907). The trachea was opened and a tube 15 mm. in diameter and 2 metres long attached. With this increase in the dead space of the respiratory tract, the breathing became deep and full, but the mesenteric arteries retained their bright red color. Thereafter, owing to the narcotic effect of the morphin which the animals had received, no further administration of anzesthetic was needed. The stomach was then distended with air ™ by way of the cesoph- agus. Bread mush slightly acidified with HCl was introduced far up in the large intestine. A part of the mush passed on into the ileum. Within ten minutes after the establishment of these conditions, in all three experiments, peristalsis was in full activity in the '* Air is not a chemical stimulus. Cf. Boar: Loc. cit. (p. 80 of this paper). Acapmia and Shock. 71 stomach, lower ileum, and upper colon. The movements were quite different from those observable when an animal is asphyxi- ated; and the characters of the movements in the stomach, ileum, and colon were in marked contrast each to the other. In the stomach the greater curvature of the antrum pylori, the pylorus itself, and a part of the fundus were visible. At the preantral groove con- strictions developed, —in one subject every fourteen seconds and in another every eighteen seconds. These constrictions, when viewed in profile as they moved toward the pylorus, were 15 mm. deep and 25 mm. long. The waves moved at a rate such that one could be seen developing in the preantral groove as the preceding reached the pylorus. The pylorus remained contracted 1° and the duodenum quiescent throughout the periods of observation in all three experiments. The picture presented was identical with the radiographs obtained by Cannon from cats in that the diameter of the antrum was reduced at the trough of each wave to about half that between the waves.’® It differed merely in the fact that only two waves, instead of three, were simultaneously visible, and that their rate was slower than in the cat. Roux and Balthazard ‘7 have found that this is normally the case in dogs, in which animals they saw, as we have, about four waves to the minute. The pictures presented by both ileum and colon were identical with those seen when the radiographs of Cannon '® are revolved in a zoetrope. In the ileum the movements were those of rhythmic segmentation. Strikingly quick and vigorous were the alternate constrictions and relaxations and the corresponding blanching and reddening of the segments, as if a column of large frog hearts were beating in such mutual co-ordination that numbers I, 3, 5, and 7 were in systole as 2, 4, 6, and 8 were in diastole, and dice versa. The contractions certainly involved complete occlusion of the lumen of the gut. In the distended colon the entire visible portion (about Io or 12 centimetres) exhibited a series of alternate rings of contraction and relaxation. The former were 8 mm. apart from trough to trough, and constricted the gut to about half the diameter of the intervening portion. They moved steadily from left to right (17. ¢., 15 Cf. Macnus, R.: Archiv fiir die gesammte Physiologie, 1908, cxxii, p. 210. 16 CanNON: This journal, 1898, i, p. 364. 17 Roux and BALTHAZARD: Archives de physiologie, 1898, xxx, p. 18. 18 Cannon: This journal, 1901, vi, p. 26s. 72 Yandell Henderson. anti-peristalsis) at a velocity of 3 centimetres per minute. In one — case the large intestine was distended merely with air for fifteen minutes before the bread mush was introduced. The anti-peristalsis which was maintained during this period was not perceptibly dif- ferent from that seen after the introduction of the bread mush. All of these movements were maintained for periods of half an hour. Then the long tube attached to the trachea was replaced by a short cannula. With this diminution in the dead space of the respiratory tract respiration became shallow, the heart rate quick- ened, and blood pressure fell slightly. Within three minutes the intestines, both small and large, became quiescent. The sheet of celluloid was then removed from the abdomen. The intestines were as fresh and pink in appearance as when the abdomen was first opened. Immediately after removal of the celluloid, however, the vascular congestion always seen when the viscera are exposed to the air began to develop. Within fifteen minutes gastric peri- stalsis had ceased and the small intestine showed a notable loss of tonus. Before the long tube and celluloid were removed two samples of blood were withdrawn, — one from the femoral artery, the other from a cannula inserted through the jugular into the right heart, — and the oxygen and CO, which they contained were determined by the Barcroft-Haldane method.'? After the removal of the long tube and celluloid the abdominal viscera were exposed for one hour to a current of air warmed to 35° to 38° and satu- rated with moisture. A continual handling of the intestines was involved in this process, but this was done gently. Hyperpnoea and tachycardia accompanied the manipulation, although ether was administered in quantities sufficient to maintain complete anzesthe- sia.. Whenever the manipulation was stopped for a few moments, respiration became shallow or even ceased altogether for a brief period. The heart rate, on the contrary, continued rapid. Arterial pressure was not diminished, although the pulse was very narrow. The stomach and intestines became greatly congested, widely re- laxed and atonic, and wholly irresponsive to stimulation. At the end of the hour two more samples of blood for analysis were taken from each animal. The percentage contents of oxygen and © Barcrorr and HALDANE: Journal of physiology, 1902, xxviii, p. 234. The flasks used by us were three times as large as those of BARcRorT and HALDANE, and the blood samples were 3.0 c.c. instead of only 1.0 c.c. ee Acapnia and Shock. 73 ’ CO, found in these and the earlier samples (calculated to 0° and 760 mm. of mercury pressure) are given in the table. The significance of the figures for the CO, content of the blood during peristalsis is emphasized by the fact that (as we have found in a series of investigations to be published in detail in a later TABLE I. SHOWING PERCENTAGE OF Og AND COg IN THE BLOOD DURING PERISTALTIS (SAMPLE 1) AND AFTER EXPOSURE AND HANDLING OF. VISCERA (SAMPLE 2). ARTERIAL BLOoop. Oxygen. Carbon-Dioxid. Animal. Sample 1. Sample 2. Sample 1. Sample 2. 18.8 20.9 50.4 24.5 22.6 21.0 43.3 29.9 14.5 23.7 43.3 26.3 VENOUS BLOOD. 10.6 8.8 16.6 paper) dogs which are etherized without morphin almost always develop, during the process of anesthetization, a hyperpnoea which reduces the CO, content of the arterial blood below 35.0 per cent. Those which have received a moderate dose of morphin and are then etherized are less liable to hyperpncea; but even these seldom, in our experience, reach the operating table with a CO, content in the arterial blood as high as 38.0 per cent. In profound anzs- thesia under morphin and chloroform we find that the CO, con- tent is 45.0 to 50.0 per cent or even higher. But when ether alone is used, in addition to the acapnia induced during the stage of ex- citement, every incision through the skin, or other operation involv- ing stimulation of afferent nerves even in complete anzesthesia, is 74 Yandell Henderson. accompanied by an augmentation of respiration and a correspond- ing diminution in the CO, content of the arterial blood. Further- more dogs under ether show very little tendency, even when left perfectly quiet, to recuperate the normal percentage. On the con- trary, it is not uncommon to find dogs in which hyperpnoea once started continues for long periods in spite of a maximum admin- istration of ether by a mask. In man in prolonged operations under ether the heart rate usually shows a progressive increase, indicating probably, as we have pointed out in a previous paper, a correspond- ing diminution in the CO, of the arteriai blood. It appears to us to be highly probable that the occurrence of acapnia affords the reason why such expert experimenters as Bayliss and Starling,?° in their well-known investigations upon the motility of the ali- mentary canal (under A. C. E. anesthesia), did not see the rhyth- mic segmentation in the small intestine and the anti-peristalsis of the colon which are certainly their normal movements.” Likewise Eliot and Barclay-Smith ?? found it necessary to destroy the spinal cord in all animals except the cat in order to observe any marked motility in the colon. Furthermore, what they saw (under ether anzesthesia) was so abnormal that they were led to doubt whether true anti-peristalsis occurs in this portion of the gut. Doubtless their animals were all in conditions of more or less acapnia. It is because of some mysterious inhibiting element in the condition of animals under operative conditions that so many of the func- tions of the abdominal viscera, etc., were not made clear until methods of observation under normal conditions were employed, especially by Pawlow and by Cannon. But the real nature of the abnormal and inhibiting condition which prevents or distorts these functions under anzesthesia and immediately after operation is un- known. We believe that fundamentally it is acapnia. From all the blood-gas analyses in the literature no one inferred that the CO, content of the arterial blood in health is a constant, yet the investigations of Haldane and Priestley have demonstrated that this must be the case. On the contrary, the percentages found in the blood varied widely, — more commonly below than above 40 per cent. Therefore these data prove that the majority of the *» BAyLiss and STARLING: Journal of physiology, 1901, xxvi, pp. 107 and 125. Cf. CANNoN: This journal, 1902-1903, viii, p. xxi. Evior and Barciay-SmitH: Journal of physiology, 1904, xxxi, p. 297. 21 we 2 “ Acapnia and Shock. ie 5s animals from which the blood samples were taken — and corre- spondingly a majority of all animals and men under anesthesia and operative conditions — were in states of more or less acapnia. Great significance attaches to the facts that the movements ob- served in our experiments were the normal forms of motility, and that the blood gases were nearly the same as in normal life, — not asphyxial. Therefore it is not reasonable to suppose that the in- activity of the gut usual under anesthesia and operative condi- tions is due to some inhibition other than acapnia, and that in our experiments the increase of CO, by the tracheal tube neutral- ized this inhibition by a stimulation. And if there is here no neu- tralization, there is then no longer need to invoke “ prolonged in- hibition’ (in the wide but vague sense of some writers) to explain the phenomena of shock in the abdominal viscera. Whether or not this is true of the suppressions of secretory activity must be left for discussion in later papers. The data here presented indicate that the failures of motility at least are the expression of lowered tonus in the tissues of the body and altered activity in the centres of the central nervous system because of diminished COg. These observations and the blood-gas analyses indicate also that the effects of asphyxia upon the intestine —the blanching, con- striction, and violent pendulum movements —are mainly due to hypercapnia. Lack of oxygen without excess of CO., in the ex- periments of Magnus ?* upon the excised intestine, caused an in- creased tonus in the circular muscle. But the effect was slow in comparison with that in ordinary asphyxia, and the possibility that there was an increase of CO, in the saline bath does not seem to have been wholly excluded. We observed repeatedly that, if the dead space of the respiratory tract was increased by so long a tube that the blood in the mesenteric arteries lost its brightness of color, the movements of the intestines were for a few seconds accentuated _and then ceased altogether. If the tube was then shortened until the blood recovered its brightness, the intestines again exhibited a brief period of exaggerated activity. These observations support the idea that lack of oxygen is not a stimulus, but that it merely paralyzes. 3 Macnus: Archiv fiir die gesammte Physiologie, 1904, cii, p. 137. 7G; Yandell Henderson. III. ExHALATION OF CO, FROM EXPOSED VISCERA. The following experiments afford a basis for estimating the rate at which CO, exhales from exposed viscera. They show that the CO, content of uncovered tissue is reduced, for in successive periods of a half hour each the exhalation diminished. Crile ** states that “where the omentum is made to cover the viscera, there is much less shock.’’ These experiments show that this protection can be in part explained by a less rapid exhalation of CO, from viscera covered by it than from organs directly exposed. Cats were placed under chloroform anesthesia. ‘The abdomen was opened by an incision in the median line. The large end of a glass funnel 8 to 10 centimetres in diameter was inserted through the cut so that the body wall and skin fitted air tight around it, like a button-hole around a button. The neck of the funnel was connected with a Pettenkofer absorption tube (2 metres in length and containing 400 c.c. of dilute baryta water) to which in turn was con- nected a suction pump on a water faucet. Air was admitted to the space above the viscera and below the inverted funnel by a small glass tube passed through the body wall. To the outer end of this tube was attached a vessel containing moist soda-lime and a wet sponge. There was no negative pressure under the funnel. Thus every minute 200 c.c. of moist air, at 18° and free from CO,, were drawn over a definite area of peritoneal surface, and the CO, which ex- haled was absorbed by the baryta. This slow current of air was continued for a half-hour. Then the baryta water was transferred to a flask. After sedi- menting over night, 25 c.c. of the clear supernatant fluid, and an equal quantity of the original baryta water were titrated with a solution of oxalic acid (2.808 gm. in a litre of water, 1 c.c. of the acid representing 0.5 c.c. of CO, at o° and 760 mm. of mercury pressure). | During the first half-hour period in two of the experiments the omentum was left covering the intestines. The exhalation of CO, under these conditions amounted to 7.2 c.c. from one cat and 6.4 c.c. from the other. In both of these experiments the funnel was then removed, the omentum pushed off of the viscera, and the funnel replaced as nearly as possible in its previous position. During the ensuing period of a half-hour the exhalation of CO, amounted to 17.4 and 15.0 c.c. respectively. The area of viscera exposed was 80 sq. cm. In two other experiments the omentum was removed at the beginning, but only 50 sq. cm. were exposed under the funnel. The exhalations from this area during three successive periods of a half-hour each were 10.8, 7.8, and 6.0 c.c. respectively, in the first experiment; and in the second 8.8, 6.3, and 6.3. Only a slight degree of congestion developed in the intestines. 24 CritE: Loc. cit. Acapnia and Shock. Wey. From these data the exhalation from the abdominal viscera into the air during the first half-hour under the conditions of the ex- periments may be estimated at 0.15 to 0.20 c.c. of CO, per square centimetre of visceral surface exposed, and less than half this quantity through the omentum. The total amounts exhaled are insignificant in comparison with the entire CO, production of the body. A slight diminution in the amplitude of respiration would counterbalance the exhalation from the viscera, and thus prevent the development of a general acapnia. But in the tissue exposed (as the last two experiments show by the lessening exhalations in successive periods) the exhalation reduces the CO, content con- siderably below the normal. Furthermore, these experiments de- termine only the minimum rate of loss, for the viscera were quies- cent and somewhat cool, and the current of air was very slow. No comparable exhalation of CO, occurs through the intact skin. penierpeck 7°) found an output of 0:35 em. (or 182.1 c.c. atuee and 760 mm. pressure) per hour from the entire skin at tempera- tures at which there was no sensible perspiration, and three to four times as much with perspiration. Taking the surface of the skin (for a body weight of 65 kilos) at 2 square metres, the output per sq. cm. of skin per half hour would be only 0.0045 c.c. of COs. The minimum rate of exhalation from a square centimetre of exposed peritoneum is forty times more. IV. Tue Errects oF AERATION OF THE VISCERA. When the abdominal viscera, or indeed any tissues, are exposed to the air, they undergo a loss of tonus and become congested. These changes occur also, although in a less degree, when the intestines are placed in a bath of warm saline. If the exhalation of CO, is a factor in these phenomena, we should expect that the conditions which would induce or prevent them could be inferred from the principles governing the diffusion of this gas. From a mass of wet paper pulp impregnated with CO, the exhalation would be more rapid if the mass were kept warm than if it were cold, more rapid if the surface were kept moist than if it became dry, more rapid if it were continually kneaded than if stagnant, more *° SCHIERBECK: Archiv fiir Physiologie, 1893, p. 119. 78 Yandell Henderson. rapid if it were placed in a draft of air or a stream of water than if the surrounding medium were quiescent, and nearly as rapid if the mass were under water (or physiologic saline) as it would be in air, for the coefficient of solubility of CO, is 1. Applying these considerations to the problem of the local inhibition of peristalsis, it is evident that the point last mentioned would afford an explana- tion of the fact (otherwise puzzling) that peristalsis is not much more persistent in a bath of warm saline than in air. Thus Meltzer finds proof of a local inhibition in the intestine — which we would explain by the local acapnia — in the following experiment: After destruction of the spinal cord in the rabbit the peristalsis in the cecum is exaggerated by the absence of the normal inhibitory control. The mere act of opening the abdomen does not, as it does in a normal rabbit, cause immediate cessation. Nevertheless very soon after exposure the movements subside, and “even in a warm bath of a physiologic solution the cecal peristalsis ceased after ten or fifteen minutes.” The three experiments cited in section II of this paper show that when both local and arterial acapnia are prevented even the more difficult task of maintaining motility in the large intestine of a dog with the spinal cord intact is readily accomplished. That this task is difficult under ordinary experi- mental conditions is shown by the fact that in the careful experi- ments of Eliot and Barclay-Smith, even after destruction of the spinal cord, anti-peristalsis was never seen in the colon of dogs.*® Professor W. B. Cannon tells me that he has recently observed two cats in which gastric peristalsis did not occur spontaneously after destruction of the-spinal cord and opening of the abdomen under saline. He found that the attachment of a piece of rubber tubing (25 cm. in length) to the trachea was effective as a means of initiating motility. The effects of exposure of the viscera, and particularly the in- tense vascular congestion, have been regarded as the results of mechanical irritation, cooling, and drying. Doubtless many in- vestigators have felt the inadequacy of these three causes to ac- count for all the phenomena observable, but no other factor has been thought of. In order to determine whether exhalation of CO, is this unrecognized fourth cause, we devised a method of treatment for the viscera (by means of the apparatus shown in 26 Eviot and BARCLAY-SMITH: Loc. cit. Acapnia and Shock. 79 Fig. 1) which fulfils the conditions, mentioned in the last para- graph, under which the diffusion of this gas would be most rapid. It consists in passing over the viscera a gentle current of air warmed to 35° or 38° and saturated with moisture. It affords an almost crucial test of the hypothesis of acapnial shock. If diminution of the CO, content were of no importance in loss of tonus, the conditions under which the viscera are placed by this Ficure 1.— This apparatus consists of a 74; horse-power electric motor, a rotary air fan, and a flask in which water is boiled. Steam is discharged into the tube from which air, thus warmed and moistened, is directed upon the viscera. For cats one triple burner is placed under the flask, for dogs three large triple burners (7. e., 9 Bunsen burners). For cats the tube is a piece of bicycle tire (‘‘single tube”). For dogs it is a piece of steam fire-engine hose 8 cm. in diameter. method would be nearly ideal. A surgeon who watched one of our experiments without knowing its purpose asked whether it would not be advantageous to use the method in the operating-room of a hospital, since cooling and drying are eliminated without me- chanical irritation. The current of warm moist air is apparently so mild that the hands of the operator after exposure to it for two or three hours are not in the least reddened, nor is their skin softened or puckered. Yet in viscera exposed to it intense con- gestion rapidly develops. Furthermore this treatment is a highly effective procedure for the production of shock. ‘The details of some of our observations on this topic are as follows: These experiments were performed upon cats. Under chloroform anezs- thesia the abdomen was opened by a median incision, and two loops of the small intestine were drawn out. One was held in the current of air. Almost immediately it blushed as markedly as does the ear of an albino rabbit when the cervical sympathetic is cut. Innumerable minute blood vessels previously invisible became apparent on the surface. At first these vessels were bright red, but at the end of fifteen minutes the tissue had acquired a peculiar dark, in- tensely congested, almost bruised appearance, although it had been handled very gently. The tonicity was so much reduced that the loop was of double its So Yandell Henderson. original diameter. Mechanical irritation by pinching between the fingers now caused no contraction. The other loop of intestine had been wrapped in cotton wool moistened with normal saline at 18°, and had been placed so as to lie on the side of the animal out of the current of air. At the end of the fifteen minute period it was found to be only slightly congested, in good tonus, and normally irritable. Parts of the intestine which had remained wholly within the abdomen showed no perceptible alteration; the effects of the aeration were confined to the parts directly exposed. Doubtless mechanical irritation was a contributory factor in the alterations noted in the exposed loop. In another experiment one loop was laid on cotton wool and left untouched by the hand during twenty minutes of warm moist aeration, while a control loop was wrapped in thin sheet rubber to protect it from the air and was handled con- tinuously for this period. Both became congested, but the former underwent by far the greater loss of tonus and motility. The animal was then asphyxiated ; and the exposed loop recovered its tonus, and developed, in common with the rest of the intestine, pendular movements. Altogether, aeration of the intes- tine was tried upon four cats. In three of these experiments the exposed loop was later placed in warm saline saturated with CO,. A few cubic centi- metres of this gas were also introduced into the lumen of the gut by means of a hypodermic syringe. In two of the three cases a marked and rapid recovery of tonus and motility were observed.”' V. Tue PRODUCTION AND RELIEF OF ACAPNIAL SHOCK. We have applied the treatment of warm and moist aeration to the entire abdominal viscera in experiments upon 15 dogs. Two examples of the production of shock in these experiments are de- scribed below. They are selected especially because in both a rapid recovery from the condition of shock was obtained by means of measures designed to relieve acapnia. In neither case was the condition of shock extreme — although both were near the point beyond which there is no return — before relief was begun. We have found that in other cases, in which the failure of the circu- lation had progressed further than in these, similar measures were not ultimately successful. The reason for the ineffectiveness of all methods of restoration in cases below an arterial pressure of 30 or 40 mm. will be shown in a later paper by data from blood-gas analyses. These two experiments are, however, especially signifi- * This result was obtained also by Boxar: Archiv fiir experimentelle Pathologie und Pharmakologie, 1887, xxiii, p. 209. Acapnia and Shock. SI cant because of the rapidity of the recovery which they show. Without measures for the replacement of the body’s store of CO, the restoration of the animals to normal condition would have been a matter of many hours, if indeed it had occurred at all spon- taneously. Even more important, however, is the demonstration which they afford that an animal is readily brought to the verge of shock by the mild procedure of aeration of the viscera. This fact is to be contrasted with the demonstration by Sollmann, Brown, and Williams ** that pouring concentrated nitric acid or caustic soda into the abdominal cavity produces little effect upon respira- tion or arterial pressure within a period of an hour. Thus it appears that corrosion of the viscera is less effective than aeration as a means of inducing shock. In Fig. 2 is shown the arterial pressure curve obtained in one of these experiments. The subject was an animal of the type which best withstands operations, a bull-dog in good condition. Yet aera- tion of the viscera for three hours lowered the arterial pressure by 50 per cent, and diminished the amplitude of the pulse curve to an even greater degree. The pulse in the femoral artery became barely perceptible to the finger. After the first twenty minutes of aeration the animal became completely comatose and no further administration of anzsthetic was needed. In the production of this condition the stimulation of afferent nerves seemed to play no part, for respiration was never accelerated. In this respect this experiment is rather exceptional. The amplitude of the respiratory movements was diminished to as great a degree as the pulse. The small intestine changed color from pink to a dark purple; it re- laxed so completely that its diameter was doubled, and it ceased to respond to a pinch. The mesenteric veins and the venous radi- cles at the base of the intestine became engorged and nearly black. During the first half-hour after the aeration was discontinued and the abdomen closed, no spontaneous improvement in the condition of the animal occurred. Then 200 c.c. of Ringer’s solution were slowly (during ten minutes) injected into the femoral vein. This fluid had previously, and while cold, been shaken thoroughly in a flask through which CO, was bubbled; it had then been warmed to 35°; and at the time of injection it was not only saturated with the gas, but contained also numerous small bubbles. The animal’s °8 SOLLMANN, Brown and WittrAms: This journal, 1907-1908, xx, p. 74. 82 Yandell Henderson. respiration increased in amplitude almost immediately. A tube I5 mm. in diameter and 0.5 metre in length was attached to the trachea, and additional pieces of tubing gradually added until the total length was 1.5 metres. The respiration became deep and full, the heart rate slower, and arterial pressure began to rise. Ringer’s solution saturated with CO, was introduced’ into the peritoneal cavity, anda stream of CO, gas from a Kipp generator was bubbled through this liquid from a tube inserted deep among the viscera. At the end of twenty minutes the animal had come out of coma, so that ether had again to be administered. Arterial pressure had risen considerably, and the pulse was of nearly normal amplitude. After an hour the Kipp generator was shut off, and the tube was removed from the trachea. Within five minutes after these changes respiration had again become shallow, the amplitude of the pulse had diminished, its rate had increased, and arterial pressure had begun to fall. The tube was then replaced on the trachea, and 150 c.c. of Ringer’s solution saturated with CO, were injected into the femoral vein. A half hour later the tracheal tube was finally removed. Thereafter the condition of the nervous system, the circulation, and respiration were almost as nearly normal as in the period.six hours earlier and before the abdomen had been opened. The viscera had recovered their normal appearance and tonus. A demonstration of the production and relief of shock which differs in some respects from the experiment above discussed is afforded by Fig. 3. During the aeration of the abdominal viscera coma. developed, respiration became shallow, the pulse narrow, the femoral artery constricted,2° and the intestines congested. At the end of one hour the animal was by all these signs in a state of shock, yet arterial pressure had not fallen. Tracheotomy was then performed, and the sciatic nerve was pinched with a pair of artery forceps. Vigorous hyperpnoea resulted, and arterial pressure fell gradually. Ether was administered during this period in quantities sufficient to maintain anzesthesia, but not in such amounts as to account at all for the fall of arterial pressure. At the end of twenty minutes of hyperpnoea the pressure had dropped from 140 down to 55 mm. of mercury. The forceps were then removed from the *® For literature showing that the fall of arterial pressure in shock is not due to arterial relaxation (7. e., vaso-motor failure), but on the contrary that the arteries are constricted, see HENDERSON, Y.: This journal, 1909, xxiii, p. 362. Acapnia and Shock. “Sve aSO]D vaysIVI] oY} JB SB UOMIpuod poo’ sv ut aq 0} porvodde yee oy) oyeI as~nd prides orout & 105 }daoxa pur ‘parins.0 pry asdejai ou g iy ‘pp Sud] Ul o1}JOU PUL PEedoulat SBM IqN} OY, “YIOYS PUB VWIOS Jo 9384S ay} WOIZ palAOraI peYy [eUMIUL ay QS'G Ty “Baye ay} 0} payoeye sea yy B puv JojoeIp “Ul “WILT CT aqn} B OE'¢ WW “UIEA [eOWAy oy} OWT payoafut A[Mo]s araM 7OD YUM surjqqnq pue poyeinyes uonnjos s,1a3uny jo ‘9 0ZT 92'S FV “peonpuy snyi sea eaoudisdAFy ‘speatoqur yuanbeay ye poyourd sem aasou ones 94} 07'S [HUN ¢ WoIg = “pautoyied sem AuI0}0 PPE SIV “OCS [HUN poyesoR oJ9M IOISIA oy} pure ‘OC's Iv pouiado sem UsMIOpgR ayy, “Ja}oWIOURUT aTyIINF & Aq papsodea4 aAind ainssaid pyorea “PUCoeS $0 UT plooat SUT, “AOYI9 Voy} “WOFOIO]YO ysay “WS ¢Q'9 ‘oyeYydjns urydiopy ‘sojry Y's Jo Soq “LOGI ‘ST [dy jo yuowiedxy —¢ anor 96°9 00" og'e 00'9 0s'g oF'9 oF’ ae'g | 96'9 02°2 ar'g a a are of ore ie) OS ool OSI 00¢ WW ‘asind ay} Jo apnyydure ay} ym Asnouosyouds pur ‘se Ajjeois se pouea “1asamoy ‘uoresidsor Jo apnyydure eyL ‘omuru rod ET uvy} OW JOU TT UeY) sso] JoAOU svM UONRAIdsos jO 9}R1 9} justuiadxa a1jua oy} NoYsnoIYT, “pasinss0 asdejar JYySys & pu saynurur aay Joy ponunuoosip cam Ady} OTS FY *(3Xe} aS) poynqsur o1aM sanssy PUL poolq ay} Jo Suryeuoqivd-a1 oy) 10J Sainsvoul YO'p TY ‘“peso[d uatopqe oy} pue pooejdar a1am eiaosiA oy} UsYy AM ‘¢Z’¢ [UN payelav o19M BIBDSIA oy} pur “O¢'ZT 1 peuedo sem usuopge ayy, “JoJoWOULUT a[yJINFY v Aq pop4soded dAIND aInssoid PHOIBD “spuodsas ¢*Q UI posal aT, “19YJ9 Way} “WULOFOIO]YD “WS SO'D eeYdins urydiopY “So[Py SOT JO Zopyinq snos0SiA “{O6T ‘8Z “eq Jo ywouedxy “(azis jeurSu10 ay Fey Moqy) —"z aunorg org 00°F 08'S 00°9 S4 Yandell Henderson. sciatic, the viscera replaced in the abdomen, and the cavity closed. During the next ten minutes the animal ‘seemed on the verge of respiratory failure, but arterial pressure rose slightly (to 65 mm.). Then 120 c.c. of Ringer’s solution, saturated and bubbling with CO,, were injected into the femoral vein. As soon as the in- creased depth of respiration induced by the CO, had developed, a tube 15 mm. diameter and 1.0 meter in length was connected with the trachea. At the completion of an interval of fifteen minutes occupied by the processes of recovery (1. e., “ recarbon- ization”’ of the tissues) respiration was deep and full, the coma was relieved, and the administration of ether again became neces- sary, the pulse regained its normal amplitude, and arterial pressure rose to 150 mm. of mercury. The tube was then removed from the trachea, and the wound closed. No relapse occurred, and ten minutes later the animal was in every respect, except a more rapid pulse, as nearly in normal condition as prior to the opening of the abdomen two and a quarter hours previously. V. CONCLUSIONS. I. Acapnia, due to hyperpnoea, plays an important part in the central inhibition of peristalsis occurring under surgical operations. Local acapnia, due to direct exhalation of CO., is a factor in the loss of tonus in exposed viscera. II. When loss of CO,, both by way of the lungs and by direct exhalation, is prevented, and the blood gases are maintained nearly normal, peristalsis can be directly observed in the stomach and in the small and large intestines. Ill. The minimum rate of exhalation of CO, from exposed peritoneal surfaces is 0.15 to 0.20 c.c. per sq. cm. in the first half hour, or 40 times the rate from the skin. IV. Exposing the abdominal viscera to a current of air at body temperature saturated with moisture rapidly induces congestion and. loss of tonus and motility. V. Aeration of the viscera in this manner is an effective method for the production of shock. Restoration of the body’s store of CO, is effective as a method of relief from all except the extreme stages of acapnial shock. VI. These observations and others which are to be presented in later papers indicate that the CO, tension in the nerve centres Acapnia and Shock. yo) 5 and in the tissues and fluids of the body is a factor in the mainte- nance of tonus (in the broad sense of the word) of the same order of importance as temperature, oxygen supply, osmotic pressure, and the equilibrium of anions and kations.*° I am indebted to Prof. Walter B. Cannon for valuable criticism upon the manuscript of this paper. Preliminary note of further investigations upon the effects of acapnia. — I have obtained data which indicate that one of the con- ditions requisite to the contractility of the uterus is the normal CO, content of the blood. The tonus and motility of this organ also appear to be inhibited by acapnia, and accentuated by hypercapnia. In normal labor the pulmonary ventilation, in spite of the suffering, is not continually excessive. During each “ pain” the partial pres- sure of CO, in the alveolar air of the lungs is even increased by the spontaneous holding of the breath. The obstetrician is able to some extent to control the duration and vigor of the contractions by com- manding or discouraging this apncea. It is probable that the in- effective ‘‘ pains”’ of prolonged labor are in part due to inability of the subject to regulate respiration properly, 7. e., to inhibit the res- piratory centre from responding to the intense stimulations of affer- ent nerves. Hyperpnoea occurs, acapnia develops, and diminution of uterine tonus and motility result. Upon this topic and upon the suppressions of gastric and pancreatic secretions by acapnia in- vestigations are under way. 30 Cf. HENDERSON, L. J.: This journal, 1908, xxi, p. 427. THE ROLE OF THE ASH CONSTITUENTS OF WHEAT BRAN IN THE METABOLISM OF HERBIVORA.! By E. B. HART, E. V. McCOLLUM, ann G. C. HUMPHREY. [Contribution from the Agricultural Chemistry and Animal Husbandry Departments of the University of Wisconsin.] N a previous publication? it has been shown that wheat. bran contains phytic acid in combination with potassium, magnesium, and calcium. In a later publication ? it was demonstrated that this complex was in all probability responsible for the well-known laxa- tive effect of wheat bran, —a phenomenon previously held to be due to the mechanical irritation occasioned by the coarse and fibrous construction of this material. In the course of the above investi- gation with milch cows data were procured indicating in certain cases a specific physiological function for this complex. Beside this laxative effect which manifested itself in all experiments, its withdrawal from the ration occasioned more or less disturbance of fat. production, a regular reduction in the volume of urine pro- duced daily, and an increased flow of milk. In addition to the above enumerated facts, withdrawal of phytin from a ration dis- turbed the cestrum periods of certain individuals, while in other individuals this disturbance was but occasional. During this early investigation a limited amount of data was collected on the metabolism of the bases associated with phytin, and their channels of excretion. It is indeed entirely conceivable that the path of elimination of the ingested bases associated with phytin is responsible for such phenomena as the laxative effect and the ' Published with the permission of the Director of the Wisconsin Experiment Sta- tion of the University of Wisconsin. ? Patten and Harr: Journal of the American Chemical Society, 1904, xxxi, p. 564. % Jorpan, Hart, and Patten: This journal, 1906, xvi, p. 268. 86 Role of Ash Constituents of Wheat Bran. 87 increased urine flow. The reversal of these two phenomena invari- ably took place on the withdrawal of phytin from the ration of milch cows. The other effects, such as disturbances of fat pro- duction, flow of milk, and cessation of the cestrum periods, — func- tions imbedded deeply in the maternal nature of the animal, — were not always highly pronounced, and when such disturbances did take place, are to be considered individualistic and at present unexplained. If it were possible to induce these last three disturb- ances with all individuals, then the specific relation of phytin to these physiological processes might be seriously considered, but such does not appear to be the case. It was to study further the action of the components of the phytin complex when administered separately as salts, their chan- nels of excretion and general relation to the phenomena of con- stipation and diuresis, with the consequent effect on milk secretion, that this investigation was undertaken. The scope of the following experiment involved: 1. A study of the metabolism and channels of excretion of the base and acid constituents of phytin, as well as equivalent quan- tities of these bases in the form of chlorides and sulphates. 2. The effect of the supply and the channel of excretion of these elements on diuresis, the character of the feces and milk flow, and the complete composition of the milk. 3. The effect of withdrawal of phytin and the reduction of the crude fibre content of the ration on the character of the excreta. 4. Physiological action of prepared potassium phytate. GENERAL PLAN. 1. Feeding to the same animal during short and long periods of time, rations differing greatly in the amount of phosphorus, magnesium, or potassium, as phytin, or rations low in natural phytin, but supplemented with magnesium and potassium, as a chloride or sulphate, or potassium as a phytate. 2. A reduction of the amount of the crude fibre in the ration of low phytin content secured by lowering the proportion of washed bran and increasing the starch content of the ration. 3. Abrupt or gradual changes from rations of high phosphorus, 88 E. B. Hart, E. V. McCollum, and G. C. Humphrey. magnesium, and potassium content, to rations low in all or only one of these ingredients. 4. The nutritive plane of the rations to be maintained the same except in the variation of the ash constituents. 5. The ration fed to be carefully weighed and samples of the excreta quantitatively collected and preserved for analysis. EXPERIMENTAL PART. The cow selected was a vigorous grade Holstein, in fair flesh and with a ravenous appetite. Her keen appetite secured a com- plete consumption of the ration during the entire experiment. The animal was kept in an especially arranged and warmed room. She was fed from a tight box which would allow recovery of all un- eaten food. The daily ration was given in two equal proportions, morning and night, and water offered at definite times. She was weighed daily; all excreta were quantitatively collected. The weights represent what was voided during the twenty-four hours from 6 A.M. The animal was milked at a definite time, twice a day, morning and night. The rations employed were made up, as in previous experiments, of oat straw, wheat bran, rice, and wheat gluten. This ration affords a high phytin intake, consequently a high phosphorus, mag- nesium, and potassium consumption. A basal ration, low in phytin and consequently low in phosphorus, magnesium, calcium, and potassium, was secured by extracting whole wheat bran with water after a period of soaking. The calcium content of the basal ration _is somewhat increased over that of the standard ration owing to the percentage increase of this element in the washed bran. Never- theless the daily consumption of this element is below the daily output. With the basal ration as a starting-point additions of materials whose influences were to be studied could be secured. Sufficient feeding materials were prepared or set aside for a presumed length of the experiment and carefully sampled, for analyses. The milk, urine, and feces were weighed and sampled after being carefully mixed. Reserve samples of milk and urine, preserved with formaline and toluol, were set aside. Five-pound samples of fresh feces were dried at 60° C. for reservation. Nitro- gen determinations were made on the fresh feces. Standard Role of Ash Constituents of Wheat Bran. 89 methods of analyses were used in most cases. Fat was determined by the Babcock method, and casein by the centrifugal process checked at intervals by the chemical method. It is only necessary to state that this method gave close agreement with the chemical determinations. SEQUENCE OF RATIONS. An initial period of two weeks was consumed in adjusting the animal to the ration used. Actual records began November 26. Ration 1. Fed from November 26 to December 5s. Transition Period. This was begun December 5 and was com- pleted on December 9. Previous experience had shown that a sudden withdrawal of phytin from the ration resulted in constipa- tion. To avoid this effect the withdrawal was made gradual by the daily substitution of two pounds of washed bran for two pounds of whole bran. The wheat gluten was slightly increased from day to day until on December 9 the animal was receiving two pounds. Her appetite remained keen. Ration 2. December 9 to December 23. Ration 1. December 23 to December 30. On the evening of December 23 the animal was suddenly changed to the ration of whole bran. Ration 2. December 30 to January 11. December 30 a sudden change from the whole bran to the washed bran ration to which was added 135 gm. of potassium sulphate and 200 gm. of magne- sium chloride. These quantities of salts supplied approximately the amounts of magnesium and potassium withdrawn by substitut- ing washed bran for whole bran. Ration 1. January 11 to January 21. This change was made suddenly. Ration 2. January 21 to February 1. A sudden change from the whole bran ration to the ration in which the crude fibre intake was equivalent to that of ration 1 was made on January 21. This ration of lower crude fibre content was secured by reducing the intake of washed bran from Io to 5.3 pounds. In addition 4.7 pounds of wheat starch were substituted for the withdrawn washed 4 The thanks of the authors are due Mr. Harry STEENBOCK for valuable so aS 13:87 59!4 TOGO 18:3 || 42:3 | 11.6 | 20.0 | 0.45 | 112 | 24.6 | 65:6 | 35.1 | 56.1 | 159.2 31.0 | 19.4 fer | 28:7, .64 5.8 2.9 | 71.9 | 58.0 | 44.6 | 13.3 | 126.2 77 | 20:1 | 38.9 22.8 26.5 | 19.2 | 16.5 | 34.4 32.4 | 22.8 | 23.2 | 39.8 drawal of phytin. From this it appears that when such disturb- ances have been produced, as in previous experiments, they must be considered as individualistic, but that we do not have in the body phytin a specific chemical entity directly regulating and imperatively concerned in the process of fat production. It is of course not here questioned but that the components supplied by phytin are important sources of the ash constituents in the ani- mal’s metabolism. With this animal there was also no disturbance in the flow of milk consequent upon the withdrawal of phytin or any other sup- plied salts. The flow of urine was directly related to the supply of phytin, as well as to certain of its components, when these were supplied as salts in the ration. This was equally manifested where the dis- placed phytin was substituted by potassium and magnesium as 100 E. B. Hart, E. V. McCollum, and G. C. Humphrey. sulphates and chlorides, or by potassium alone as a chloride or phytate. The increased output of potassium in the urme which always accompanied an increased intake strongly suggests a close relation between the diuresis produced with the whole bran and its high potassium content. The percentage of potassium in the urine, however, does not remain constant, even with increased flow. When the volume increased, the percentage of potassium often rose to I per cent, while in the period of low output it was but one tenth that amount. The constipating effect incident to withdrawal of ong was always manifest. When, however, the phytin was replaced with magnesium sulphate, a laxative effect was produced, but when this substitution was made with potassium sulphate or chloride, an unmistakable dryness of the feces resulted. Again, when the dis- placed natural phytin was substituted by a prepared potassium phy- tate, a laxative condition resulted. A glance at Table VIII will reveal the fact that the principal channel of excretion of magnesium in the cow is the gut. The inference from a consideration of the facts lends strong color to the theory that the phosphorus and magnesium carried by whole wheat bran are largely responsible for this action. It is to be observed that when the washed bran was supplemented with crude phytin, the excretion of magnesium in the gut was increased. The lime supply in the ration of the entire period was manifestly deficient. The output was approximately 50 gm. daily, while the intake was but 25 gm. The popular notion that wheat bran is particularly useful as a building material for growing animals, due to high ash content, needs qualification. It is high in total ash, but its content of lime is relatively low. Ten pounds of wheat bran supplied but 8 gm. of calcium oxide. This period of feeding covered one hundred and ten days, and consequently entailed an approximate loss of 2500 gm., or 5% pounds of lime. The data from the Rothamsted Experiment Station on the ash constituents of various animals affords an approximate esti- mate of the total lime in our animal: At the beginning of the experiment it was about 24.2 pounds. This means that during the period of our experiment there had been a loss of about 25 per cent of the entire lime content of the animal. A certain and definite percentage content had been maintained in the milk, and an apparent waste, possibly indicating general cell metabo- Role of Ash Constituents of Wheat Bran. 101 lism had been excreted in the feces arid urine. This large loss above that used for milk production could have had no other source than the skeleton. This supports what our experiments with pigs have shown, namely, that the skeletal tissue can vary its ash content within quite wide limits, thereby acting as a supply house over considerable periods of time for certain ash constituents that may be deficient in quantity in the food. During the periods of high phosphorus feeding there appeared to be a storage of phosphorus and especially of magnesium. This condition prevailed more particularly during the later periods of the experiment. This would make clear how it was possible to withdraw the needed supply of lime from a calcium-phosphorus complex and still retain the phosphorus. However, it is also possible that the constant withdrawal of lime did not involve the calcium phosphate of the skeleton, but that the calcium carbonate, which is also supposed to be present, was involved in this interchange; but this is mere supposition. During periods of low phosphorus feeding there was constantly an increased output of calcium in the urine, apparently again in- volving the metabolism of a calcium-phosphorus entity. The total output of calcium was greater during these periods than during periods of high phosphorus feeding. It is apparent from the data that the daily loss of calcium oxide due to milk production and cell metabolism was at least 50 gm. The animal weighed 1150 pounds and produced about 30 pounds of milk daily. The periods of low phosphorus feeding also afford data on the output of this element at periods of phosphorus starvation. The average amount of phosphorus pentoxide lost daily through the milk and cell metabolism was approximately 60 gm. It is an interesting fact that in spite of deficient supplies of these two elements during considerable periods there was nevertheless an apparent waste. A part was constantly being metabolized and passed out of the reach of the reconstructive processes. Whether this means that the form in which the ash elements are presented to the cell before metabolism is different from that existing-after such processes have occurred, is, of course, mere conjecture. The supply of magnesium and potassium compounds in the food at all periods was equal to or greater than that excreted. 102 E. B. Hart, E. V. McCollum, and G. C. Humphrey. SUMMARY. 1. A high potassium intake accompanied by a high phosphorus intake gave a high potassium content in the feces, although a con- siderable portion of potassium was also excreted in the urine. 2. A high potassium intake accompanied by a low phosphorus intake gave a low potassium content in the feces, with a high out- put of this element in the urine. 3. A high potassium intake accompanied by a low phosphorus and a high magnesium intake gave a high potassium output in the urine. 4. Magnesium when supplied as a chloride or as a phytate was largely excreted in the gut. 5. Phosphorus and calcium were also principally eliminated by this channel. 6. A low phosphorus intake was accompanied by a high calcium output in the urine. 7. A deficient calcium intake was nevertheless accompanied by a considerable output of this element in the gut. This same state- ment is also true of phosphorus. 8. When calcium or phosphorus was deficient in quantity in the food, the skeletal tissues appeared to be ready sources of supply. The average quantities of calcium oxide and phosphorus pentoxide metabolized and excreted daily by this animal during periods of deficient supply were, respectively, 50 and 60 gm. g. The supplies of potassium and magnesium were in all periods equal to or above the amounts eliminated. 10. Variations, within wide limits, in the form and quantity of supply of potassium, magnesium, or phosphorus, did not influence the percentage content of these elements in the milk. 11. With this animal there was no appreciable fluctuation in the percentage of organic constituents in the milk relative to the supply of phytin. 12. Marked diuresis was produced by the quantity of phytin supplied. A high potassium and magnesium intake, as sulphate and chloride, produced a similar effect, as did potassium alone when supplied as a chloride. This would indicate that the high potassium intake accompanying the whole bran ration was re- sponsible for this phenomenon. Réle of Ash Constituents of Wheat Bran. 103 13. Sudden withdrawal of phytin produced constipation. This was even manifested when the intake of crude fibre was reduced to that of normal bran. 14. The laxative action is more easily understood when it is remembered that the channel of excretion of phosphorus, calcium, and magnesium especially, and a part of the potassium, when sup- plied in wheat bran, is by way of the gut. 15. The “margin of safety”? provided in the skeletal tissues in the animal precludes against immediate disastrous results conse- quent on a sudden deficit in the intake of phosphorus or calcium. THE EFFECT OF SMOKING UPON THE BLOOD PRES= SURES AND UPON THE VOLUME OF. THE HAND? By JAMES W. BRUCE, JAMES R. MILLER, anp DONALD R. HOOKER. [From the Physiological Laboratory of the Johns Hopkins University.] HIsToORICAL, HE effect of nicotine upon the cardio-vascular system is very well defined. Wertheimer and Colas? state that the acceler- ation of the heart rate which follows the injection of nicotine is present whether the extrinsic nerves to the heart are intact or severed, and that an increase of arterial tension occurs even after the destruction of the medulla and spinal cord. The action upon both the heart and blood vessels is, however, greater when the nervous system is intact, demonstrating a central as well as a pe- ripheral effect. Langley * concludes, in an investigation published in the same year (1901), that “nicotine stimulates sympathetic nerve cells; the evidence that it paralyzes them is incomplete.’ * He observed, further, that nicotine in dilute solutions (0.1 per cent in 0.7 per cent NaCl) when applied to sympathetic ganglia produces excit- ing effects of longer duration, although slower in onset than in stronger solutions. Of recent years attention has been focused upon the action of tobacco, especially of tobacco smoke. Cushny ® states that in what- ever form tobacco is consumed, nicotine absorption occurs. That Reported before the American Physiological Society, December, 1908. WERTHEIMER et Cotas: Archives de physiologie, 1891, p. 341. LANGLEY: Journal of physiology, 1901, xxvii, p. 224. Lest this statement, as quoted, should be misleading in regard to the work of Langley and others in tracing nerve paths by the use of nicotine, we ought perhaps to call attention to the fact that the paralyzing action of nicotine is presumably, in large part at least, upon the endings of the preganglionic fibres. ® Cusuny: Pharmacology and therapeutics, 1903, p. 276. - 104 1 2 3 4 oe — Se. Effect of Smoking upon the Blocd Pressures. 105 tobacco exerts a deleterious effect upon the animal organism when administered in sufficient quantity is evident from the recent litera- ture on the subject. Thus Richon and Perrin ® observed that the subcutaneous injection of the aqueous extract retarded growth in rabbits. After the injections were stopped the animals continued to develop normally. Fleig’ studied the effect of inhalation of smoke, the subcutaneous injection of aqueous extracts of smoke and of nicotine, and the salts of nicotine on rabbits. The results of such procedures were qualitatively the same, causing the animal to abort or to produce weakly viable young. When the dosage was low, the effect was less pronounced, and with weak fumes the young were able to return to normal some time after the treatment was stopped. Lesiur® states that denicotinized tobacco is entirely without toxic effect upon animals. This statement is supported by the work of Lehmann,°® who concludes that nicotine is the most important and practically the only toxic substance in tobacco smoke. Investigating the effect of tobacco on man, Boveri! has noted a distinct effect upon the muscular power of healthy individuals after smoking. He observed, using Mosso’s ergograph, that the use of tobacco resulted in a short, preliminary period of slightly in- creased power, followed by a long period during which the power was distinctly less than in control experiments. Finally, mention should be made of the clinical picture known as Toxic angina."! This condition is apparently limited to smokers, and is attributed to disturbances in the coronary arteries brought on by the use of tobacco. Upon the blood pressure and heart rate the inhalation of tobacco smoke acts similarly to the injection of nicotine. Thus Fleig and De Visme 1? give a tracing which shows the effect upon a curarized dog. After a transient fall in blood pressure, accompanied by a decrease in the heart rate and a decrease in the volume of the kidney, there is a considerable rise in carotid blood pressure, co- 8 RICHON et PERRIN: Comptes rendus de la Société de Biologie, 1908, Ixiv, p. 563. 7 FLe1c: Comptes rendus de la Société de Biologie, 1908, Ixiv, p. 683. 8 LesturR: Comptes rendus de la Société de Biologie, 1908, Ixiv, p. 9. ® LEHMANN: Miinchener medicinische Wochenschrift, 1908, lv, p. 723. BovertI: Deutsche medicinische Wochenschrift, 1906, xxxli, p. 1439. OsLER: Practice of medicine, 1gor, p. 764. Fie1c et Dr Visme: Comptes rendus de la Société de Biologie, 1907, Ixiii, p. 578. 10 11 12 106 J. W. Bruce, J. R. Miller, and D. R. Hooker. incident with an increase in the rate of the heart. As the blood pressure begins to rise, the volume of the kidney momentarily increases. This slight increase in renal volume is followed by a distinct decrease in volume which occurs while the arterial pres- sure is still rising. Finally, dilatation of the kidney sets in before, and appears, on the tracing, to initiate the fall in arterial pressure. The preliminary effect in this experiment is undoubtedly, as Pachon 1% points out, due to vagus action and presumably is of central origin. Attention may also be called to the clinical view hat tobacco is one of the etiological factors in the production of arterio-sclerosis. Numerous attempts have been made to produce arterio-sclerosis experimentally in animals by repeated injections of nicotine and by feeding extracts of tobacco.1*— This work has resulted in demon- strating degenerative changes in the muscular coats of blood ves- sels unaccompanied, however, by involvement of the intima. The evidence therefore is negative so far as true arterio-sclerosis is concerned. Whether the injury to the muscular coats is the result of mechanical or toxic effects cannot be stated. The only reference to the immediate effect of smoking upon the blood pressure in man with which we are acquainted was published by Hesse?® in 1907.1° This observer determined the maximum ‘3 PacHon: Comptes rendus de la Société de Biologie, 1908, Ixiv, p. 116. ‘ SattyKow: Zentralblatt fiir die gesammte Physiologie und Pathologie des Stoffwechsels, 1908, p. 654. 1° Hesse: Deutsches Archiv fiir klinische Medicin, 1907, Ixxxix, p. 565. 6 Since this was written a paper by LEE (Quarterly journal of experimental physiology, 1908, i, p. 335) has become accessible to us. LEE observed a continuous rise in blood pressure (systolic) during the smoking period. Immediately after smok- ing was stopped the pressure began to fall and continued to fa!! until the normal was reached. Moderate and excessive smokers were less affected than novices. In the latter the rise in pressure was sharp, and in one case gave way to a fall of 50 mm. Hg with the symptoms of general collapse. In all of LEr’s experiments (seven in num- ber) the smoke was inhaled. The rise in pressure noted was, for novices, 1o-20 mm. Hg, for moderate smokers about 10 mm. Hg, and for excessive smokers 2-4 mm. Hg. This observer also studied the effect of tobacco smoke upon animals, his results con- firming the previous work cited above, except in the case of one rabbit which was sub- mitted to 70 inhalations of fifteen to twenty minutes each, over a period of five months. This rabbit, which had gained steadily in weight, showed artheromatous changes post- mortem. ‘The conclusion is reached that ‘‘arterial disease may result from prolonged tobacco smoking” brought about by the mechanical injury to the vessels caused by sudden and repeated elevations of arterial tension. r Effect of Smoking upon the Blood Pressures. 107 and minimum blood pressures and the heart rate before and after smoking on about 25 individuals. The maximum pressure was determined by the palpation method, the minimum pressure by Strassburger’s method. The Riva-Rocci instrument was used. Hesse’s results varied somewhat, but he observed that, as a rule, both pressures rose and the heart rate was increased. The maxi- mum pressure tended to rise more than the minimum, so that the pulse pressure was increased. In two cases this author compared the effect of natural with denicotinized tobacco. The effect was practically the same. Also in both of these cases, comparing natu- ral with denicotinized tobacco (the only cases in which the blood pressure was observed subsequently to the use of tobacco), the blood pressure fell to or below normal within twenty minutes. The heart rate also showed a marked tendency to fall below normal, although the change was somewhat more gradual than in the case of the pressures. It is difficult to explain these results except upon the assumption that by far the major part of the effect observed is due to psychical influences, in which case it would not be justi- fiable to assume that tobacco smoke as such exerts any influence upon the cardio-vascular system. This negative inference is sup- ported by the similarity of action between the natural and nicotine- free cigars, although it must be borne in mind that commercial nicotine-free tobacco doubtless contains some nicotine. EXPERIMENTAL. The following experiments were performed to ascertain the demonstrable effects upon the heart rate, arterial pressure, and the volume of the hand which follow moderate smoking. Subjects. — Both were healthy men, accustomed to moderate smoking, twenty-one years old. Methods. — Arterial blood pressure was determined upon the right arm with the Erlanger sphygmomanometer, the palpation method being used for systolic pressure. Volume changes in the left hand were recorded with Mosso’s plethysmograph, a rise in the curve indicating vaso-constriction, and a fall in the curve vaso- dilatation. The work was conducted in a private room, and all extraneous factors were, as far as possible, excluded. The ex- periments lasted from one and a half to three hours, the subject 108 J. W. Bruce, J. R. Miller, and D. R. Hooker. sitting quietly during that time. During the observations it was found convenient to permit the subject to aimlessly draw pictures on the arm of the chair. This procedure offered sufficient diversion to maintain a fairly even psychic condition. In some of the experi- ments an attempt was made to determine venous pressure. The method employed was not, however, suited to the experimental BLOOD-PRESSURE Supyect M. Max. Min. Ie ee PERS B. | De | Al Be) Deck. VB. Dis | eae Baal eee a ie 113 75 38 60 alts 70 43 62 119 75 | 44 84 120 * | | 72 119 75 44 | 72 120 75 | adie see 64 110 ae sane lta 740) ve |+5| 40 = | 41°68 +_ 4 108) | 2 | +8 | “70 ae heouieoe | --° | +3) 60 | 2225 SE aO f12 0.2) 3°)! 70, | oe [463 424) 2 | Olea en 114 | ee Neve el isco all st-o5s | CSO lees 0; 66 | + 18/4 14 | | | Max. = maximum pressure. Min. = minimum pressure. P. P. = pulse pressure. P.R. = heart rate. B = before smoking. D = during smoking. A = after smoking. conditions and was finally abandoned. Pipes, cigars, and cigar- ettes were used. There was no evident difference in the effects, consequently special notes in this regard are not always entered in the protocols. An electric signal was employed to write upon the plethysmographic record, so that the observer could readily indicate the time when events, such as blood-pressure determina- tions and psychical disturbances, took place. Effect of Smoking upon the Blood Pressures. 109 Experiments. — After the subject was entirely accustomed to the experimental conditions, thus insuring in so far as pdssible an even psychic condition, the records of the experiments were pre- served. The number of experiments culminating satisfactorily is not large, nevertheless the results are suggestive. In the accom- panying table we have gathered together all the blood-pressure DETERMINATIONS. SuBJEcT B. Max. Min. eles pavike determinations, together with the observations of the heart rate. It will be seen that, after smoking, there is, as a rule, an increase in the heart rate. The increase is not great and may be absent altogether, even when the other records (blood pressure and hand volume) show distinct changes. The maximum and minimum pressures usually rise slightly, the maximum the more, so that the pulse pressure tends to increase. 110 J. W. Bruce, J. R. Miller, and D. R. Hooker. The following experiments are given in detail, since they cor- respond to the plethysmographic tracings reproduced in the ac- companying plate. In all of the protocols the same nomenclature + m/\; + : + wii 3 i 4 1 he : t+ 1s ww 3 5 4 \ iw 7 9 Ww i2 By 4 W/V Rhos Ma 5 2 9 5 ' 2A3 4 6 9 25 1 10 a Ms tihatt a 7 De een ane ae 40 FicurE 1. — Plethysmographic tracings of the left hand. The Roman numerals refer to the experiments, the Arabic numerals to the experimental procedure. The vertical lines (indicated by +) on the tracings delimit the smoking periods. The curves read from left to right. A rise in the curve indicates decrease in the hand volume. The scale at the side gives volume changes in cubic centimetres. is used. Under ‘‘ Procedure” are given the time of observations and the various experimental steps, together with notes of extrane- ous factors. Under “ Ref. pleth.”” are given the numbers corre- sponding to those on the plethysmographic tracings. “ Max.,” “Min.” and “ P.P.” indicate maximum and minimum arterial blood pressure and pulse pressure. “ P. R.” indicates the heart rate. In the brief discussion of these experiments reference will be made to the plethysmographic tracings. The curve shows (Exp. I) considerable variations, two of which (3 and 4) may be due to the hot water drunk. Its general trend is, however, quite even until the smoking period (5-6), after which it first tends to elevate itself and then to remain constant. The sudden rise just before 5 is presumably of psychic origin. The heart rate remained quite constant. Apparently the peripheral vaso-constriction following the smoking period was either com- pensated or insufficient to markedly affect the blood-pressure read- ings by the method employed. There is, however, a slight in- crease in maximum and minimum pressure and heart rate over the determinations made just before smoking was begun. The curve shows (Exp. III) a sharp and considerable rise just Effect of Smoking upon the Blood Pressures. 111 Procedure. EXPERIMENT I. May 7, ’08. Susyecr B. Ref. pleth. 10.26 A.M WONS2ZPALMi 5s) 10.40 a.m 10.46 am... . MOSSSPAS Msc sae Cup hot water . Second cup water. . 11.10 a. u Third cup water 1U Gy Ais ea 11.30 a.m Smoked two and a half Cigarettes 11.55 a.m Procedure. EXPERIMENT III. May 11, 708 Susyecr M. | Ref. pleth. | WUODPA SM... see Smoking period 11.30 A. um 11.45 a.m Mes ATM. ke 112 J. W. Bruce, J. R. Miller, and D. R. Hooker. preceding and coincident with the beginning of the smdking period. These sudden changes are undoubtedly psychic in origin. It is to be noted, however, that the curve remains consistently elevated through- out the remainder of the experiment. There is a decided increase in both pressures, accompanied by an increase in the pulse pressure. The heart rate apparently played no part in the change. There is a transient increase in rate, amounting to 16 per minute, which disappeared within fifteen minutes and was even converted into a lessened rate without affecting the plethysmographic curve or the blood pressures. EXPERIMENT IV. May 11, 08. Susyecr B. Procedure. Ref. pleth. Moved arm Smoking begun. . . Shows the effect of prolonged smoking. The first part of the curve is quite regular, except for two elevations, which were pre- sumably coincident with the determination of maximum and mini- mum pressures on the opposite arm. Smoking was begun at 3, after which there is distinct evidence of a slow but progressive vaso-constriction. Upon the curve thus elevated, we again see the transient effects of extraneous factors. The blood pressures show little change. There is a slight increase in the maximum, the Effect of Smoking upon the Blood Pressures. 113 minimum remaining stationary, so that the pulse pressure varies directly with the maximum pressure. The heart rate shows a con- sistent tendency to slowly increase. EXPERIMENT VIII. May 18, ’08. Supnject B. Procedure. Ref. pleth. Max. | Min. Pes Pik. LOAORAGMS (2. 2 se 1 92 Z 90-94 be ue a 3 65 27 | Wentebresss Det. - 4 | | 5 93 lit “46 88 Ven. Press. Det. if Talked of smoking 8 11.10 A. M. started pipe | 9 Weepspreathy a5 . - 10 Two deep breaths . | ll Inhaled smoke . . . 12 | Said head dizzy . . 13 Smoking stopped . . 14 is S< se 96 Deep breath . . .. 15 | 100 meta (iii gS 70 30 Wen Press, Det. . - 17 18 BOcepibreath: . . . - 19 =e ae a 90 20 98 Two deep breaths. | 21 l ms 70 28 11.50 a.m. Exp. ended 22 The curve exhibits many slight irregularities which find no ex- planation in the protocol. The most marked of the sudden changes, 114 J. W. Bruce, J. R. Miller, and D. R. Hooker. occurring at 13, is accompanied by the note “ Said head was dizzy.” The effect of smoking was to cause a slight vaso-constriction. Both pressures were affected. There was a slight increase in the pulse pressure. The heart rate was greatest just at the end of the smok- ing period, but approached the control value as the experiment progressed. EXPERIMENT IX. May 18, 708. Susyect M. Procedure. Ref. pleth. i H. left room . . nt S60 SG, H. entered Talked ‘ ee, et 4.06 P.M. cigar . . . H Oo No apparent cause 4.25 p. mM door opened 4.30 p.m. smoking stopped Moved Deep breath Ven. Press. Det. Exp. ended The tendency of this curve to rise (from 5 to 9) before smoking was begun makes the interpretation difficult. The general effect is, however, that the smoking results in vaso-constriction. The Effect of Smoking upon the Blood Pressures. 115 pressures are uncertain, but indicate a slight rise of both maximum and minimum. The heart rate is slightly increased. EXPERIMENT XII. May 19, ’08. Susyecr M. Procedure. Ref. pleth. 118-120 120 118-120 118-121 Ven. Press. Det. This is one of the control experiments. The curve shows a slow and very slight vaso-dilatation throughout the experimental period. The blood pressures are strikingly constant. The heart rate shows a steady decrease. Discussion. — Our results show clearly that a change in the cardio-vascular system follows smoking. Other factors, however, and especially psychic factors, of necessity complicate the results. How far, therefore, we are justified in stating that such change is a tobacco effect depends upon the possibility of excluding these factors. In describing our experimental procedure it was stated that special effort was made to maintain an even psychic condition. . This effort consisted (1) in repetition until the subject was entirely accustomed to the procedure observed and (2) in permitting the subject to aimlessly draw pictures on the arm of the chair during 116 J. W. Bruce, J-.R. Miller, and D: K- Hooker. the experiments. The plethysmographic tracings show in conse- quence only transient irregularities as the result of psychic dis- turbances. As compared with the results which follow smoking these irregularities appear of insignificant moment. Their insig- nificance is further emphasized by the fact that they are not absent after smoking, but appear superimposed upon the elevated curve. The duration of the vaso-constriction which followed the use of tobacco may be conveniently summarized in the following figures: From end of From beginning of smoking period. smoking period. WDE a6 8 | 5 6 Alana 35 min. er. artes of yl Spo 30" <* DV SS sage See ee 60.“ Ve eeu ws) S25) mee Boe TX ee Sd EMSS SS 60 “ It is to be noted also that in none of these tracings is there evidence of a tendency to fall after the constriction has once set in. Un- fortunately, we have no tracings to show the duration of such vaso-constriction, since it was not possible to continue our experi- ments over sufficiently long periods. In this connection we wish again to call attention to the results of Hesse cited above. He observed no difference in the effect upon the blood pressures of smoking natural and denicotinized cigars, and, what is of interest here, both the maximum and minimum pressures returned to or fell below normal (control) within twenty minutes. Our pressure determinations fail to show this quick reaction. Even if we have shown that smoking as such does influence the cardio-vascular system, we do not believe that our evidence is suf- ficient to lend much support to the theory that tobacco is an etio- logical factor in arterio-sclerosis, at least in so far as this theory assumes a mechanical injury to the vessels. The effects of mod- erate tobacco smoking upon a man accustomed to its use would seem to be very little, if any, greater than the effects of those stimuli which are the necessary consequences of civilized life. hae PRODUCTION ._BY .HYDROGEN PEROXIDE OF RHYTHMICAL CONTRACTIONS IN THE MARGIN- EESS BELL OF GONIONEMUS. By ©: Be TERRY. [From the Marine Biological Laboratory, Woods Hole, and the Physiological Laboratory of Purdue University.] HE normal movements of the body of a normal Gonionemus’ consist of a series of five to twenty-five or more rhythmical contractions, or pulsations, which squeeze water from the interior of the bell, and in so doing force the animal, apex of the bell for- ward, through the water. In general these contractions resemble closely the beat of the heart of the higher animals. As in the heart, these pulsations of Gonionemus are brought about by muscles under the control of a nervous system. In Gonionemus the ner- vous system consists of a ring of nervous tissue at the margin of the bell, in which are found many nerve cells, and from which nerve fibres ramify toward the apex of the bell. This nervous system has for one of its functions the initiation and co-ordination of the pulsations. That the impulse to beat normally originates in this ring of nerve cells is easily proved, as was first done by Romanes, by cutting off this margin of the bell. At once all rhythmical movements of the bell cease when that bell is placed in normal sea water. Such a marginless bell constitutes the material upon which the following experiments were performed. In describing the experi- ments the pulsations are called co-ordinated when the muscles of the marginless bell contract simultaneously, giving the appearance of the contractions of a normal non-mutilated specimen. ‘This implies immediate relaxation of the muscles after a contraction. The pulsations are described as inco-ordinated, or clonic, when a wave of contraction starts at some point and travels rather slowly over the whole organism (sometimes back and forth several times). 117 118 OL OE erry: Pulsations are described as tonic when, after a contraction, the organism remains contracted in a small hard ball. Loeb! carried on experiments with a jelly fish, Polyorchis, found along the coast of California. His experiments dealt with the effects of various salt solutions on the pulsations of the marginless bell. It was also found that rhythmical pulsations can be initiated by the addition of a trace of carbon dioxide or any acid to the sea water. “ About 1.5 to 2 c.c. of m/10 H€l to 100 c¢.c. of Seanwaten is required for this purpose. Alkalies have the opposite effect.” I have confirmed these observations regarding the effect of acids and carbon dioxide upon Gonionemus. These results seem to be directly opposed to those upon the heart muscle, in which acids have an inhibitory action and alkalies the opposite. Lillie 7 remarks: “‘ Acids are known, even in low concentrations, to be very injurious to contractile processes: weak solutions of acid quickly suppress muscular activity. . . . Again, it [acid] must be present in relatively low concentration in the external medium.” Lillie says accumulation of acids, produced by metabolic processes, causes polarization of the surface of the contractile elements and inhibits further contraction. This is hard to reconcile with the fact that an excess of carbon dioxide and other acids in the sur- rounding medium stimulates the marginless bells of Polyorchis and Gonionemus to contract. Experiments with hydrochloric acid.? — n/too HCl— No 1. p.,* fairly co-ordinated pulsations in 2 sec., con- tractions ended by tonic condition. n/250 HCl— No l. p., 10 co-ordinated pulsations in 15 sec. n/300 HCl — (1) L. p. 30 sec., go co-ordinated pulsations. (2) L. p. to sec., at the end of the 20th co-ordinated pulsation the solution was diluted with an equal amount of fresh sea water and 150 more fairly co-ordinated pulsations resulted. : n/500 HCl —No stimulation. ' Lores, J.: The dynamics of living matter, 1906, p. 87. ? Litre, R.S.: This journal, 1908, xxii, p. 75. * In all of the following experiments the reagents were added to fresh sea water in such a way that a minimum of change occurred in the normal sea water. For ex- ample, in making up 2/500 HCl, 2 c.c. of u/10 HCl in distilled water were added to 98 c.c. of sea water. The freshly trimmed bells were then gently placed in the solution. * “T,. p.” stands for latent period, being the time elapsing between the placing of the bells in the various solutions and the response by contraction. Production of Rhythnucal Contractions. 119g Experiments with nitric acid. — n/250 HNO, — No Ll. p., 46 inco-ordinated pulsations. n/300 HNO, — Nol. p., 75 fairly co-ordinated pulsations. n/400 HNO, — L. p. 5 sec., 13 co-ordinated pulsations, followed by inco-ordinated pulsations for 14 min. n/450 HNO, —L. p. 20 sec., 200 co-ordinated pulsations in 2 min., inco-ordinated pulsations for 2 min. more. Solution was then diluted to n/600, and 103 good pulsations followed in groups of 20, 45, and 38 with a few seconds intervening between the groups. Faint rhythmic pulsa- tions followed for about 10 min. n/500 HNO, — L. p. 1 min., 29 slow co-ordinated pulsations. n/600 HNO, — No stimulation. Experiments with acetic acid. — n/250 HC,H,O, — L. p. 10 sec., 30 inco-ordinated pulsations. n/300 HC,H,O, — L. p. 1 min., 40 co-ordinated pulsations. n/400 HC,H,O, — L. p. 34 min., 75 co-ordinated pulsations. After 25 more fair pulsations solution was diluted to n/500, and 250 slightly inco-ordinated pulsations resulted. n/4so HC,H,O, — L. p. 2 min., 81 co-ordinated pulsations in 1 min. Solution then diluted to 2/600, and 20 more good pulsations resulted. Three drops of m/ro acetic acid were then added, and pulsations began again. Immediately diluted solution to about m/600, and go more co- ordinated pulsations followed. n/600 HC,H,O, — No stimulation. Experiments with sulphuric acid. — n/too and n/200 H,SO,—No 1. p., 7 to 10 co-ordinated pulsations. Dilution did not prolong the contractions. n/400 H,SO,— L. p. 70 sec., 27 co-ordinated pulsations, followed by a few inco-ordinated pulsations; pause for 1} min., and 105 slightly inco- ordinated pulsations resulted. n/500.H,SO,— L. p. 2} min., 50 co-ordinated pulsations in 35 sec. n/700 H,SO,— No stimulation. Experiments with carbon dioxide..— 1 part CO, charged sea water plus 1 part fresh sea water — L. p. 1} min., 3 fair pulsations. = - rt part CO, charged sea water plus 3 parts fresh sea water — (1) L. p. ri min., 75 co-ordinated pulsations in 45 sec. followed by 75 slightly inco-ordinated pulsations. 5 The carbon dioxide used was generated in a coinmercial portable siphon in sea water; an overcharge resulting. 120 O. P. Terry. (2) L. p. 14 min., 8 co-ordinated pulsations in 6 sec., followed by inco- ordinated pulsations. Then specimen was placed in fresh sea water, and slightly inco-ordinated pulsations followed slowly for 44 min., when 10 co- ordinated pulsations resulted, and then all movements stopped. Twenty- four hours later this same specimen was placed in 1 part CO, charged sea water plus 2 parts fresh sea water. L. p. 2 min., then 30 co-ordinated pul- sations resulted in 20 sec. Solution was then diluted with abundant fresh sea water and more or less rhythmical, co-ordinated pulsations occurred for 7 min. : 1 part CO, charged sea water plus 5 parts fresh sea water — L. p. 20 sec., 30 co-ordinated pulsations. Dilution with equal parts fresh sea water caused 70 co-ordinated pulsations to occur. Lingle ® has shown that strips of the ventricle of the turtle’s heart beat rhythmically for a time in a pure sodium chloride solu- tion; the contractions may be prolonged by the addition of calcium ions or of pure hydrogen peroxide, or by merely allowing pure oxygen to bubble through the solution. Hydrogen peroxide will not initiate beats in strips of turtle’s ventricle in Ringer’s solution. With the idea of prolonging the pulsations initiated by an acid in the marginless bell of Gonionemus, I added to an n/500 HNO, solution (in sea water) a little neutral hydrogen peroxide (“ Dioxy- gen’’) and found that the pulsations were practically uninfluenced, so far as I was able to determine, either in number or strength. It was possible, however, to initiate pulsations with a much weaker strength of acid if a trace of hydrogen peroxide was also added. Experiments with nitric acid plus hydrogen peroxide. — n/500 HNO, in 50 c.c. of sea water plus 1 drop of hydrogen peroxide — L. p. 2 min., 18 co-ordinated pulsations; a short pause, then 5 more co- ordinated contractions. n/2000 HNO, in 50 c.c. of sea water plus 1 drop of hydrogen peroxide— L. p. 1? min., 27 co-ordinated pulsations; pause of 14 min., 35 co- ordinated pulsations; pause of 4 min., 23 co-ordinated pulsations; pause of ? min., 3 co-ordinated pulsations. 10 c.c. CO, charged sea water plus 4o c.c. dilute hydrogen peroxide (x part H,O, and 500 parts fresh sea water) — L. p. 2 min., 11 fair pul- sations only. The effect of neutral hydrogen peroxide was then determined upon the marginless bells when in otherwise fresh sea water. It ® LinctE: This journal, 1900, iv, p. 265; 31902, viii, p. 75. Ee ———— Ee Production of Rhythmical Contractions. 121 was found that it alone caused rhythmical pulsations even in very dilute solution. The pulsations were, if anything, more in number and greater in individual strength than those caused by carbon dioxide or acids alone. Experiments with neutral hydrogen peroxide alone. — I part 3 per cent H,O, in 100 parts of sea water — L. p. 10 sec., 30 co-ordinated pulsations. Removed to fresh sea water and a few more pulsations followed. I part 3 per cent H,O, in 200 parts of sea water — L. p. 1 min., 20 co- ordinated pulsations. I part 3 per cent H,O, in 400 parts of sea water — L. p. 2 min., 103 co- otdinated pulsations. I part 3 per cent H,O, in 500 parts of sea water — L. p. 1 min., 50 co- ordinated pulsations; pause of 30 sec., 40 co-ordinated pulsations; pause of 5 sec., 21 co-ordinated pulsations; pause of 14 min., 10 co-ordinated pulsations; pause of 15 sec., 5 co-ordinated pulsations; pause of 15 sec., 77 co-ordinated pulsations. (A very few tiny bubbles of oxygen gas were observed on some of the bells after about 10 min. immersion.) I part 3 per cent H,O, in 1000 parts of sea water — No response for 30 min. at least. Specimen unobserved then for four hours, at the end of which time 35 slow, weak, rhythmic pulsations were observed. Loeb has said that acids cause pulsations in the marginless bells of Polyorchis. ‘‘ Alkalies have the opposite effect.” This I have shown to be true with Gonionemus, but also that if a trace of hydrogen peroxide is added to alkaline sea water, pulsations (clonic, then tonic contractions) can occur and continue. The amount of alkali added may be as great as n/40’ NaOH. The alkali, how- ever, precipitates the magnesium ions, and in this strength the Mg(OH), is evident as a gelatinous suspension which slowly settles. In this case a number of the hydroxyl ions are lost from solution. The beat in such a solution may be due in part to the removal of the inhibitory action of the magnesium ions. This is indicated, according to Loeb,’ by the clonic, then tonic condition of the contractions; and also by the fact that the marginless bells in sea water made alkaline with NaOH (no H,O, being added) tend to contract slowly into a small hard ball. 7 Logs, J.: This journal, 1902, viii, pp. 81, 91. 122 On Berry: Experiments with sodium hydrate and hydrogen peroxide. — 40 c.c. 2/400 NaOH in fresh sea water plus 2 drops H,O, — L. p. 1} min., 168 co-ordinated pulsations in alternating series of slow and fast contractions; pause of ro sec., then 5 good pulsations; pause of 2 min., then 11 more slow pulsations at intervals of about 2 sec.; pause of 8 min., then 1 pulsation; pause of ro min., then 1 more pulsation. 40 c.c. 2/200 NaOH in fresh sea water plus 2 drops H,O, — L. p. 14 min., 31 co-ordinated pulsations; pause of 5 sec., 41 co-ordinated pulsa- tions; pause of 3 sec., 78 co-ordinated pulsations; pause of 2 sec., I co- ordinated pulsation; pause of 5 min., 6 co-ordinated pulsations. " 4o c.c. n/100 NaOH in fresh sea water plus 2 drops H,O, — L. p. 14 sec., 30 co-ordinated pulsations; pause of 5 sec., 27 co-ordinated pulsa- tions; then followed by 40 co-ordinated pulsations at intervals varying from } to 5 min. 40 c.c. 2/50 NaOH in fresh sea water plus 2 drops H,O, — L. p. 1? min., t co-ordinated pulsation; pause of 5 sec., 54 co-ordinated pulsations ; pause of 3 sec., 70 co-ordinated pulsations; pause of 2 sec., 2 co-ordinated pulsations; pause of 5 sec., 6 co-ordinated pulsations; pause of 3 sec., 13 co-ordinated pulsations; pause of 2 sec., 11 co-ordinated pulsations; pause of 3 sec., 3 co-ordinated pulsations; pause of ro sec., 2 co-ordinated pulsations; pause of ro sec., 2 co-ordinated pulsations; pause of 1 min., 18 co-ordinated pulsations; pause of 2 sec., ro co-ordinated pulsations ; pause of 35 min., 3 co-ordinated pulsations; pause of 14 min., 2 co-ordi- nated pulsations. Benedict ® claims that NaCl exhaustion in strips of turtle’s ven- tricle is due to loss of tonus (which may be possible in the bell of Gonionemus after cutting off the nervous system), and that cal- cium chloride, sodium carbonate, and oxygen increase this tonus and thus initiate the contraction. Alkalies also tend to increase tonus, at least aid recovery after sodium chloride exhaustion. “ It is very interesting to note that not one of these substances (cal- cium chloride, dextrose, lithium chloride, cane sugar, oxygen, air, hydrogen peroxide, sodium carbonate, and Ringer’s solution) is capable of producing beats in a fresh strip.” I have shown above that hydrogen peroxide initiates pulsations in the marginless bell of Gonionemus, even in alkaline solution. Experiments with pure oxygen bubbling through the fresh sea water gave negative results, no pulsations resulting during four hours’ observation. However, lack of time prevented complete $ BENEDICT: This journal, 1908, xxii, p. 16. Production of Rhythmical Contractions. 123 experimentation; it is possible that pulsations may have begun six or eight hours later, as, in one case where hydrogen peroxide was used in minute traces (1 drop to 40 c.c. sea water), pulsations were not observed until five or six hours later. When the bells were placed in sea water to which sufficient sodium hydrate was added to make 7/50 solution (not considering the loss by precipitation of magnesium) and oxygen allowed to bubble through, eight or ten isolated co-ordinated pulsations oc- curred at intervals of one to ten minutes. Experiment with sodium hydrate plus pure oxygen. — n/50 NaOH in sea water through which pure oxygen bubbled for 25 min. — L. p. 16 min., 1 co-ordinated pulsation; pause of 1 min., 1 co- ordinated pulsation; pause of 1 min., 1 co-ordinated pulsation; pause of 2 min., 1 co-ordinated pulsation; pause of 10 min., 1 co-ordinated pulsation. SUMMARY. These experiments indicate that hydrogen peroxide initiates the pulsations in the marginless bell of Gonionemus in normal sea water by increasing the oxidative processes. Oxidations occur much more easily in alkaline than in neutral or acid media, which may explain the pulsations caused by pure oxygen in alkaline solu- tions and not in fresh sea water. Also it is unreasonable to suppose that exactly the same effects are produced by oxygen when in solu- tion in normal amount for sea water and when there is a super- saturation. The stimulating action of hydrogen peroxide and the non-stimulation by oxygen in normal sea water may be due to the fact that oxygen is liberated from the hydrogen peroxide in an atomic condition. The stimulating action of oxygen in alkaline solution is probably due to increased tendency to oxidation. STUDIES IN THE PHYSIOLOGY OF THE CENTER NERVOUS SYSTEM.—TI. THE GENERAL Via NOMENA OF SPINAL SHOCK-+ By PED PERE: [From the Physiological Laboratories of the University of Chicago and of the Harvard Medical School.] HE generally accepted teaching for many years has been that the spinal cord is the great. reflex centre of the icemtcal nervous system. In the lower vertebrates, such as the tortoise, there is little doubt that the spinal cord is a true reflex mechanism, since nothing short of the death of, or severe injury to, its con- stituent neurones will stop such reflexes. In the frog and higher vertebrates any sudden injury to the cord which completely blocks conduction to the brain or to the medulla oblongata, such as crush- ing it or cutting it across transversely, will cause a cessation, more or less temporary, of the reflexes of the skeletal muscle supplied by nerves arising below the level of the injury. “ This phenomenon is shock.” ? Two explanations are open to us at this point: (1) that the reflex mechanism in these animals normally involves other struc- tures in the central nervous system lying above the level of the injury and that the reflexes have stopped because their normal pathway is blocked, or (2) that the reflex mechanism is the same in all vertebrates, despite the profound morphological changes occurring in the central nervous system as we ascend in the phylum, and that the reflexes stop because of some other reason than the first given. The second explanation has been the one generally, but not unanimously,®? accepted, and much time has been given to ' A preliminary note appeared in Science, 1908, N. Ss. xxvili, p. 808. * MarsHALt Hatt: Synopsis of the diastaltic nervous system, London, 1850. The earlier literature is given by EckHArD, Beitrige zu Anatomie und Physiologie, Giessen, 1881, ix, pp. 29-192. * BastrAN: Medico-chirurgical transactions, London, 1890, xxiii, pp. 151-217; ROSENTHAL and MENDELSSOHN: Neurologisches Centralblatt, 1897, xvi, p. 978, and the papers there cited. Also, PANpt, Archiv fiir die gesammte Physiologie, 1895, 1xi, p. 465, and Neurologisches Centralblatt, 1904, xxiii, p. 440. 124 a ee Physiology of the Central Nervous System. 125 the consideration of the nature and cause of spinal shock and to the development of a symmetrical hypothesis. The second explana- tion is widely regarded to-day as a definite settlement of the point. Edinger,* for example, states that all instincts and reflexes are functions of the palzencephalon, and not of the neéncephalon. That the cause of spinal shock.lies in the central nervous system itself and not in the fall of blood pressure produced by spinal tran- section is shown by the experiments of Owsjannikow,® who found that section of the splanchnic nerves did not cause spinal shock, although the blood pressure was greatly lowered; and raising the blood pressure by stimulation of the peripheral end of the splanchnic nerves did not remedy the condition of shock when the spinal cord was transected. It has been pointed out by Sherrington ® that the whole animal is affected alike in the matter of low blood pressure, but the reflexes of the brain and the part of the spinal cord anterior to the lesion are not affected. The immediate effects of the transec- tion are, therefore, not due to the low blood pressure. But any one who has worked with decerebrated animals and has noticed the slowly failing, irregular, or even spasmodic respiration which comes on after removal of a large part of the spinal cord has no doubt been impressed with the remote effects of a very low blood pressure. That the respiratory disturbances are not due to the stimulation of afferent fibres in such cases may be shown by freezing the cord and transecting it while frozen, as will be described later in the paper. Subsequent removal of the cord is followed by the same conditions as before, although there has been no stimulation of afferent fibres. Such remote effects of continued low blood pres- sure are shown by Goltz’s * dogs, which died when the entire lumbar cord was destroyed at one time, even as late as several days after the first transection. Such remote effects have, in my opinion, been too generally overlooked in discussing the nature and causes of surgical shock. Further literature on spinal shock is given by Rosenthal, Loeb, Goltz, Sherrington, and Walton in the various articles to be cited * EpINcER: Journal of comparative neurology and psychology, 1908, xviii, p. 437. 5 OwsJANNIKOoW: Arbeiten aus den physiologischen Anstalt zu Leipzig, 1874, P. 374. 6 SHERRINGTON: SCHAFER’S Text-book of physiology, 1900, li, p. 847. 7 Gortz: Archiv fiir die gesammte Physiologie, 1873, vili, p. 460. 126 Fo Five: in this paper. A more formal discussion of the literature will be given in later papers of the series. In. the series of studies on the central nervous system, I purpose to examine into the question of shock from the point of view of the phylogenetic development of function of this system. The role of the central nervous system in the evolution of the verte- brate phylum has been dealt with very little from the functional side, and most of the interpretations of the adaptation of animals to their environment have been made by morphologists. It is a truism that the one best fitted to interpret’'in terms of function the various adaptations existing in nature is the trained student of function, but the paucity of literature upon this phase of the sub- ject is evidence of the minimal extent to which the physiologist has entered into his heritage. It does not seem possible that the application of the principles of evolution to the functional study of the central nervous system can be much longer deferred. But it is clear that whoever traces the functional evolution of the central nervous system, endeavoring to bring some order into the chaos of literature on the central nervous system, perhaps by establishing functional types that shall render to physiology the same service that structural types have rendered to morphology, must reckon with the idea of shock. And, conversely, the defenders of the idea of shock must reckon, more fully than has been done hitherto, with the facts of organic evolution, morpho- logical and functional. The idea of shock has not yet been exam- ined in its proper biological perspective. In the light of its phylogenetic development we may also in- quire into the two theories of the organization of the central nervous system now current, (1) the segmental theory as developed by Goltz * and Loeb,? and (2) the theory of cortical localization of function. As Goltz has shown, the theory of cortical localization of function cannot be true if his idea of shock, 7. ¢., a long inhibi- tion resulting from the stimulation of efferent fibres, is true, and, although the doctrine of cortical localization of function has not been without defenders, e. g., Hitzig,’° and is even now the domi- nant theory, the hypothesis of spinal shock has passed almost un- ’ Gotz: Archiv fiir die gesammte Physiologie, 1892, li, pp. 570-614. ® Lors: Comparative physiology of the brain, New York, 1900, passim. © Hirzic: Physiologische und klinische Untersuchungen iiber das Gehirn; ge- sammelte Abhandlungen, Berlin, 1904. Physiology of the Central Nervous System. 127 challenged. The predominant tendency has been, not so much to demand a proof of the existence of spinal shock, but to assume its existence and debate about the mechanism and nature of a more or less purely hypothetical entity. I shall show in later papers (1) that any idea of shock is inconsistent with the theory of cerebral local- ization, and (2) that it is possible to develop a theory of the com- parative physiology of the central nervous system without postulat- ing shock, In connection with Professors Stewart and Guthrie,'! I have pub- lished some experiments in which the main phenomena of spinal shock were duplicated by methods which did not involve the an- atomical rupture of any conduction pathways, and with doubtful stimulation of efferent inhibitory pathways. In subsequent experi- ments the phenomena of spinal shock have been duplicated in still greater detail by the method of cerebral anemia, which does not involve the anatomical rupture of any conduction pathways, al- though it totally blocks them physiologically, and by freezing the cord, —a procedure which does not stimulate any efferent inhibi- tory fibres. These experiments are given in detail in the present paper. It may be pointed out here that what is said applies to spinal or experimental shock, as distinguished from surgical shock or “collapse.” In my opinion they are entirely different phenomena. Neither will the general phenomena of so-called protoplasmic shock or “ block” be considered here.'? I wish to express here my obligation to Professor G. N. Stewart in developing the line of experimental work which has led up to my present conception of the central nervous system; to my col- leagues in the University of Chicago for many criticisms and sug- gestions, adverse as well as favorable; to Professor W. B. Cannon for many suggestions and also for the privilege of working in the Harvard laboratory, and to many clinical friends for data on the condition of the reflexes in the human subject in cases of disease of, or injury to, the central nervous system. If progressive changes have occurred in the phylogenetic development of the central nervous system, we might reasonably expect to find certain phenomena of shock more marked in the human than in the monkey — the highest 11 Pree, GUTHRIE, and STEWART: This journal, 1908, xxi, p. 359. 12 SHERRINGTON: SCHAFER’S Text-book, 1900, ii, p. 846. 128 F. AH. Pike. animal type so far studied in the physiological laboratory. Clinical data acquire, therefore, a peculiar interest when considered from the point of view of the evolutionist. THE DoctTRINE OF A LonG INHIBITION AS THE CAUSE OF SPINAL SHOCK. A subject which has so many bearings on the fundamental con- ceptions of the central nervous system as spinal shock may afford the opportunity for a searching examination which, under other circumstances, might be hypercriticism. Accordingly, we may in- quire into (1) the conditions which such a hypothesis must fulfil, and (2) the evidence in favor of such a hypothesis. That the idea of shock, surgical and spinal, has been the excuse for much loose thinking, and that in all probability it has been erroneously held responsible for certain respiratory phenomena, are well shown by Potters As generally expressed, the hypothesis of inhibition postulates a depression of function of the neurones below the level of the transec- tion. It is evident to any one who has worked with the exposed spinal cord that this depression cannot be general. If the brain stem of a dog be divided transversely at the level of the anterior corpora quadrigemina and the anesthetic withdrawn, the exposed spinal cord soon becomes irritable, and a mere touch will cause extensive movements of the muscles supplied by motor nerves orig- inating in that particular region, but it may be impossible to obtain any reflexes of the skeletal muscles. It has been observed !* that stimulation of the peripheral end of the cut pyramidal tract in the monkey will give as great a response, or even a greater, than when these pathways are stimulated with the brain and cord intact. There can be, therefore, no “ motor paralysis”? nor any depression of the neurones in the pyramidal paths. It has been shown that the pyramidal fibres do not end?®° about motor cells in the anterior horn, but about cells in Clarke’s column and probably about other 8 PorteR: Journal of physiology, 1895, xvii, p. 455; Boston medical and sur- gical journal, 1908, clviii, p. 73. “* SHERRINGTON: SCHAFER’s Text-book of physiology, 1900, ii, p. 847. ® ScHAFER: Journal of physiology, 1899, xviv, p. xxxii; von Monakow: Archiv fiir Psychiatrie, 1895, xxvii, pp. 1, 386. See also RepiicH: Neurologisches Central- blatt, 1897, xvi, pp. 818-832. — Physiology of the Central Nervous System. 129 cells in the gray matter. Von Monakow’s view that the fibres of the pyramidal tract do not enter into direct connection with the motor neurones, but act upon the latter through intercalated den- draxones, receives a certain degree of confirmation. Impulses pass- ing down the pyramidal tracts must therefore pass over at least two synapses before reaching the motor neurones. The conduc- tivity of these synapses has not been affected in the least, and there is here no discoverable depression sufficient to account for the failure of the reflexes. Any such effect must have been exerted in some other locality. It would appear to be more than questionable, as Beevor 1® suggests, whether the pathway between the endings of the pyramidal tracts and the motor neurones is exactly the same pathway as that involved in the reflexes, supposing for the moment that the reflex arc lies through the spinal cord alone. If the two pathways were identical, there is no apparent reason why the re- flexes should fail after spinal transection, since the efferent motor path is so obviously open. The inhibitory effect must therefore be exerted either upon some part of the afferent pathway or upon some undiscovered part of the efferent pathway. There is no apparent decrease in conductivity or excitability of the afferent nerves, and the only remaining places where one might look for a break in the arc are (1) the synapse between the afferent fibre of the posterior root and the cell of Clarke’s column, or (2) in an intermediate neu- rone interposed between the posterior root fibre and the cell of Clarke’s column; (3) it is conceivable, perhaps even probable, that the reflex arc might consist of the posterior root fibre, an interme- diate neurone other than the one in Clarke's column and the motor neurone in the anterior horn.1* But, whatever constitutes the reflex arc, the inhibitory effect must be exerted upon the afferent rather than upon the efferent or motor part of the pathway. All reflex arcs are not affected alike by such inhibition. The visceral reflexes are affected scarcely more in man and the monkey by transverse lesions of the cord than they are in the rabbit or the cat,1§ but the return of the reflexes of the skeletal muscles, as will be pointed out in greater detail in a later paper, is far less in the former than in the latter.19 And in the monkey weak electrical or 146 BEEvoR: Journal of the American Medical Association, 1908, li, p. 89. 11 yon LENHOSSEK: Der feinere Bau des Nervensystems, 2 Auf., Leipzig, 1895, p. 405. Cited by SHERRINGTON, SCHAFER’S Text-book, 1900, li, p. 810. 18 SHERRINGTON: SCHAFER’S Text-book, 1900, ii, p. 847. 19 Moore and OErTEL: This journal, 1899, iii, p. 245. 130 FP. HPaike. mechanical stimulation of the central end of the posterior root of a spinal nerve will elicit reflex movements of the skeletal muscles at a time when far stronger stimuli applied to skin or to afferent nerve trunks causes no response.” I have seen reflex movements in cats after transection of the cord on stimulation of the central end of a posterior root (lumbar) at a time when there was no reflex response to stimuli applied to skin or nerve trunks. On high spinal transec- tion or on decerebration the reflexes of the hind limbs return sooner than those of the fore limbs. The hind limbs suffer less than the front on removal of the cerebral motor cortex. The inhibitory effect must be exerted in increasing degree upon the reflex arcs for the skeletal musculature as we ascend in the vertebrate phylum. To bolster up the theory of inhibition, it must be shown that (1) there are more inhibitory fibres in man and monkey than in any other animal, or (2) that these fibres are more easily excitable and pro- duce far greater effects in man than in the lower vertebrates. It is well known also that after the reflexes have returned following transection of the cord, a second transection has no further effect.?? Why these hypothetical efferent fibres should be capable of but one effective stimulation is a matter of considerable interest. Shock is exerted in the aboral direction only. There are many afferent fibres the stimulation of which produces well-known in- hibitory effects. Any theory of inhibition must explain why or how these afferent fibres escape stimulation by a process which is sup- posed to be such a tremendous stimulus to efferent nerves. And, again, it is a well-known fact that the tetanizing current from an induction coil is a far more effective stimulus for most efferent nerves than simple cutting; yet no one has ever been able to pro- duce spinal shock by any sort of stimulation which did not destroy the activity of the neurones. Why the less efficient stimulus should produce so much greater effect when applied to the spinal cord is a point requiring further explanation on any hypothesis of inhibition. A theory of shock must also explain why shock is less severe in young animals. Babak ** has noticed that in larval frogs transection of the spinal cord produces no shock. To interchange premise and conclusion, as Babak does, and explain the absence of shock by *” SHERRINGTON: SCHAFER’S Text-book, 1900, ii, p. 847. *! SHERRINGTON: Integrative action of the nervous system, New York, 1906, p. 246. * BaBak: Zentralblatt fiir Physiologie, 1907, xxi, p. 9. ey Physiology of the Central Nervous System. 131 saying that the inhibitory fibres in the cord have not developed is hardly the most rigid kind of a demonstration. It is necessary to show that some of these fibres are as yet incapable of conductivity or excitability. The whole necessity for any theory of shock lies, as has been pointed out, in the fact that another assumption has already been made, — the assumption that the spinal cord has exactly the same function throughout the vertebrate phylum. It is scarcely necessary to point out here that no independent proof has ever been adduced that the reflexes for the skeletal muscles in higher vertebrates occur through an arc involving the cord alone when the whole central nervous system is intact, and Rosenthal’s. results ** cast doubt upon the validity of such a postulate even in the frog. It may be men- tioned here, too, that the doctrine of spinal shock and spinal re- flexes has proved unsatisfactory from the clinical point of view.?* That certain reflexes — the maintenance of a peculiar state of tonus of all the skeletal muscles — present in normal and in decerebrated frogs permanently fail when the basis of the mid-brain is severed from the medulla oblongata has been observed by Verworn.”° THE EXPERIMENTAL EVIDENCE ON SPINAL SHOCK. It has recently been pointed out °° that it is possible to duplicate many of the phenomena of spinal shock by occlusion of the head arteries. The functions of the brain and medulla oblongata totally fail during the consequent anemia. The reflexes of the skeletal muscles are abolished and reflex vaso-motor effects are no longer obtainable. In experiments where the occlusion period was long and the damage to the cerebral cells consequently great, the scratch reflex was often noticed in the period of recovery. The statements in the literature 2" are that anemia of the brain does not produce 23 ROSENTHAL and MENDELSSOHN: Loc. cit. 4 Warton: Journal of nervous and mental disease, 1902, xxix, p. 337; WALTON and Pau: Jbid., 1906, xxxili, p. 681. 2° Verworn: Archiv fiir die gesammte Physiologie, 1896, Ixv, p. 63. 76 Pike, GUTHRIE, and Stewart: This journal, 1908, xxi, p. 359; Journal of experimental medicine, 1908, x, p. 490. 27 AsHER and Liscuer: Zeitschrift fiir Biologie, 1899, xxxviii, p. 499; ASHER and ARNOLD: Ibid., 1900, xi, p. 271. 132 fo ee oragee: shock in rabbits. Also, removal of the cortex has a smaller effect than section below the pons.?® Section of the brain stem by the knife in cats just above the tentorium, 7.¢., through the anterior corpora quadrigemina, pro- duces no increase of shock over that produced by anzemia of the brain and cervical cord, if practised so late in occlusion or so early in resuscitation that the afferent paths to the respiratory centre are non-conductive, even though the efferent paths from the respiratory centre still remain open.?® This is perhaps an additional argument against the view that mechanical excitation of efferent inhibitory fibres is a factor in spinal shock. When the anatomical section is made at an earlier period in occlu- sion, when the anemia has not yet produced block of the conducting paths, it causes shock no deeper than that due to the anemia. Our first statement, that anemia of the brain and cervical cord did not itself produce shock in the remainder of the cord, was based upon a study of the reflexes after short periods of anzemia. This study did not begin sufficiently early in the resuscitation period, as we had not then appreciated the fact that after a short period of anzmia the shock phenomena disappear rapidly, not, as after shock produced by anatomical section, through the opening of spinal reflex paths, but by restoration of the normal long paths to the brain in the re- suscitation period. The facts really show, however, that our first statement was incorrect. Further proof of this will appear below. If, then, cerebral anzemia produces spinal shock in cats, it should be possible to reproduce, in greater detail than we had previously done, the classical phenomena of spinal shock. The failure of the reflexes should be as complete during cerebral anzemia as after tran- section of the spinal cord. It should be possible, for example, to get back as many of the reflexes of the skeletal muscles when the head arteries were permanently ligated as could be obtained in the same time after transection of the spinal cord. It would, of course, be hopeless to attempt to keep an animal alive by means of artificial respiration for a number of days, and any such comparisons are, of necessity, limited to a few hours. All the animals were deeply etherized before beginning the experiment. A tracheal cannula was then inserted to provide for artificial respiration when it should 8 SHERRINGTON: Integrative action of the nervous system, New York, 1906, p. 246. ty 9 Stewart and Pike: This journal, 1907, xx, p. 61. Physiology of the Central Nervous System. 133 become necessary. After the first cessation of movement in animals whose cerebral arteries had been ligated, the anesthetic was discon- tinued, as total anzemia of the brain is equivalent to decerebration. Similarly, the anzesthetic was discontinued after decerebration. As illustration of these facts, we submit the following protocol, in addition to those previously published: *° Experiment of February 23, 1908.— Adult female cat. Ether. Tracheotomy. Stimulation of the central end of left sciatic nerve before ligation of the head arteries gave a marked increase in blood pressure. 3.25 P.M. Head arteries permanently clamped. 3.38 P.M. Stimulation of the sciatic has no effect. 3.45 P.M. Stimulation of the sciatic has no effect. The mght hind foot is drawn up when pinched. 3.51 P. M. Stimulation of the sciatic causes slight rise in pressure. The reflex contraction of the hind legs is also increased somewhat at this time. From this time to 433 P.M. The effect of stimulation of the sciatic nerve increased grad- ually, the vaso-motor response being slightly greater at successive intervals. The first appearance of the scratch reflex occurred at this time. 4.44 P.M. Stimulation of the left sciatic nerve caused rise in blood pressure. The right hind leg was strongly contracted during this stimula- tion, and kicked vigorously several times when the stimulation was stopped. 4.55 P.M. Spinal cord exposed in mid-dorsal region. Touching it with the point of the knife caused a great rise in blood pressure. 4.59 P. M. Divided spinal cord transversely. Right hind leg drawn up when pinched 30 seconds later. 5.01 P. M. Stimulation of the sciatic causes rise in blood pressure. 5-03 P. M. Scratching ribs with point of forceps causes kicking of hind legs and rise of blood pressure. 5.07 P. M. Result of the stimulation of sciatic was doubtful. Irregular- ities began to appear in the heart rhythm. The beat became slower, caus- ing a greater excursion of the mercury column of the manometer. In two minutes the heart had ceased to beat. Post mortem examination showed that section of the spinal cord_was complete, the knife passing just below the roots of the sixteenth spinal nerve. The fatal effect of section of the spinal cord during the period of anemia of the brain or early in the resuscitation period has been 8 PIKE, GUTHRIE, and STEWART: This journal, 1908, xxi, p. 359. 134 Flake. noticed previously,*' but the discussion of the significance of these results cannot be given here. It will be noticed (1) that all the reflexes fail at a certain time, more or less variable according to the individual peculiarities of the animals; *? (2) that there is a gradual return of reflex contrac- tion of the hind legs on pinching the foot, the contraction occurring (a): on the same side and (0) involving the other side later. The return of the homolateral and crossed reflexes occurs here in the same order as we have observed in the fore limbs on reestablishment of the cerebral circulation; ** (3) irregularities resembling Traube Hering curves appear in the blood-pressure tracing, and soon a reflex rise of pressure appears in response to stimulation of the central end of the sciatic. Whether this is a true vaso-motor reflex . in all cases is questionable. In the curarized cat, mentioned later in the paper, true vaso-motor reflexes were not obtained as early after occlusion of the head arteries as such blood-pressure changes usu- ally appear in non-curarized animals. (4) The movements of the hind legs become increasingly facile and the scratch reflex may be evoked by stroking the ribs. (5) The whole posterior part of the animal becomes increasingly irritable, so that a light touch may cause kicking of the hind legs, movements of the tail, and contrac- tion of the abdominal and lower intercostal muscles. The return of the reflexes is often somewhat slower when the cerebral circula- tion is stopped than it is after transection of the cord. Occasionally apparently contradictory results are obtained. The reflexes of the skeletal muscles may fail for a time, although slow, convulsive respiratory movements occur and the blood pressure remains well up, possibly higher than normal. Stimulation of the central end of the sciatic nerve causes an unusually great change in the blood pressure, although it may evoke no reflex response of the hind limbs. Such a condition is shown in the — Experiment of May 14, 1908.—Dog, nearly or quite grown. Ether. Tra- cheotomy. Blood pressure from cannula in left carotid. Central end of left sciatic and right vagus nerves prepared for stimulation. 2.33 P.M. Stimulated sciatic nerve; rise of blood pressure. 37 P.M. Stimulated right vagus nerve. Rise of blood pressure. *! Stewart ef al.: Journal of experimental medicine, 1906, viii, p. 311; This journal, 1907, xx, p. 71. *° Stewart et al.: Ibid., viii, pp. 289-321. 38 Stewart et al.: Loc. cit. Phystology of the Central Nervous System. 135 2.39 Pp. M. Tied off right carotid and right subclavian arteries. 2.43 P.M. ‘Tied off left subclavian artery. 2.45 P.M. Started artificial respiration. 2.54 P.M. Stimulated the central end of vagus. Marked rise of blood pressure. 2.55 P.M. Stimulated sciatic nerve. Marked rise of blood pressure. Very faint reflex in right hind leg. 2.57 P.M. Front legs perfectly limp. No reflexes. 3.00 P. M. Stimulated sciatic. Marked rise of blood pressure. The front leg perfectly limp, while the left hind leg is rigid, 7. e., extensor mus- cles are tonically contracted. 3.03 P. M. Stimulated vagus nerve. Rise of blood pressure. Corneal reflex still present. 3.05 P.M. Thefront limbs are becoming rigid, 7. e., tonically contracted. 3.06 P.M. Both fore limbs rigid; the right drawn up and the left extended. 3.10 P.M. Draws up left front foot when pinched, and pushes it out again when stimulation ceases. 3.46 Pp. M. Stimulated sciatic nerve. Some rise in blood pressure. 3.49 P. M. Cut spinal cord across. 3.50 P.M. Reflex in left hind leg as strong as ever. 3.51 P. M. Stimulated sciatic. Possibly a slight rise of blood pressure. 4.30 P.M. Stimulated sciatic. Slight rise of pressure, then a fall, with return to normal before stimulation ceased. Hind limb reflexes very active by pinching. Wags tail after stimulating sciatic. 4.41 P.M. Stimulated sciatic. Slight rise of pressure. Dog raised tail up and held it rigid for a moment. No corneal reflex at this time. 4.45 P.M. Stimulated vagus. Slight rise of pressure with return to normal. Hind limbs very active. 4.55 P. M. Cut spinal cord across the second time. Blood pressure rose to twice its usual height and then fell. 4 4.56 P.M. Shook tail violently when foot was pinched. 5.00 P.M. Stimulated sciatic. Slight fall of blood pressure with re- turn to normal. The first section of the spinal cord was at the level of the third or fourth dorsal segment. The second section was about two seg- ments lower. The attempt was made to eliminate all the bulbar and cephalic centres by anzemia, as we had done in cats. Instead of this a dif- ferential effect probably resulted. The limpness of the fore limbs ' after tying the head arteries could not have been due to any failure 136 F. H. Pike. of function in the upper part of the spinal cord, since the respira- tory and vaso-motor centres, as shown by stimulation of the vagus, were active. Nor could it have been due to the effect of ether, since the hind legs were rigid. It must have been due, therefore, to the paralysis by anemia of the cerebral cortex and possibly also the subcortical and basal ganglia as low as the posterior corpora quadrigemina, or to the anzemia of the fore limbs themselves. That the posterior corpora quadrigemina were not affected was shown by the persistence of the corneal reflex. Some shock might have been expected on section of the cord, since the reflex arcs might have been established above this level. The effect on the reflexes of the skeletal musculature was, however, very slight. The effect of the second transection was inconsequential. The respiration became slow and labored immediately after the first transection and soon ceased. Every part of the encephalon anterior to the pons, or, at most, to the posterior corpora quadrigemina, as indicated by the wide pupils (due to failure of third nerve centre) and final absence of the corneal reflex, had been rendered inactive by the anzemia, and the in- activity of the anterior part of the encephalon was sufficient to bring about the failure of the reflexes of the skeletal muscles. Such cases are not very rare. The results are very similar to those following anatomical transection of the brain through the posterior corpora quadrigemina. Later, the respiration and vaso-motor tone may also fail. These cases, as stated before,** are probably due to individual differences in the anastomotic channels to the brain and medulla oblongata. More rarely, the blood pressure, nearly normal or occasionally higher than normal, and respiration may persist throughout the occlusion period. If, at the end of thirty or forty minutes, when the muscular reflexes have returned in some degree, the spinal cord be transected in the upper or mid-dorsal region, the skeletal re- flexes are unaffected, while vaso-motor reflexes are abolished. As an illustration of this point, we may cite the — Experiment of May 12, 1908.— Young cat, nearly grown. Ether. ‘Tra- cheotomy. Blood pressure from left carotid. Central end of right sciatic nerve prepared for stimulation. Central stimulation of the sciatic before ligation of the cerebral arteries caused rise in the blood pressure. 34 Stewart ef al.: Loe. cit. Physiology of the Central Nervous System. 137 11.06 A.M. Ligated head arteries in the usual way. 11.27 A.M. Blood pressure low, but respiration still continues. Arti- ficial respiration kept up all the time. 11.28 A.M. Touched right hind leg. Blood pressure rose very high, and animal struggled. Pupils are widely dilated, with furrowed and sunken cornea. No corneal reflex on ordinary stimulation, but eyelids tremble when strong pressure is made on eyeball near lids. 12-12.30 Pp. M. The spinal cord, which had been exposed, was now divided transversely at level of third or fourth dorsal segment. 12.15 P.M. Stimulation of the central end of the left sciatic, which had been freshly prepared, caused a very slight rise in blood pressure. Stimulation with a slightly weaker current just before section of the cord caused a marked rise in pressure. The reflexes of the hind legs were unaffected by transection of the cord. 12.23.30 P. M. Artificial respiration stopped. 12.25.30 P. M. Asphyxial convulsions. Blood pressure, which rose at first, has now fallen again. Micturition. 12.30.30 P. M. Hind legs straightened out spasmodically. Experiment stopped. This condition corresponds, as nearly as can be expected from the diversity of the methods employed and the inability to control the degree of anemia in various parts of the brain, to successive tran- sections of the neural axis at (1) a high and (2) afterward at a low level. Very rarely, in fact only once in the entire series of ex- periments on cerebral anemia begun nearly four years ago, have we found an animal in which the vaso-motor response to stimulation of the sciatic was not completely abolished for a time either by the anemia or by subsequent anatomical transection of the spinal cord. In this animal stimulation of the central end of the sciatic during the period of anzemia produced some change, although at times a relatively small one, in blood pressure each time it was stimulated. And furthermore, transection of the spinal cord in the dorsal region did not affect these vaso-motor reflexes. It is worthy of remark that this was in a young animal. Although the phenomena were perfectly definite in this case, attempts to verify the facts on other young cats have so far failed. It is possible, therefore, to reproduce with great exactness the phenomena of spinal shock observed after transection of the neural axis at various levels, by methods which do not involve the anatomi- cal rupture of any conduction pathway. 138 F. H. Pike. The long-continued stimulation of efferent inhibitory pathways by anzemia is an improbable occurrence. It may be shown that anaemia does not stimulate efferent motor fibres. As has been pointed out in a previous paper,*? the upper part of the phrenic nerve becomes anemic during occlusion of the head arteries, and, while its excit- ability may increase, I have never seen a case of spasm of the diaphragm or a twitching of its muscular fibres due to the anzemia of the phrenic nerve. Again, after the cessation of the first con- vulsions immediately following the occlusion of the head arteries, the hind limbs relax for a considerable time, usually twenty to thirty minutes or even longer. But during this time there is a transi- tion area of the cord which is receiving but little blood intervening between that portion which is receiving a full blood supply and that part which receives no blood at all. Somewhere in this transition area there would be excitation if anemia, either partial or complete, excites nerves. But, as we have seen, there is no excitation of motor fibres. If the anemia excites efferent inhibitory fibres alone, we have a method of rare exactness in physiology. Such stimulation of efferent inhibitory fibres is, however, extremely improbable, since anzmia of the upper part of the vagus does not cause any noticeable inhibition of the heart after the first rapid fluctuations are over.*® Asphyxia alone does not cause prolonged spinal shock. An animal may be etherized and afterward asphyxiated until the heart ceases to beat. When the heart is again started and the blood oxygenated by artificial respiration, the animal rapidly recovers, and the reflexes quickly return in all their former vigor. I shall show, in another paper, that asphyxia temporarily abolishes vaso-motor as well as muscular reflexes, and that its general effects are closely similar to those of anzemia. If the hypothesis of a long inhibition is true, and if anzemia really does cause spinal shock by stimulation of the efferent inhibitory pathways, such a stimulation should produce shock once for all, and the restoration of the cerebral pathways should have no influence on the supposed spinal reflexes. But such is not the case. We *? have already quoted, in considerable detail, experiments in which there was a relatively rapid and extremely complete return of all the reflexes after restoration of the cerebral circulation following 8° STEWART and PIKE: This journal, 1907, xix, p. 339. 36 Stewart and Pike: Ibid., xix, p. 349. 37 Stewart et al.: Journal of experimental medicine, Joc. cit. Physiology of the Central Nervous System. 139 rather short periods of anemia. There could have been, therefore, no permanent effects of the stimulation of such efferent inhibitory fibres. It might be supposed that any stimulation of efferent in- hibitory pathways due to anemia would cease upon the reéstablish- ment of the cerebral circulation and the spinal reflex arcs again be- come passable. There seems little reason for such a supposition, however, as the convulsions, described in previous papers, occurring during the resuscitation period are, in all probability, due to central stimulation of motor neurones. That anemia should stimulate efferent inhibitory fibres only, and that the blood, upon its return to the temporarily anzemic cells, should stimulate motor neurones only, is too fanciful a conception to be entertained. Any stimulation of efferent inhibitory fibres during the resuscitation period must cer- tainly be overbalanced by the stimulation of motor fibres. It is note- worthy, also, in this connection that the strychnine-like effect, pre- viously described, obtaining apparently throughout the spinal cord, so long as it is anatomically intact, is seen when only the brain and a few upper segments of the spinal cord have been subjected to anemia. ‘These results are incompatible with the hypothesis of a long inhibition or depression of function by stimulation of efferent fibres, but they may in some degree be harmonized, as will be shown later, with the idea of tonus changes in spinal shock. But granting for the moment that anzemia does stimulate efferent inhibitory pathways, can we produce spinal shock by any other method which does not necessarily involve the anatomical rupture of the conduction pathways or the stimulation of inhibitory fibres ? To answer this question, I have frozen the spinal cords of rabbits, cats, and dogs by means of an ethyl-chloride spray or by pouring liquid air directly on to the cord. The first data known to me on this subject are from an unpublished experiment by Professor D. J. Lingle. The brain of a frog was frozen by ethyl chloride. The reflexes ceased just as they did in a decerebrated frog. In my own experiments the animals were etherized and the spinal cord exposed in the upper dorsal region. If ethyl-chloride was used for freezing, the cord was, as a rule, gently lifted up and a sheet of rubber tissue passed under it, the nerve roots on each side of one segment being cut to facilitate the passage of the rubber. If the cord lay down in the spinal canal surrounded by the warm blood which oozed out from the vessels of the vertebral column, some difficulty was ex- perienced in freezing it through with the ethyl-chloride spray. Im- 140 F. A. Prke. mediately before the spray was applied to the cord, the vaso-motor response was tested by electrical stimulation of the central end of one sciatic nerve. This often sufficed for the reflexes of the skeletal muscles also, but both crossed and homolateral reflexes were usually tested by pinching the foot or the tail or some area of the skin. It was shown, by this means, that the division of the dorsal nerve roots had no appreciable effect on the reflexes tested. When liquid air was used for freezing, the spinal cord was left undisturbed, the dura mater being opened. No difference in the effects could be ob- served because of the different manner of handling the cord. Section of the nerve roots in the dorsal region, as already stated, had no noticeable effect upon either the crossed or the homolateral reflexes of the hind legs. That section of the posterior root may affect the tone of muscles connected with that segment ** is well known. It was in order to meet the objection which might be raised against the experiments that a number of the cords were left un- disturbed while freezing. When thoroughly frozen, the cords were usually divided. Their consistency at this time resembled wood or soft metal. The phenomena attendant upon freezing the cord were in every case entirely comparable to those following cerebral anemia or anatomical transection without freezing. The blood pressure, as measured by a manometer connected with a cannula in the carotid, fell more or less rapidly, depending upon the rapidity with which the cord was frozen. When liquid air was employed, the pressure fell almost perpendicularly (Fig. 1). With ethyl chloride a gradual fall might-occur during most of the ten or fifteen minutes occupied in freezing the cord. The reflexes of the skeletal muscles failed as completely as after anatomical transection or after cerebral anzmia. In illustration of these facts we may cite the protocols of two ex- periments on the rabbit and on the cat, respectively : Experiment of February 25.— Adult rabbit. Ether. Tracheotomy. 10.45 A.M. ‘The spinal cord was exposed in mid-dorsal region. The central end of the left sciatic nerve was stimulated to test reflexes of the hind legs. Crossed reflex present. 10.55 A.M. Began freezing cord with ethyl-chloride spray. °8 yon Cyon: Berichte der kéniglichen sachsischen Gesellschaft der Wissen- schaft, Leipzig, 1865, reprinted in ‘‘Gesammelte physiologische Arbeiten,” Berlin, 1888, pp. 197-202; WARRINGTON: Journal of physiology, 1898, xxxiii, p. 112. Physiology of the Central Nervous System. I4I 11.05 A.M. Cord now frozen through, and divided transversely while frozen. There were no movements of any part of the animal while the cord was being frozen, nor when it was divided. No crossed reflex after freezing, but there was a slight contraction of the leg on the same side when right hind foot was pinched. 11.15 A.M. Fairly good reflex on same side when right hind foot is pinched or on stimulating the central end of the sciatic nerve, with the suggestion of a crossed reflex. 11.20 A.M. Good reflex on same side, and fairly strong crossed reflex. 11.35 A.M. Good reflex contraction of both hind limbs on stimulating central end of left sciatic nerve. 11.42 A.M. Strong reflexes of both hind limbs. Experiment stopped. Section of spinal cord complete below fifteenth spinal nerve. I 2 3 Figure 1.— Dog. April 9. Showing fall of blood pressure and absence of vaso-motor reflexes. Spinal cord frozen with liquid air at 1. Subsequent stimulation of the central end of the sciatic nerve at 2 and 3 one and one-half and three minutes respec- tively after freezing the cord was without effect. Four fifths the original size. Experiment of February 25, 1908.— Adult cat. Ether. Tracheotomy. Spinal cord exposed in mid-dorsal region; two spinal roots cut on each side, so that a piece of sheet rubber could be drawn under it, separating it from the blood in the spinal canal. Blood pressure from cannula in left carotid. Central end of left sciatic nerve prepared for stimulation. Effect of stimulation on blood pressure was slight. The reflex contraction of the hind legs was good immediatély before starting to freeze the cord. 3.48 Pp. M. Ethyl-chloride spray turned on cord. 4.07.30 P. M. The cord had been cut about half through. Spray main- tained continuously. No movements of hind limb or tail during the time the spray was playing. The cord was now completely divided. 4.08 P.M. Ethyl-chloride spray stopped. 142 F. H. Ptke. 4.16 P.M. Stimulated sciatic nerve. No effect on blood pressure and no reflexes of the hind limbs. The right hind foot was drawn up slightly on pinching. 4.27 P.M. Stimulated sciatic. Slight rise in blood pressure. Good reflex contraction of right hind leg on pinching foot. No crossed reflex. 4.55 P.M. Stimulation of sciatic. Some fall of blood pressure. 5.07 P.M. Reflex contraction of right hind leg now much stronger. 5.20 P.M. Respiration stopped, possibly from ether, although little was given. Artificial respiration was begun. Natural respiration soon returned. 5.30 P. M. Scratching ribs on left side now causes movements of right hind foot. 5.35 P.M. Stimulated sciatic. Got crossed reflex of right hind leg. 5.37 P. M. Cut cord below first section. Rise of blood pressure. Struggles of hind limbs. Reflex contraction of right hind foot quite as strong as before. 5.43 P.M. Crossed reflex of right hind leg on stimulating left sciatic. 5.48 p. M. Reflex in right hind foot now shows excellent clonus. Ether, which had been given for a few minutes, was now stopped, and reflexes soon improved. 5.51 P.M. Clonus of right hind foot quite as strong as ever. Experiment stopped. We may cite in further illustration the — Experiment of April 9, 1908. — Dog of 15 kilos. Ether. Tracheotomy. Can- nula in left carotid artery. Spinal cord exposed in upper dorsal region. Central end of left sciatic nerve prepared for stimulation. 5.21 P.M. Stimulated sciatic. Rise of blood pressure. 5.25 P.M. Froze cord (about level of first or second dorsal) with liquid air. No movements of any kind occurred. Blood pressure fell rapidly (Fig. 1). 5.26.30 P. M. Stimulated sciatic. No effect. 5.28 p.M. Stimulated sciatic. No effect. The anesthetic is an objectionable and confusing feature in all experiments in which the cord is transected below the medulla ob- longata. Attempts have been made to eliminate these factors from the problem, but without good results. For exampie, if the cord is frozen in the upper dorsal region and divided while frozen, varia- tions in the degree of anzesthesia cause corresponding variations in blood pressure and in the reflex excitability of the part of the spinal Physiology of the Central Nervous System. 143 cord below the level of transection. It is evident that decerebration of the animal at this time can have no direct effect on the part of the spinal cord below the level of the first transection. Decerebra- tion, when the spinal cord is intact, has a very slight immediate influence on blood pressure, vaso-motor reflexes, or respiration. Freezing the spinal cord does not stimulate any efferent inhibitory fibres, but when decerebration is done after previously freezing the cord and dividing it while frozen, the respiration soon becomes ir- regular and often ceases. Artificial respiration greatly prolongs life, but the hemorrhage incident to decerebration added to the fall of blood pressure following division of the spinal cord is incompatible with high reflex excitability for long periods of time. In another series of experiments undertaken for an entirely different purpose, ligation of both carotid and both vertebral arteries in the dog some- times caused total insensibility for several hours without greatly affecting blood pressure or respiration. More often it produced mere drowsiness without insensibility. Somewhat similar results are shown in the protocol of the experiment of May 14, as quoted above. The method is therefore very inconstant in its results, but would seem to promise well for the few cases in which insensibility is produced. I have not so far had a really successful combination of the two methods (1) of producing insensibility by ligation of the four cerebral arteries and (2) of freezing the spinal cord. Some method of anzesthesia, however, is necessary when the blood-pressure changes caused by the stimulation of the sciatic nerve and of the vagus are to be compared. In some of the following experiments on the effect of the second transection of the cord, anzesthesia was maintained by ether, and in others the animals were decerebrated soon after the introduction of the tracheal cannula. When the animal is decerebrated, the knife passing through the anterior corpora quadrigemina, the vaso-motor reflexes, as already stated, are not depressed, but may even be increased. The respira- tion continues unaffected unless the division is made so far back as to involve the part of the mechanism situated in the anterior or posterior corpora quadrigemina. The “ shock” of the skeletal mus- cular reflexes is, however, severe. Early in the recovery, what is probably decerebrate rigidity makes its appearance. The fore limbs are rigidly extended and the jaws are tightly set. In cats the rigidity is less noticeable in the hind legs, and reflex drawing up of the feet on pinching makes its appearance here earlier than in the 144 Fda: fore limbs. A second transection of the neural axis, e. g., in the lower cervical or upper dorsal region of the spinal cord, may-affect the skeletal reflexes in some small degree, but never, in my experi- ence, has it completely abolished them. The vaso-motor reflexes, on the contrary, are completely abolished below the level of the second transection and all respiratory movements, except those of muscles whose nerves arise above the level of the second section, cease per- manently. After a time the vaso-motor reflexes return. Further transections, provided they are not too far removed from the level of first transection of the cord, are without effect. But the combina- tion of the two operations — decerebration and transection of the spinal cord — seems, as already stated, to have a greater effect than either one alone. These facts are illustrated in the protocol of the experiment of March 14, 1908. Experiment of March 14, 1908. — Young adult cat. Ether. Tracheotomy. Blood pressure from left carotid artery. Respiratory tracing from tam- bour connected with tracheal cannula. Skull trephined and calvarium removed. The cerebral hemispheres were removed rapidly, but an at- tempt was made to spare the basal ganglia. The object was to remove the cerebrum which, it might be supposed, would not produce shock but would destroy sensibility, allow the animal to recover from the effects of the anesthetic, and then see whether the corneal reflex and swallowing would be affected by destruction of the basal ganglia. 10.05 A.M. ‘Trephined skull. 10.15 A.M. Removed cerebrum. No further anesthetic. 10.21 A.M. No reflexes in hind limbs, but strong corneal reflex present. “10.26 A.M. On pinching tail get slight movement at root. 10.32 A.M. Corneal reflex strong. 10.35 A.M. Put cotton in mouth; cat apparently swallowed. 10.50 A.M. Crossed reflex present in hind limbs. Slight movement of tail. 10.54 A. M. Movement of fore limb by pinching. 11.05 A.M. ‘Twisting the hind legs is more effective in producing re- flex movements than pinching the foot. Same also for the fore legs, but the hind-limb reflexes are more marked than the fore-limb reflexes. Strongly pinching the tail causes the head to be raised, drawn back, and rotated to the right. 11.10 A.M. The left ear is moved without obvious stimulation. There are a few small blood clots on the tip. The eyes and vibrisse are moved when the ear is pinched. Pinching the nose also causes movements of the Physiology of the Central Nervous System. 145 vibrissze, but they do not move much when they are pulled. The eye and the ear reflexes are much more active than the hind-limb reflexes. 11.28 A.M. Blowing in ear causes movement of it. On striking the board, the animal moves. 11.35 A.M. Pinched tail. Cat raised head. General movements of body, followed by deep respiratory gasp. 11.40 A.M. Pinched fore foot; cat raised head and body. Movements were such as a cat would make in attempting to rise. 11.43 A.M. Same observation as above on pinching hind legs. 11.49 A.M. Scratching side behind fore leg causes much stronger res- piratory movements. 11.55 A.M. On gently tapping hind foot the cat made considerable effort to raise itself on to its feet by drawing hind feet off the board, raising its head, and pulling its fore feet forward. 12.12 P.M. Eyelids closed. Nictitating membrane well drawn over. Eyes looking well forward. 12.15 P.M. Cut spinal cord. Reflexes of skeletal muscles unaffected. 12.29 Pp. M. No reflexes could be elicited in fore or hind limbs, nor was there any corneal reflex. Post mortem. Cord cut below fourteenth spinal nerve. Section complete. Head put into 4 per cent formalin and hardened in situ. The piece of cotton could not be found at post mortem. The stomach was so full of meat that the small piece of cotton might easily have been overlooked. The cesophagus was opened from cardia to pharynx, but no trace of cotton was seen. April 8. The examination of hardened brain showed that the posterior and ventral portions of both occipital lobes were present, but more was present on the left side than on the right. Both anterior corpora quadrigemina intact, but the knife passed close to anterior portion of right corpus. A small portion of the optic thalamus was left, particularly on the left side, where a part of it projected out 4 mm. beyond the corpora mamillaria. From the profound shock effects produced by decerebration it might be thought that the basal ganglia had been injured. There must have been more of the brain left, however, than in the ordinary experiments, since raising the head had never been seen before. Although the second tran- section produced no immediate effects except the fall of blood pressure, the reflexes soon stopped, both anterior and posterior to the section, and the respiration became slow and labored. It is a question whether the failure of the reflexes and the respiration were an immediate or an indirect effect of the second transection. 146 F) A. Prke: Although the basal ganglia were not subsequently destroyed in the above experiment, transection through the anterior corpora quadrigemina was done in a similar experiment as soon as the corneal reflex returned after removal of the cortex and the with- drawal of the anesthetic. The second transection (through the anterior corpora quadrigemina) had no effect on the corneal reflex. Strychnine spasms, as has been pointed out previously,?? may occur at any time during the occlusion period. It has been shown, also, that strychnine spasms will occur immediately after transection of the spinal cord in dogs. What may be mistaken for vaso-motor reflexes are also easily elicited at this time, since stimulation of the central end of the sciatic nerve will often cause a great rise in blood pressure. That this rise is due to the contraction of striated muscle is easily shown. No rise of blood pressure occurs in a strych- ninized and curarized animal at this time. Strychnine does not, therefore, have the same effect upon the spinal vaso-motor reactions that it has upon the reflex mechanism for the skeletal muscles. Al- though detailed consideration of this point cannot well be entered upon without taking into account certain other experimental data on . the vaso-motor mechanism, best presented in a separate paper, the following condensed protocol will be given here for the sake of completeness : ‘ Experiment of August 7, 1908. — Three-fourths grown cat. Ether. Tra- cheotomy. Blood pressure from cannula in left carotid. Central end of left sciatic nerve prepared for stimulation. 3.13 P.M. Head arteries ligated in the usual way. Ether then discon- tinued as soon as reflexes ceased, as cerebral anemia is equivalent to decerebration. 3.16 Pp. M. Injection of strychnine (about 1/10 grain) subcutaneously. 3.24 P.M. More strychnine injected. 3.28 P.M. First strychnine convulsion appears. Stimulation of sci- atic causes good rise of blood pressure, with extremely violent spasms. 3.30.30 P. M. About to c.c. of a 1 per cent solution of curare injected subcutaneously. Stimulation of sciatic at intervals from 3.40 to 4.10 after the spasms of the skeletal muscles had ceased, caused no rise or only an extremely slight rise of blood pressure. 39 STEWART ef al.: Journal of experimental medicine, 1906, viii, p. 289. Physiology of the Central Nervous System. 147 DISCUSSION OF RESULTS. A detailed discussion of spinal shock would, at this stage, be quite premature, but a brief réswmé does not seem out of place here. As is well known, anemia of the lower dorsal or lumbar cord alone does not produce permanent “ shock.’ Spronck #° and others have found that no permanent loss of reflexes and no permanent lesions of the spinal cord follow temporary ligation of the abdominal aorta if the period of anzmia has not been so prolonged ‘as to cause death of the cells. Temporary anemia of the upper part of the spinal cord and the brain will cause temporary “ shock ”’ every- where below the region of anemia. Freezing the spinal cord or brain causes shock below the frozen region. All the phenomena of spinal shock can be reproduced by interruption of the long con- duction pathways of the spinal cord, regardless of any stimulation, or lack of stimulation, of efferent inhibitory pathways, if the time the animals are kept under observation is approximately the same after operation. Inhibition, therefore, has no important share in the production of spinal shock. The only essential factor in the pro- duction of spinal shock is the rupture of the long conduction path- ways. Two possible interpretations are open to us: (1) It is con- ceivable that impulses from above are necessary to maintain the conductivity of the synapse between the afferent and efferent path- ways concerned in the reflex arc. When the efferent fibres from the cerebrum, or whatever other portions of the brain which may be considered to give rise to them, are ruptured, or when their con- ductivity is blocked, the reflex arc becomes non-conductive. Under changed conditions it may regain more or less completely the primi- tive properties which it possesses in such animals as the turtle, but which it has lost in the gradual evolution of the vertebrate phylum. This idea enables us to retain the primitive reflex mechanism in the spinal cord, and to account in a general way for the increasing severity of spinal shock as we ascend in the vertebrate phylum. Von Cyon *! observed that a skeletal muscle lost part of its tonus when the posterior root of its spinal nerve was divided. Loeb *” suggests that this may throw some light on the question of spinal 49 SproNnckK: Archives de physiologie, 1888, p. 1. “ von Cyon: Loc. cit. * Lors: Comparative physiology of the brain, 1900, p. 274. 148 EOE dike: shock. This idea is more or less prevalent under various guises, such as maintenance of tonus. The form which I have given is that suggested by Professor W. B. Cannon. (2) Another possible explanation is that, in the morphological development of the nervous system as outlined by Herrick,** there may have been a concomitant shifting of function, and that the functional reflex arcs in higher animals pass through some part of the central nervous system above the spinal-cord. The primitive pathways through the spinal cord persist morphologically, and may, under changed conditions, regain a part of their primitive function. I have previously suggested such a contingency in connection with the possible spinal origin of accelerator impulses to the heart.** The possible effect of phylogenetic changes in modifying the reflex activities of the spinal cord has also been recognized by Professor Stewart.*° In discussing the course taken by impulses leading to reflexes it is suggested that the question of the exact pathway is not so important as the question “ whether, as a matter of fact, the spinal motor cells are most easily discharged by the im- pulses that reach them directly, or by the impulses that come down by the roundabout way of the cortex and the efferent fibres that connect it with the cord. It is evident that the answer to this ques- tion need not be the same for all kinds of animals. It may well be that in the higher animals, in which the cortex has undergone a relatively great development, the spinal motor mechanisms are more easily discharged from above than from below, while in lower ani- mals the opposite may be the case.’’ As I shall show later, certain of these primitive arcs probably persist in all animals. The higher the type of animal, or, in other words, the more the central nervous system has departed from the primitive segmental type, the less complete will be the recovery from injury to, or disease of, the long conduction pathways, on this second hypothesis. We may, on this basis, readily explain the “ deficiency phenomena ”’ which Sherrington *® describes in the monkey, without postulating any mysterious ‘““ Hemmungserscheinungen.” And, as we would ex- * Herrick: Journal of comparative neurology and psychology, 1908, xviii, p. 393. ** Prxe, GuTHRIE, and STEWART: Journal of experimental medicine, 1908, x, Pp. 494. * Srewart: Manual of physiology, rgoo, 4th ed., p. 706: 5th ed., 1906, p. 706. *© SHERRINGTON: Philosophical transactions of the Royal Society, 1897, cxc, pp. 139-141. « Physiology of the Central Nervous System. 149 pect on the hypothesis of a progressive change in the reflex path- ways, the “ multiformity and complexity of reflexes ” in the monkey, “feeble and poverty-stricken as it is,’ when compared with the reflexes found in the cat or the dog on recovery from spinal transec- tion, is still affluence when compared with the reflexes obtainable in the human after total transverse lesions. It makes little difference, from the theoretical point of view, whether we say, with Bastian 4% and others, that no reflexes return in the human subject after total transverse lesions of the spinal cord, or accept Senator’s ** verdict that, while it is proved beyond question that the tendon reflexes of the lower extremities may fail completely after lesions of the upper part of the spinal cord, even when the reflex arc is not demonstrably affected, such a total failure is probably not the absolute rule. On a priori grounds, one would expect a certain amount of variation here, as in other biological phenomena. Such questions are of in- terest to the clinician, but, whatever be his ultimate decision, the fact is plain that the effects of spinal transection are more severe in the human than in the monkey, and vastly more severe than in such mammalian types as the dog or cat. It is difficult for me to believe that a reflex arc, with no demonstrable lesions, should re- main non-conductive for eleven years,*® lacking only the tonus im- pulses from above to facilitate the passage of the afferent impulse, if it had been fully conductive up to the time of the accident to the spinal cord. The loss of certain afferent impulses, such as result from the division of the posterior spinal roots, for example, leads to chromatolysis and other degenerative changes in the motor cells of the cord,®® but the loss of impulses from above does not, as Senator concedes, necessarily produce any such permanent effect in any neurones which might be regarded as forming a part of the spinal reflex arc, although Sherrington *! had, the year previous to the publication of Senator’s paper, put forward his hypothesis of “isolation dystrophy ” to account for the permanent failure of cer- tain spinal reflexes in the monkey after spinal transection. It may be objected that a sufficient amount of time does not elapse between the transection of the cord and the reappearance of the 41 Bastian: Loc. cit. 48 Senator: Zeitschrift fiir klinische Medicin, 1898, xxxv, p. 18. 4° Bow py: British medical journal, 1899, i, p. 1132. 5 WARRINGTON: Loc. cit. 51 SHERRINGTON: Philosophical transactions of the Royal Society, 1897, Joc. cit. 150 PF eee: reflexes to permit of any change in the nature of the neurones con- stituting the spinal reflex arc. In reply, we may cite a single in- stance drawn from'embryology. ‘The accessory optic vesicles of the chick arise, reach their maximum development, and disappear within a space of three hours.°? There is little reason for thinking that whatever functional change is necessary in the intermediate neurone in the spinal reflex arc should require a vastly greater time for its consummation, particularly in animals somewhere near the chick in the taxonomic scale, unless the neurone in question has so far de- parted from the primitive type that a resumption of primitive func- tion is impossible. In older animals and in animals whose period of embryonic development is longer than that of the chick, cell processes might well be slower. In the monkey, however, the period of recovery from spinal transection does not apparently last more than five or six weeks.** Porter °* has shown that the commissural path at the level of the phrenic nuclei, which is probably closed to the passage of respiratory impulses ordinarily, almost immediately opens up when the spinal cord is hemisected at the level of the first or second cervical nerve, e. g., on the right side, and the opposite, c. g., the left, phrenic nerve is cut. The evidence is perfectly clear, therefore, that the Severtiayen spinal shock is greater in the monkey and in the human than in any other mammalian form, and more severe in mammals than in the lower orders of vertebrates. Spinal shock begins in those ver- tebrate forms in which the neopallium, as distinguished from the archipallium, first appears as part of the cerebral cortex, and the transverse segmentation of the spinal cord has added to it the longitudinal segmentation represented by the long conduction path- ways of the cord.°® It increases in severity and duration as the cerebral cortex and the long conduction paths of the cord increase in complexity. The effects of removal of the cerebellum, which appears much earlier, phylogenetically, than the part of the cere- brum known as the neopallium, are more severe and more permanent than those following removal of the motor areas of the cerebral cortex (Luciani). These facts, thus briefly stated, and many Locy: Anatomischer Anzeiger, 1897, xiv, p. 113. SHERRINGTON: Philosophical transactions, Joc. cit. PorTER: Journal of physiology, 1895, xvii, p. 455. °° Herrick: Loc. cit. Physiology of the Central Nervous System. 151 others, the enumeration of which would require too much space at this time, have led me to believe that the mechanism of spinal shock consists, not in stimulation of efferent inhibitory pathways, nor in the loss of tone in levels below the transection, but in cutting off the normal conduction pathways for the reflexes. Recovery from shock, whether it be from spinal transection, removal of the motor areas or of other areas of the cerebral cortex, or removal of the cerebellum, consists, therefore, not alone in the failure or dis- appearance of a hypothetical inhibition nor in the regaining of tonus, but far more in the more or less complete assumption by more primitive structures in the line of phylogenetic development, of the function of the parts lost. These primitive parts have lain more or less dormant under ordinary conditions, but may again become active in some degree when the more recent structures have suffered ir- reparable injury. And the more complex the development of the structures appearing last in the phylogenetic chain, or, in other words, the more the central nervous system has departed from the primitive palzéncephalon, the less completely can the more primi- tive portions regain their primitive function. As regards the motor areas, we may consider this hypothesis to be such a modification or extension of the views of Luciani and Tamburini °° as the phylogeny of the central nervous system may demand. The decidedly less marked severity of spinal shock in young animals is sufficiently explained, not by assuming, as Babak °’ has done for the larval frog, that the efferent inhibitory fibres have not yet become excitable, but by supposing that Von Baer’s law of re- capitulation, known also as the law of biogenesis,°* applies to func- tion as well as to structure, and that those parts of the central nervous system which are the last to appear phylogenetically are also the last to reach their full functional development in ontogeny. The idea that the law of biogenesis applies to cerebral function 1s rendered far more than an assumption by the work of von Bech- terew,°® who showed that the myelination of the fibres of the pyram- idal tract begins about the tenth day after birth. 86 Lucrant and TampBurINt: Ricerche sperimentali sui centri psico-motori corticali; Reggio-Emilia, 1878; Brain, 1878-1879, i, p. 529; 1879-1880, li, p. 234. 37 Basak: Loc. cit. 88 SepGWICK: Quarterly journal of microscopical science, 1894, XXxvi, P 35; EIGENMANN: Mark anniversary volume, 1903, pp. 197-200. + 59 yon BECHTEREW: Neurologisches Centralblatt, 1888, vii, p. 14; ibid., 1889, Vili, p. 513. 152 Po id Pie. Considered in their relations to the spinal cord alone, it is diffi- cult, perhaps impossible, to decide between the two hypotheses. Con- sidered in their wider relationships, the evidence, in my opinion, favors the latter hypothesis. The detailed consideration of these points must, however, be postponed until I have presented further evidence drawn from the study of special reflex mechanisms, more particularly the vaso-motor mechanism, and the phenomena follow- ing destruction of the cerebral motor cortex by operation or disease, after which we may consider both hypotheses in their relation to the central nervous system as a whole along the lines suggested above. | ds HYDROLYSIS OF VITELLIN FROM THE HEN’S EGG.’ By THOMAS B. OSBORNE anp D. BREESE JONES. [From the Laboratory of the Connecticut Agricultural Experiment Station.] HE yolks of a large number of eggs were strained through cheesecloth, an equal volume of saturated sodium chloride solution added, and the mixture shaken with ether, containing a very little alcohol, until nearly all of the ether soluble substance was removed. The aqueous solution was then filtered through a felt of paper pulp and the perfectly clear solution dialyzed until all the globulin had precipitated. This precipitate was filtered out, redis- solved in 10 per cent sodium chloride solution, and the clear solution again dialyzed. When the globulin had been completely reprecipi- tated, it was filtered out, washed with water, then with 50 per cent alcohol, and digested with a mixture of equal parts of absolute alcohol and ether. After digesting several days the vitellin was col- lected on a filter and washed thoroughly with ether. When dried over sulphuric acid, a colorless powder was obtained which was used for the hydrolysis. HypDROLYSIS OF VITELLIN. Four hundred grams of vitellin, equivalent to 345.6 gm. ash and moisture free, were dissolved in 1000 c.c. hydrochloric acid, specific gravity 1.1, by heating on a water bath for two hours. The hydrolysis was then completed by boiling the solution in an oil bath for fifteen and a half hours. The hydrolysis solution was then concentrated, under diminished pressure, to a thick syrup and esterified according to the directions of Emil Fischer. The free esters were liberated, shaken out with ether and dried in the usual manner. After the salts were removed from the aqueous layer, the remaining amino-acids were subjected to a second esterification. After distilling off the ether from the ' The expenses of this investigation were shared by the Connecticut Agricultural Experiment Station and the Carnegie Institution of Washington, D. Be: 153 154 Thomas B. Osborne and D. Breese Jones. combined esters, at 760 mm. pressure, there were obtained 278 gm. of esters which were fractionally distilled as follows: Temperature of Fraction. bath up to Pressure. Weight. f 80° 15.00 mm. 5-32 gm. II Loo? 12/007 “ PU O yon Il (a) 96° o20) “ ASG ean (0) 106° roy A eo) IV 145° rome 9 a ASO 7 ae V 200° o20 R2STOr BG tal eG ee eee 183.73 gm. The undistilled residue weighed 58.00 gm. Fraction I. — Repeated attempts were made to separate glycocoll ester hydrochloride from this fraction, but none was obtained. The amino-acids were then regenerated by boiling with water, the chlo- rine removed, and the free acids added to the amino-acids from Fraction II. Fraction II. — This fraction was treated with concentrated hy- drochloric acid, and an attempt was made to separate the hydro- choloride of glycocoll ester, but none was found. After saponifying by boiling with water, the solution was evaporated under dimin- ished pressure and the hydrochloric acid removed with silver sul- phate and the sulphuric acid with baryta. The filtrate from the barium sulphate was evaporated to dryness under diminished pres- sure, and the amino-acids thus regenerated together with those from Fraction I were extracted with boiling alcohol to remove proline. By fractional crystallization of the part insoluble in alcohol, 5.34 gm. of leucine, 6.46 gm. of valine, and 2.59 gm. of alanine were obtained. The leucine was analyzed as follows: Carbon and hydrogen, 0.1539 gm. subst., gave 0.3090 gm. CO, and 0.1410 gm. HO: Calculated for C,H,,0,N = C 54.96; H 9.92 per cent. Found =). rk. =Crsay5 Eiroare ~ “ The valine crystallized in the characteristic plates. Carbon and hydrogen, 0.1448 gm. subst., gave 0.2728 gm. CO, and 0.1264 gm. HO» Calculated for C;H,,O,N = C 51.28 H 9.40 per cent. Bound! <2 (Meme mats = 6 51.38; El o7O..-0 8 Hydrolysis of Vitellin from the Hen’s Egg. 155 The alanine crystallized in hard, dense prisms, which decomposed sharply at 290°. Carbon and hydrogen, 0.1117 gm. subst., gave 0.1654 gm. CO, and 0.0792 gm. H-O©: Calculated for C;H,O,N = C 40.45; H 7.86 per cent. Bound: Aires js. = = = C 40.38; H785 “ “ The last fraction of amino-acids which separated from the final filtrate should contain glycocoll, if this were present. It was accordingly analyzed without recrystallizing, for if it contained glycocoll this should be indicated by a low percentage of carbon. As the analysis showed 42.36 per cent of carbon, little if any glycocoll could have been present, as the whole fraction weighed only about I gm. Fraction III. — This fraction was saponified by boiling with water until the alkaline reaction had disappeared. The solution was then evaporated to dryness under diminished pressure and the proline extracted from the residue by boiling with absolute alcohol. The part insoluble consisted chiefly of leucine, which weighed 28.77 gm. Carbon and hydrogen, 0.1500 gm. subst., gave 0.3026 gm. CO, and 0.1307 gm. He ©: Calculated for C,H,,0,N = C 54.96; H 9.92 per cent. IOUMCL: ws cc) =. eae = re O2) El o0S) sy From the filtrate of the leucine there were further obtained 3.77 gm. of copper aspartate which crystallized in the characteristic sheaves. Copper, 0.1020 gm. subst., gave 0.0292 gm. CuO. Calculated for C,H,O,N Cu 44 H,O = Cu 23.07 per cent. enn oboe: tor 23) lay? ct —eeoo tO yim. bar From the alcoholic extracts of Fractions II and III the proline was separated and weighed in the form of its copper salts, which contained 13.38 gm. of /-proline and 1.09 gm. of 7-proline. The /-proline copper was converted into the free acid and identified in the form of the phenylhydantoine which crystallized in beautiful prisms, melting sharply at 142°. 156 Thomas B. Osborne and D. Breese Jones. Nitrogen, 0.2154 gm. subst., required 20 c.c. N/1o HCl. Calculated for C,,.H,,O,N, = N 12.96 per cent. Bound! 2. ete ye ae = N t2:60m = = The air-dried racemic copper-salt was analyzed as follows: Water, 0.1403 gm., lost 0.0155 gm. H,O at 110°. Calculated for C,,H,,O,N,Cu * 2 H,O = H,O 10.99 per cent. MOUNC ows seek o Koo pikes eee =H Oki05 * ee Copper, 0.1446 gm. subst., gave 0.0348 gm. CuO. Calculated for C,,H,,O,N,Cu* 2 H,O = Cu 19.40 per cent. Pound Meese es en oe eee =O KO 230 oe Fraction Iv. — Phenylalanine was removed from this fraction by shaking with ether in the usual way, and 6.38 gm. of the hydro- chloride were obtained. The free phenylalanine decomposed at about 270° and gave the following analysis: Carbon and hydrogen, 0.1709 gm. subst., gave 0.4111 gm. CO, and 0.1039 gm. ~ H,O. Calculated for C,H,,O,N = C 65.45; H 6.66 per cent. Found Ss SOS = Cc 65.60; H 6.75 “c a From the aqueous layer, after saponification with baryta, there were obtained 5.57 gm. of aspartic acid, in the form of the barium salt. The free aspartic acid which resulted on decomposing the barium salt reddened at about 300° and gave the following analysis: Carbon and hydrogen, 0.1765 gm. subst., gave 0.2350 gm. CO, and 0.0853 gm. HO. Calculated for C,H,O,N =C 36.09; H 5.26 per cent. Houndiety ee eee aw OxrYorcn tds bry) pa re The filtrate from the barium aspartate, when freed from barium, concentrated to small volume and saturated with hydrochloric acid gas, yielded 6.99 gm. of glutaminic acid hydrochloride. Fraction V. From the ether extract of this fraction 4.59 gm. of phenylalanine hydrochloride were separated. The aqueous layer, after shaking out with ether, yielded 7.35 gm. of glutaminic acid in the form of the barium salt and 3.15 gm. as the hydrochloride. The free glutaminic acid decomposed with effervescence at about 2052 Hydrolysis of Vitellin from the Hen’s Egg. 157 Carbon and hydrogen, 0.2001 gm. subst., gave 0.2984 gm. CO, and 0.1109 gm. 1,0, Calculated for C;H,O,N = C 40.81; H 6.12 per cent. Found ge ee ee — 40.67 ; H 6.16 “ ‘cc No copper aspartate could be obtained from this solution. Owing to an accident, a considerable part was lost at the outset of the attempt to separate glutaminic and aspartic acids, but since it con- tained no aspartic acid the loss fell only on glutaminic acid. As the quantitative determination of this acid was made by direct isola- tion of the glutaminic acid hydrochloride from a separate portion of the vitellin, the final result of the analysis was not affected by the loss. Drrect DETERMINATION OF GLUTAMINIC ACID. There were taken for hydrolysis 86.26 gm. of ash and mois- ture free vitellin, which was dissolved by heating on a water bath for two hours in 300 c.c. of HCl, specific gravity 1.1. The hy- drolysis was completed by further heating in an oil bath at about 130° for eighteen and one-half hours. The solution was then concentrated and saturated with HCl gas. After standing in an ice box for several weeks there separated 13.51 gm. of glutaminic acid hydrochloride, which, after deducting the ammonium chloride, is equivalent to 25.25 gm. of glutaminic acid, or 11.95 per cent. ETHER DISTILLED FROM THE ESTERS AT 760 MM. The ether was treated with alcoholic hydrochloric acid, and the substance which separated after standing for several days was boiled with baryta until free from ammonia and worked up for glycocoll ester hydrochloride in the usual manner, but without success. THE RESIDUE AFTER DISTILLATION. The residue remaining after distillation of the esters, which weighed 58 gm., was dissolved in boiling alcohol, and after cooling I gm. of crystalline substance separated from the solution. The filtrate from this was saponified by boiling for five hours with 75 gm. of baryta, and, after removing the barium with an equiva- 158 Thomas B. Osborne and D. Breese Jones. lent quantity of sulphuric acid, 7.49 gm. of glutaminic acid hydro- chloride were separated by the usual process. TYROSINE. A quantity of the vitellin equal to 45.04 gm. of ash and moisture free substance was hydrolyzed by boiling for eighteen hours with 150 gm. of sulphuric acid and 300 c.c. of water. After removing the sulphuric acid with an equivalent quantity of baryta, the solu- tion was concentrated to crystallization, and after twenty-four hours the substance which separated was filtered out and recrystal- lized. The tyrosine after a second recrystallization weighed 1.50 gm., equal to 3.37 per cent. Nitrogen, 0.1544 gm. subst., required 8.65 c.c. N/1o-HCI. Carbon and hydrogen, 0.1509 gm. subst., gave 0.3289 gm. CO, and 0.0866 gm. HO: Calculated for C,;H,,0,N = C 59:67; 116.08; Ni 7-73: Mout dh iiejtayer uae varus ==C 50:43); EL 6:375 N172384: The filtrate from the tyrosine was used for determinations of the bases according to the method of Kossel and Patten. HISTIDINE. The solution of the histidine = 500 c.c. Nitrogen, 50 c.c. sol., required 2.32 c.c. 5/7 N-HCl = 0.2320 gm. N in 500 c.c. = 0.8561 gm: histidine = 1.90 per cent. The histidine when converted into the dichloride melted at 233°. Chlorine, 0.1076 gm. subst., gave 0.1335 gm. AgCl. Calculated for C,H,,O,N,Cl, = Cl 31.14 per cent. Hound Fes ee == "Ol 2o1G8) 6 ie ARGININE. The solution of the arginine = 1000 c.c. Nitrogen, 50 c.c. sol., required 5.13 c.c. 5/7 N-HCl = 1.0260 gm. N. in 1000 c.c. = 3.1878 gm. arginine + 0.1700 gm. = 3.3578 gm. = 7.46 per cent. Hydrolysis of Vitellin from the Hen’s Egg. 159 The arginine was converted into the copper-nitrate double-salt for identification. Copper, 0.1053 gm. subst., gave 0.141 gm. CuO. Calculated for C,,H,,0,N,Cu (NO,), 3 H,O = Cu 10.79 per cent. ER Te eave kd i nay oa a, = Om abo 5 LYSINE. The lysine picrate weighed 5.5650 gm. = 2.1665 gm. lysine = 4.81 per cent. Nitrogen, 0.1590 gm. subst., required 20.8 c.c. N/1o HCl. Calculated for C,H,,0,N,*C,H,0,N, = N 18.67 per cent. Bina ys a 8 ao SS = NP asian, oo ee Hydrolyses of vitellin from hen’s eggs have been made by Hugouneng,”? Levene and Alsberg,? and by Abderhalden and Hunter. Since very wide differences exist between the results obtained by these investigators, we determined to make the hydrolysis just described in order, if possible, to obtain more definite information as to the proportion in which this important protein yields the several amino-acids. Hugounengq’s results for most of the amino-acids are so far below those found by all the others who have analyzed this protein that they need no further consideration. Those reported by Levene and Alsberg for the mono-amino-acids are unquestionably too low, for they made their distillation at 15 mm. and the quantity of esters which they obtained corresponded to only about one half the yield which we secured. Levene and Als- berg call attention to their small yield of esters and state that much time elapsed between the first and second esterification. In view of these facts it is evident that a comparison of their results cannot be made with those of Abderhalden and Hunter or with ours. In respect to the total amount of esters recovered after distilla- tion, the quantity obtained by Abderhalden and Hunter, 158 gm., agrees fairly well with the 184 gm. which we obtained, and the. 2 HucouNENQ: Comptes rendus des Séances hebdominaires de |’Académie des Sciences, 1906, cxlii, p. 173. 8 LEVENE and ALsBERG: The journal of biological chemistry, 1906, ii, p. 127. 4 ABDERHALDEN and Hunter: Zeitschrift fiir physiologische Chemie, 1906, xlviii, P- 595; 160 Thomas B. Osborne and D. Breese Jones. quantity obtained below 105°, namely, 97 gm., also agrees very closely with that which we got below 106°, namely, 103 gm. The results given by Abderhalden and Hunter for most of the amino- acids agree satisfactorily with those which we obtained, as shown by the following figures: HYDROLYSIS OF VITELLIN FROM THE HEN’s EGG. | Abderhalden | Osborne and Hunter. | and Jones. Abderhalden} Osborne Substance. aad clans eee Jones. Substance. Glycocoll, “2 27 1.10 0.00 iyrosine semen 1.60 A824) Alanine ... .| + 0.75 Cystine? 22° Pe 2 not deter. Valine 2.40 1.87 | Histidine. . - Nees 1.90 Weucine ma eae 11.00 9.87 | Arginine: ey eee Hants 7.46 Prolsie™ =) == 2] 3.30 4.18 || Lysme- 5 2 eee 481 Phenylalanine .| 2.80 2.54 Ammonia .. See 1.25 Aspartic acid . . 0.50 213 Tryptophane . Pes present Glutaminic acid 12.20 12.95 Phosphorus”. Soot 0.94 Serine | ; ? | [ Total The most striking difference, and one which we cannot explain, is shown by glycocoll, which we were wholly unable to find, although persistent efforts were made to do so. The difference between the percentage of aspartic acid in these two analyses is relatively large, but from such data as are now available it would appear that deter- minations of aspartic acid are among the most uncertain of all of the protein decomposition products. We made no attempt to de- termine cystine or oxyproline. HYDROLYSIS OF THE MUSCLE OF SCALLOP (PECTENS IRRADIANS).} By THOMAS B. OSBORNE anp D. BREESE JONES. [From the Laboratory of the Connecticut Agricultural Experiment Station.] HITTENDEN ? found in the large adductor muscle of the American scallop (Pectens irradians) a relatively large amount of free glycocoll. It is therefore of interest to know whether the muscle substance of this mollusk is relatively rich in glycocoll or presents any other peculiar feature in the propor- tion of the amino-acids which it yields by hydrolysis. A quantity of the muscles of the scallop were carefully removed from the still living mollusk and freed from all the other tissues. The muscles were then suspended in water, and each separately re- moved and dropped into a large volume of water containing toluol. After standing over night the muscles had swelled greatly and ab- sorbed about two-thirds of the water. They were next dropped, one by one, into about 3 litres of yater which was kept con- stantly boiling. The water in which they had stood over night was then added, and the whole boiled for fifteen minutes. The coagulated muscle substance was strained out on cheesecloth and disintegrated by crushing with the hand. It was then washed twice with distilled water and strained and pressed on cheese- cloth after each washing. It was next suspended in water satu- rated with toluol, allowed to stand over night and then thoroughly pressed in a hydraulic press. After breaking up the press cake, it was digested for forty-eight hours with 95 per cent alcohol, squeezed out on cheesecloth, and again digested with 95 per cent alcohol. After digesting a third time with absolute alcohol, the residue was extracted with ether and dried in the air. Thus prepared, it contained 13.13 per cent of moisture and 0.66 ? The expenses of this investigation were shared by the Connecticut Agricultural Experiment Station and the Carnegie Institution of Washington, D. C. ? CHITTENDEN: American journal of science and arts, 1875 (3), x, pp. 26-32. 161 162 Thomas B. Osborne and D. Breese Jones. per cent of ash. The ash and moisture free material contained 17.05 per cent of nitrogen. HypDrROoLysIs OF SCALLOP MUSCLE. Three hundred grams of the scallop muscle, equivalent to 258.6 gm. of ash and moisture free substance, were hydrolyzed in three separate portions by heating on a water bath for three hours with 600 c.c. of hydrochloric acid, specific gravity 1.1. The hydrolysis was then completed by boiling for twenty-four hours in an oil bath. The glutaminic acid was separated from each of the hydrolysis solutions as the hydrochloride in the usual way, and after deduct- ing the ammonium chloride which it contained, 14.08 gm., 13.83 gm., and 13.47 gm., respectively, were obtained from the three por- tions. With the 6.66 gm. subsequently obtained from the esters, the total glutaminic acid hydrochloride weighed 48.04 gm., equivalent to 38.5 gm. of glutaminic acid or 14.88 per cent of the muscle. The free acid decomposed with effervescence at 202°—203°. Carbon and hydrogen, 0.1701 gm. subst., gave 0.2546 gm. CO, and 0.0977 gm. is ROP Calculated for C;SH,O,N = C 40.81; H 6.12 per cent. Found). 22 . 4 = "'Co.825 Ease a The filtrate from the gluttaminic acid hydrochloride was con- centrated to a thick syrup, freed from water by evaporation with alcohol under reduced pressure, and esterified in the usual way. After separating the free esters with ether, the inorganic salts were removed from the aqueous solution and the esterification re- peated. The esters thus obtained, which were freed from ether by distillation from a water bath at atmospheric pressure, were then distilled i vacuo with the following results: Temperature of Fraction. bath up to Pressure. Weight. I 100° 9.0 mm. 28.49 gm. II 100° OS) oe 2222 oT 200° co fy. a 55232 Total 2. <-:. kee 109.03 gm. The undistilled residue weighed 13.00 gm. Fraction 1. — The esters of this fraction were saponified by boil- ing with water until the alkaline reaction had ceased, when the FAlydrolysis of the Muscle of Scallop. 163 solution was evaporated to dryness under strongly diminished pres- sure and the proline extracted from the dried residue by boiling with absolute alcohol. From the amino-acids insoluble in alcohol, there were isolated 8.77 gm. of leucine. The filtrate from the leucine was evaporated to dryness and the residue esterified with alcohol and dry hydrochloric acid gas. After long standing at 0°, no glycocoll ester hydrochloride separated, even after repeating the process of esterification several times. The amino-acids, when regenerated and freed from chlorine, yielded 2.20 gm. of leucine and 4.0 gm. of a mixture from which nothing definite could be separated. The leucine had the following composition: Carbon and hydrogen, 0.1381 gm. subst., gave 0.2772 gm. CO, and 0.1238 gm. HO: Calculated for C,H,,NO, = C 54.96; H_ 9.92 per cent. mance sek ee =—"Cce7a, we roids, °° Fraction I1.— This fraction consisted almost entirely of the esters of leucine and proline. After saponifying by boiling in the usual way with water, the solution was evaporated to dryness under diminished pressure and the proline extracted with boiling alcohol. The part which was insoluble in alcohol yielded on frac- tional crystallization 10.93 gm. of not quite pure leucine. On account of a very slight admixture of some substance which could not be readily separated, this leucine was converted, into the copper salt, which gave the following results on analysis: Copper, 0.1264 gm. subst., gave 0.0308 gm. CuO. Calculated for (C,H,,NO,.),Cu = Cu 19.64 per cent. Bee eat Goce. so. = Curg4ar = * Copper, carbon, and hydrogen, 0.1645 gm. subst., gave 0.0400 gm. CuO, and 0.2677 gm. CO., and 0.1079 gm. H,O. Calculated for (C,H,,.NO,),Cu = C 44.46; H.7.47; Cu 19.64 per cent. it rs A =e AA 85s) El. 94s -Cuto.4s. = ee The alcoholic extracts of Fractions I and II were joined and worked up for proline. After evaporating this solution under diminished pressure to dryness, the residue was extracted with ab- solute alcohol and the evaporation and extraction with alcohol re- peated until the residue was completely soluble in absolute alcohol 164 Thomas B. Osborne and D. Breese Jones. and yielded no deposit on standing. The proline was then con- verted into the copper salt, and the lzevo separated from the racemic by boiling with absolute alcohol. The racemic salt was freed from copper with hydrogen sulphide and the solution evaporated to dry- ness under diminished pressure. A small residue remained, in- soluble in absolute alcohol, which was filtered out and the proline again converted into the copper salt. Water, 0.2050 gm. subst., air dried, lost 0.0223 gm. H,O at 110°. Calculated for C,,H,,O,N.Cu 2H,O = H,O 10.99 per cent. - Pound)... -esce Cons ae wie @ TOrs amet. 1 Copper, 0.1819 gm. subst., dried at 110°, gave 0.0496 gm. CuO. Calculated for C,,H,,0,N,.Cu = Cu 21.81 per cent. Pound 5.5 isin! s vere es = Cuysor7ss res The copper salt of the levo-proline, which was soluble in alcohol when dried at 100°, weighed 7.49 gm., equivalent to 5.91 gm. of leevo-proline. The free acid was then regenerated and converted into the phenylhydantoine derivative. This latter crystallized from water in characteristic long prisms which melted at 143°. Carbon and hydrogen, 0.1899 gm. subst., gave 0.4635 gm. CO, and 0.0966 gm. HO: Calculated for C,.H,.N°O; —) © '06:67— 15°57 percent: ound te. 7s foe cee =" @.66:57 bt 5-005 be Fraction 111. — The phenylalanine ester was removed by shak- ing this fraction with ether in the usual manner. The residue left by evaporating off the ether was saponified by heating with strong hydrochloric acid, and 10.92 gm. of phenylalanine hydro- chloride, equivalent to 8.94 gm. of phenylalanine, were obtained. The free phenylalanine, which resulted by treating the hy- drochloride with an excess of ammonia, had the following composition : Carbon and hydrogen, 0.1685 gm. subst., gave 0.4037 gm. CO,, and 0.1038 gm. HO: Calculated for C,H,,O,N = C 65.45; H 6.66 per cent. Found: (cere. =/(C 65.34; T116:80) ane The final filtrate from the phenylalanine hydrochloride yielded a thick syrup, from which nothing could be directly separated. Hydrolysis of the Muscle of Scallop. 165 This syrup was dissolved with boiling water, an excess of ammonia added, the solution decolorized with bone black, and concentrated. There was thus further obtained 3.74 gm. of free phenylalanine, making a total of 12.68 gm., or 4.90 per cent. The aqueous solution, from which the ester of phenylalanine had been removed by shaking with ether, after saponifying in the usual manner with baryta, yielded 3.90 gm. of aspartic acid, as the barium salt, which was converted into the copper salt for anaiysis. Copper, carbon, and hydrogen, 0.2013 gm. subst., gave 0.0583 gm. CuO, 0.1313 gm. CO,, and 0.0920 gm. H,O. Calculated for C,H;0,NCu 44 H,O = Cu 23.07; C 17.41; H 5.12 per cent. NG Re cy Sg» es bx. 2) Sis s = Cwes14; Cr 7 os s-Ccoe ls The filtrate from the barium aspartate was freed from barium, the solution concentrated, and saturated with hydrochloric acid gas. There separated 6.66 gm. of glutaminic acid hydrochloride. Chlorine, 0.1696 gm. subst., gave 0.1329 gm. AgCl. Calculated for C;H,,O,NCI = Cl 19.35 per cent. otnd, s\>. coy se = Clergy hi | The free glutaminic acid obtained from this hydrochloride de- composed with effervescence at 202°—203°. Carbon and hydrogen, 0.2016 gm. subst., gave 0.3018 gm. CO, and 0.1127 gm. reo. Calculated for C;-H,O,N = C 40.81; H 6.12 per cent. LVS ois a rs = C 40.83; H 6.26 ‘“ “c The filtrates from the glutaminic acid hydrochloride, after being freed from chlorine, yielded 10.50 gm. of copper aspartate, which separated from a large volume of water in the characteristic sheaf- like groups of crystals. Copper, 0.1609 gm. subst., air dried, gave 0.0461 gm. CuO. Calculated for C,H;O0,NCu 44 H,O = Cu 23.07 per cent. CINE o> oy of Mat Je er ean Gh == ue. ba), The filtrate from the copper aspartate, after removing the cop- per with hydrogen sulphide, yielded further 0.89 gm. of nearly pure leucine: 166 Thomas B. Osborne and D. Breese Jones. Carbon and hydrogen, 0.1115 gm. subst., gave 0.2261 gm. CO, and 0.0917 gm. HO: Calculated for C,H,,NO, = C 54.96; H 9.92 per cent. Pound’ 2. 222/ Jo. =.€ s5:30s4H ozo fs “ Nothing could be separated from the undistilled residue, which weighed only 13 gm. The ether, which was distilled from the esters at 760 mm., was carefully examined for glycocoll, but none was found. TYROSINE. A quantity of ash and moisture free scallop muscle, weighing 25.87 gm., was hydrolyzed by boiling for twenty-four hours with a mixture of 49 c.c. of concentrated sulphuric acid and 181 c.c. of water. After removing the sulphuric acid quantitatively with baryta and washing the barium sulphate until free from tyrosine, the solution was concentrated on a water bath to crystallization. The substance which separated after twenty-four hours was dissolved in water, the solution decolorized by boiling with animal charcoal, and the tyrosine separated by concentration and cooling. By re- crystallization 0.5058 gm. of pure tyrosine was obtained = 1.95 per cent. Nitrogen, 0.1008 gm. subst., required 0.77 c.c. 5/7 N HCl. Calculated for C,H,,O,N = N 7.73 per cent. Hound: 2030. es = — aN OA The mother liquor and washings from the tyrosine were exam- ined for bases according to the method of Kossel and Patten. HISTIDINE. The solution of the histidine = 500 c.c. Nitrogen, 50 c.c. sol., required 1.43 c.c. 5/7 N HCl = 0.1430 gm. N in Soo c.c. = 0.5277 gm. histidine = 2.04 per cent. The histidine was converted into the dichloride for identification. Chlorine, 0.1004 gm. subst., gave 0.1266 gm. AgCl = 0.0313 gm. Cl = 31.19 per cent. Calculated for C,H,,O,N,Cl, = Cl 31.14 per cent. Found =< 02 arctan = = Cl 2nito) Ss *: Hydrolysis of the Muscle of Scallop. 167 ARGININE. The solution of the arginine = 1000 c.c. Nitrogen, 50 c.c. = sol., required 2.91 c.c. 5/7 N HCl = 0.5838 gm. N in 1000 c.c. = 1.8083 gm. arginine, + 0.1026 gm. = 1.9108 gm. = 7.38 per cent. The arginine was converted into the copper-nitrate double-salt for identification. Copper, 0.2195 gm. subst., gave 0.0300 gm. CuO. Calculated for C,,.H,,0,N,Cu(NO,), 3 H,O = Cu 10.79 per cent. 2 LIE WOR aie Se eee adn = Cu idigg? "2. a: LYSINE. The lysine picrate weighed 3.8323 gm.=— 1.4918 gm. lysine = 5-77 per cent. Nitrogen, 0.2565 gm. subst., required 4.74 c.c. 5/7 N HCl. Calculated for C,H,,O,N, ° C,H,0;N; = N 18.67 per cent. SIGE Foo! oy meet: Jew! “Sy pas 2 = N wae oe PARTITION OF NITROGEN. The different forms of nitrogen yielded by hydrolyzing the scal- lop muscle by Hausmann’s method, as modified by Osborne and Harris, were as follows: Nitrogen asammonia ..--.-.--..- 1.08 per cent. Basicwitrogen - — 5. .'s.-,2 - & =» = = yA a mam-pasie TiktrOpen 5.5.8 ere. Pron see ce “ Nitrogen in magnesium oxide precipitate . 0.40 Paral murogels (7/5. a) a. eo va -e 17.05 per cent. The nitrogen contained in the histidine, arginine, and lysine is 4.02 per cent, or 0.50 per cent less than the nitrogen precipitated by phosphotungstic acid, —a difference similar to that found for fish and chicken muscle, and probably due to basic substances of non-protein origin. The results of this hydrolysis of scallop muscle, together with 168 Thomas B. Osborne and D. Breese Jones. those previously obtained with chicken and halibut muscle, are given in the following table: Scallop Halibut Chicken per cent. per cent. per cent. Gilycocolly ria) -o yo ese toe tec 0.00 0.00 0.68 ANGIE A ra. pessoas eer as Be ee 2.28 Walin@ cS c4 ia ueh cs ete me eis 0.79 inte eueiuey, poo. soe oe eee 8.78 10.33 II.19 LONER pose ones foie ee 2.28 B17 4-74 Phenylalanine:s = 2 5.2 ise." 4.90 3.04 3.58 AASParlicsacid | oy.a oe. = eee 3.47 2.93 B28 Gintaminicacid 25.2 sens 14.88 10.13 16.48 SOHC) ai. oe Sued aes Se ees ee aes ae SPYTOSING 4. 2 ewiteees ic eee 1.95 2.39 2.16 Arcininey os 2h sa = ieee 7.38 6.34 6.50 ENSGiNe Teese ro ae ws ent 2.02 2.55 2.47 TS V SUNG oad — minutes for the temperature to rise from 45° to 75°, 50° 60° 70° In a small muscle like the frog’s sartorius most of the myogen must be coagulated after twenty-four hours’ immersion in 96 per cent alcohol. But a muscle in which a large proportion of the myo- gen has been so coagulated is more capable of shortening at 55° than is a fresh muscle. Figs. 1 and 2 represent the heat curves of the sartorii from opposite legs of the same frog. The two muscles were heated in the same vessel to 47°. One was then arranged to FicurE 2.— Curve obtained from the sartorius fellow to that of which the heat shortening is recorded in Fig. 1. This muscle, after being heated along with its fellow to 47°, was immersed for twenty-four hours in 96 per cent alcohol. It was then trans- ferred for five hours to 0.7 per cent NaCl solution, and heated —— in the salt solution at such a rate that it required four minutes for the temperature to rise from 40° to 75°. This curve and that of Fig. 1 were made under the same experimental conditions. In both cases the magnification of the writing lever was 5; the muscle was weighted with 0.6 gm.; and the length of muscle between attachments at the beginning of the curve was about 8 mm. write its changes in length on a kymograph, and further heated in 0.7 per cent NaCl solution to 75°. Fig. 1 represents the curve so obtained. The other muscle was also arranged to write its changes in length on a kymograph, care being taken that the length of muscle between attachments was the same as in the first case and was im- mersed for twenty-four hours in 96 per cent alcohol; during this period it lengthened slightly. It was then immersed for five hours in 0.7 per cent NaCl solution, which produced no change in its length. The muscle was treated with salt solution, because the heat short- ening will not take place in tissue saturated with 96 per cent alcohol 184 Edward B. Meigs. and heated in that fluid. If 70-per cent alcohol be substituted, a shortening takes place at about 70°, but is slower and not so high as in fresh muscle. But if the alcohol be entirely substituted by water, the shortening occurs just as it does in untreated muscle. After having remained for five hours in the salt solution, the piece of muscle was gradually heated in that fluid and gave the heat curve shown in Fig. 2. It will be observed that its heat shortening is con- siderably larger than that shown in Fig. 1. This is usually the case when two such muscles are compared with each other, though the difference is often less marked than in the curves given. The following experiment shows that a muscle will still exhibit a marked heat shortening at 55° after all its proteins have been coagulated by alcohol. The muscles of a frog’s leg were kept for eleven days in 96 per cent alcohol. At the end of that time they were immersed for about twenty hours in 0.7 per cent NaC! solution. From the muscle which had been treated as described with alcohol and 0.7 per cent NaCl solution a small piece 8 mm. long was arranged to write its changes in length on a kymograph, and gradually heated in the salt solution to 80°. The curve so obtained is reproduced in Fig. 3. The other leg muscles were extracted with 0.7 per cent NaCl solution, and the extract was filtered. The filtrate, which was very slightly opalescent, was heated gradually to boiling; it showed not the slightest sign of precipitate or in- crease in opalescence at any temperature. The initial shortening which occurs in striated muscle under the influence of heat is almost certainly of an entirely different nature from the shortening which occurs at about 55°. It is quite characteristic of striated muscle; no other tissue shortens 30 per cent or 40 per cent of its original length at a temperature below 40°. It has the same height as the greatest shortening which can be produced in living muscle by the tetanizing current.’* It is not accompanied by the tendency to lose water, which has been shown to be characteristic of the shortenings which occur at about 55° in smooth and striated muscle and in catgut. Finally, it is ac- companied by,'® and probably the result of, the formation of a large quantity of lactic acid within the muscle. It is not surprising, 14 VERNON: Loc. cit., p. 276. 18 Meics: Loc. ctt., p. 10. 16 FLETCHER and Hopkins: Journal of physiology, 1907, xxxv, p. 247. 1 Meics: Loc. cit., pp. 12 and 13. Protein Coagulation and Heat Shortening. 185 therefore, that muscles which have been killed by the long-continued action of strong alcohol do not shorten when heated to 40°. It may be shown, however, that the proteins of a fresh muscle can be coagulated gradually within the muscle without causing any shortening comparable to the initial heat shortening. An irritable ee 40° 50° 60° 70° 80° 40° 350° 60° 70° 70° FicurE 3. — Curve showing the heat shortening Ficure 4.— Curve showing the heat of a piece of muscle, which had been kept for eleven days in 96 per cent alcohol, and then for twenty hours in 0.7 per cent NaCl solution. The heating was carried out at such a rate that it required eight minutes for the tempera- ture to rise from 40° to 80°. Magnification of writing lever, 5; weight, 0.6 gm.; length of muscle between attachments at beginning of experiment, 8 mm.; proportional shortening about 25 per cent. The shortening is some- what greater in this case than in Figs. 1 and 2, shortening of a piece of muscle which had been treated first with weak and then with strong alcohol. The heat- ing was carried out at such a rate that it required six minutes for the temperature to rise from 40° to 80°. Magnification of writing lever, 5; weight, 0.6 gm.; length of muscle between attachments at beginning of experiment, 14 mm. ; proportional shortening, about 36 per cent. because this muscle had not been heated to 47° before immersion in the alcohol. frog’s sartorius was weighted with 0.6 gm. and arranged to write its changes in length ona kymograph. It was then immersed for two hours in a mixture containing 9 parts of 0.7 per cent NaCl solution and 1 part 96 per cent alcohol. During this period it underwent no shortening. The mixture of alcohol and salt solution was then replaced by pure 96 per cent alcohol, in which the muscle remained for forty-four hours. During this period it underwent an insig- nificant shortening of about 4 per cent; even this could probably have been avoided by a longer treatment with the mixture of alcohol and salt solution. After the treatment with alcohol, the muscle was placed for five hours in 0.7 per cent NaCl solution, and then gradually heated in that fluid; it gave the curve shown in Fig. 4. This result shows that the proteins cf muscle may, with proper precautions, be coagulated without causing any considerable 186 Edward B. Meigs. shortening, and that muscles of which the proteins have been coagulated under these conditions show the most marked shortening when heated to 55° and above. There is, then, no reason to believe that there is any connection between protein coagulation and the heat shortening which occurs at temperatures above 50° in striated muscle, smooth muscle, and connective tissue; and these are the tissues which exhibit the phe- nomenon most markedly. The experiments of Brodie and Halli- burton ?§ with nerve and liver are hardly sufficient to -demonstrate that a connection between the two phenomena exists in those tissues. These experiments are open to many of the objections which have been urged against the conclusions of Brodie and Richardson, and to some others besides. The authors, in determining the tempera- tures at which the various shortenings occur in nerve, select “ typ- ical’’ examples instead of taking the average from a number of experiments. The curves given represent the heat shortening of the sciatic nerve, while the temperatures at which the proteins of “ ner- vous tissues’ coagulate were determined in extracts of the brain. Finally, too little attention is paid to the fact that both “ nerve” and “liver,” as prepared by the authors, are complex structures, containing certainly connective tissue, and possibly elastic tissue also, and that the steps which they obtain in the heat shortening may be the result of simple shortenings occurring in these various tissues at different temperatures. The facts which have been reported in this article do not, of course, preclude the possibility that the precipitation of protein from its solutions and the shrinkage of animal tissues under the influence of heat may be fundamentally more or less similar processes. They do show, however, that the shortening of striated muscle at tem- peratures above 50° is independent of the coagulation of myogen, and they make it seem probable that the heat shortening of most animal tissues is dependent, not on the aggregation of the particles of coagulable protein, but on some other process. 8 BRoprie and HALIBURTON: Journal of physiology, 1904, xxxi, p. 473. MERCURIAL POISONING OF MEN IN A RESPIRATION CHAMBER. By THORNE M. CARPENTER anp FRANCIS G. BENEDICT. [Nutrition Laboratory, Carnegie Institution of Washington, Boston, Massachusetts.] N connection with an extended series of investigations on the respiratory exchange and heat production of man a respiration calorimeter embodying a respiration apparatus of the type Regnault and Reiset and an adiabatic constant flow calorimeter has been used for a number of years at Wesleyan University, Middletown, Con- necticut. Formerly this apparatus was on the open circuit plan of Pettenkofer, although in 1902 it was modified so as to include the direct determination of oxygen.’ The apparatus was in constant use from March 24, 1903, to March 29, 1905, there being a total of 75 experiments with 115 days of experimenting. Thirty-one different subjects were used in these experiments in the apparatus as thus used. The air leav- ing the respiration chamber was forced by means of a rotary blower through a valve system which allowed it to pass through one of two sets of purifiers. The purifying system consisted, first, of a porcelain vessel filled with sulphuric acid for the removal of water ; second, a silver plated can containing soda lime for the removal of carbon dioxide; and third, another sulphuric acid vessel to remove the water taken up by drawing the air as it passed through the soda lime. The air leaving the purifiers was considered anhydrous and free from carbon dioxide. It was then caused to return to the chamber, oxygen being admitted as the oxygen was consumed. In adapting the apparatus to the Regnault and Reiset plan it was ab- solutely necessary to secure perfect closure of all connections. Since the purifying apparatus for removing water and carbon dioxide 1 The experiments here reported were made in the Chemical Laboratory of Wes- leyan University, Middletown, Connecticut. 2 W. O. AtwaTER and F. G Benepict: A respiration calorimeter with appliances for the direct determination of oxygen: Publication No. 42, Carnegie Institution of Washington. 187 188 Thorne M. Carpenter and Francis G. Benedict. from the ventilating air must be changed frequently during long experimental periods, a valve system was necessary. Recourse was had to a valve of special construction employing a mechanical seal bathed with mercury to insure perfect closure. The valves were placed at both ends of the purifying system. Since in its passage through the purifying system the air must be forced through a con- siderable layer of sulphuric acid, the pressure of the air as it entered the first purifying vessel was not far from 40 mm. of mercury. The purifying vessels were changed every two hours, consequently one side of the valve was under a pressure of 40 mm. of mer- cury, while the other side was under atmospheric pressure during the time that the absorbing system was disengaged for weigh- ing. Under these conditions we found it exceedingly difficult to secure valves with a large enough aperture which could be abso- lutely tight. The type of valve finally used is described in consider- able detail and shown in Fig. 10 on page 21 of Publication No. 42 of the Carnegie Institution of Washington. The closure is made by forcing a steel disk on the end of a threaded spindle against the end of the steel tube about 30 mm. in diameter. To make the closure all the more mechanically perfect a hard rubber or fibre gasket was placed in the steel disk, and as a fimal precaution a reservoir of mercury connected through a rubber tube with the valve chamber was raised in such a manner that mercury could flow about the mechanical closure and thus produce a mercury seal. Under these conditions the closure was perfectly satisfactory. When it was desired to open the valve, the mercury reservoir was first lowered, during which process mercury drained away from the valve chamber, and then the valve was opened by turning the spindle. In thus raising or lowering the mercury it was impossible to prevent small globules of mercury from adhering to the inside of the iron valve chamber, although all but a very small proportion of the mercury flowed into the lowered reservoir. This valve was originally designed to meet extreme conditions of pressure at the entrance end of the absorber system, but inasmuch as it had proven absolutely tight to this high pressure it was found expedient also to use them at the exit end, where the pressure to be guarded against was very much less and amounted to only a few millimetres of mercury, 7. ¢., a pressure necessary to overcome the resistance of the air current in its passage through 9.2 metres of iron pipe 50 mm. internal diameter on its return to the respiration Mercurial Poisoning of Men. 189 chamber. It is particularly with the set of valves at the exit end of the purifying system that we have to deal. General plan of experiments. — In order to bring the calorimeter into temperature equilibrium, the subjects usually-enter the chani- ber some few hours before the experiment proper begins. In many instances the experiment had been previously planned to study the effects of inanition on metabolism. In nearly all cases, how- ever, the subjects had had food but a few hours prior to being in the experiment. In those experiments in which the subject slept in the chamber over night he usually entered the apparatus about 9 P.M. and retired at 11 P.M., sleeping on the cot in the chamber. He was awakened at 7 a.m. In certain other experi- ments the object was simply to study the normal respiratory ex- change of persons sitting quietly at rest, and the subject had no preliminary preparation other than one hour inside the chamber before the experiment proper began. Finally, some of the experi- ments were made with the special purpose of studying the question at hand, namely, toxic poisoning resulting from a sojourn in the chamber. The special conditions under which these experiments were made are discussed in more detail in connection with each experiment. The subjects were in most cases young men, students in the University, although two were men of more mature years. Indices of physical conditions when subjects were inside the respira- tion chamber. — The complete hermetical closure of this respira- tion apparatus makes it practically impossible to come in phys- ical contact with the subject. Our observations of his welfare were confined to the use of the telephone and inspection through a glass window. Certain factors, however, could be determined with as great accuracy as they could be were the subject outside of the chamber. Thus it was possible to determine with great accuracy the pulse rate, respiration rate, and body temperature. The pulse rate and respiration rate were determined by using a pneumograph. The body temperature was at times taken in the usual way with a carefully calibrated mercurial clinical thermometer under the tongue. In general temperature observations were made with a much higher degree of accuracy by means of an electrical resist- ance thermometer * which was inserted deep in the rectum. The * F. G. Benepicr and J. F. Sneti: Archiv fiir die gesammte Physiologie, 1901, Ixxxviii, pp. 492-500. 190 Thorne M. Carpenter and Francis G. Benedict. variations in resistance were measured by an observer outside the calorimeter and they were recorded every few minutes. They are accurate to o°.or C. Since this thermometer could *be worn without discomfort during sleep, the records are continuous. Personal impressions. —In any series of experiments such as these here reported, subjects are continually influenced by per- sonal impressions which are for the most part without scientific foundation. Thus it is a common matter for the men to state that there are marked changes in the temperature inside of the chamber and that the ventilation is better at different times, while, as a matter of fact, the temperature remains constant hour after hour and the rate of ventilation remains likewise unchanged. How- ever, with the symptoms appearing in some of these cases, there is no question but that the men were absolutely wretched and their personal impressions were borne out by the data secured from the pulse rate, respiration rate, and body temperature. After the subjects left the calorimeter, they were under the con- stant supervision of a skilled physician, and the personal impressions were again substantiated by his careful observations. Collaborators. — The striking and characteristic symptoms appear- ing in these experiments attracted the attention of many scien- tific men in the neighborhood of Middletown. We were fortunate to have in consultation with us a number of experts whose atten- tion to this problem we wish especially to emphasize. Dr. A. R. Diefendorf, formerly pathologist of the Connecticut Hospital for the Insane, and at present lecturer in psychiatry at Yale University, devoted considerable time to the problem, and Dr. John E. Love- land was in constant attendance on these subjects. Appearance of toxic effects. — While, as is usual with any number of experiments with men of varying temperaments, there had been in the previous experiments occasional complaints that the subjects had slight headaches or were not thoroughly comfortable inside of the chamber, no particular attention was paid to them until one experiment in May, 1905, when a subject, who was of unusually stolid temperament, was taken violently ill with nausea and fainting several hours after the experiment. From this time on, the com- plaints of discomfort were very frequently observed, and actual disturbances of temperature regulation were soon noticed in a large number of experiments. These experiments, all of which were made in the spring and fall of 1905, are given below. Mercurial Poisoning of Men. IOI EXPERIMENTS WITH Toxic SYMPTOMS. Experiment I. — Subject, C. R. Y., March 29-30. This experiment was originally planned to cover three days to study the influence of sleep and body position on metabolism. On the first day the subject was to lie in bed all day but not to sleep if possible, although he was expected to ob- tain the normal amount of sleep on the preliminary night. The experi- ment proper was supposed to begin at 7 A.m., March 30. The subject entered the chamber at 9 p. m., March 29, and the observations began at 11 P. M., the subject going to sleep on the cot provided inside the cham- ber. In the morning when he awoke at 7 A. M., he reported he had slept fairly well, waking up two or three times, and he ate breakfast consisting of cereal, crackers, and milk. During the forenoon it was noticed that the respiration was rapidly increasing, and at one o’clock the subject was nauseated and vomited, but preferred to stay inside the chamber, think- ing he would feel better later on, although he said his lungs were sore at 3 P.M. It was necessary to remove him from the chamber at this time. His temperature as indicated by the clinical thermometer under his tongue was 39°.19. The electrical resistance thermometer had shown a distinct rise in temperature during the whole day. The pneumograph for recording the pulse and respiration was unfortunately not used in this experiment. After the subject left the laboratory there were no observa- tions regarding his general condition. As a matter of fact, he was subse- quently used for a number of experiments inside the chamber and showed no idiosyncrasies. The record of body temperature as measured by the electrical resistance thermometer is as follows: 192 Thorne M. Carpenter and Francis G. Benedict. Experiment IT. — Subject, B. F. D., May 5-7. A two-day fasting experiment was planned with this subject, who went into the respiration chamber at 7 p.M., May 5. Nothing abnormal in the experiment was noticed during the night or, indeed, the next day until late in the afternoon, when a slight hacking cough developed. At 10 Pp. M. in the evening there was consider- able pain in the chest, and the subject began to vomit. This continued for an hour or more, and the subject was finally taken out of the chamber at 1 A.M., May 7. He complained of great shortness of breath and diffi- culty of respiration, with a pain in the chest and a persistent nausea. The subject remained in the laboratory too uncomfortable to leave until 6 P. M., _May 7. The pain in his chest and nausea did not disappear completely until during the day of May 8. The subject took his own pulse rate in- side the chamber from time to time, but no material increase was noticed, neither was there any temperature rise in this experiment. The ill effects seem to be wholly confined to respiratory disturbances and to nausea. Experiment IIT. — Subject, R. D. M., May 12. In order to Secure the normal heat production of man at rest, the subject, who was an editorial assistant in connection with the nutrition investigations in progress, entered the respiration chamber at 8 A.M.; the experiment proper began at 8.48 and ended at 2.48 p.m. During the experiment proper practically no ab- normal conditions were noticed, but a few hours after leaving the labora- tory the subject was taken violently ill. At about 4 Pp. M.a cough appeared, with difficulty in breathing. At 9.20 p.m. he vomited and fainted. A physician was called and a stimulant administered at 11.30 p.m. The next morning his lungs were slightly sore. No observations regarding temperature, respiration, or pulse rate were obtained other than those by the attending physician, who found the patient in a state of collapse. Experiment IV.— Subject, H. E. B., October 19-20. A two-day fasting ex- periment with this subject had been planned. He entered the respiration chamber at 7 Pp. M., October 19, the experiment proper beginning at 9g P. M. Although sleeping as well as could be expected under for him abnormal surroundings, he experienced no discomfort until he arose from the bed at 7 A.M. Upon moving about he became dizzy and lost consciousness for a few minutes, falling upon the bed and unconsciously voiding a small amount of urine. When consciousness returned, he felt nauseated and endeavored unsuccessfully to vomit. The feeling of nausea persisted for about one hour, then after coming out of the chamber he began to feel somewhat better. He remained in the calorimeter laboratory until 3 P.M., during which time, as the table shows, there was a marked in- crease in the respiration rate. The subject did not complain of any cough or soreness of the lungs. Body temperature. Time. Pulse. Respiration. Oct. 19-20. Rate per min. Rate per min. °C, 9.00 Pp. uw “4 ae 36.75 11.00 p. um. 21 36.51 12.00 P. M. 22 36.58 1.00 a. M. ~ 23 36.73 2.00 A. M. 26 37.16 3.00 A. M. : 29 37.59 4.00 A.M. oe 33 37.92 5.00 A. M. 56 32 38.13 6.00 a. M. 34 38.30 7.00 A. M. 38.44 8.15 A. M. : 37.39 * 11.00 a. um. aC 37.45 } 2.30 P. M. ely ae ? Sublingual. Experiment V.— Subject, C. F.S., October 24-25. The experiment was planned to investigate the metabolism and energy production during typewriting. A man of about thirty-five years entered the chamber at g A.M. The experiment proper began at 10.18 A. M., ending at 2.18 P. M. At 1 Pp. M. he ate 450 gm. of whole milk and 142 gm. of graham crackers. During the experiment proper there was nothing abnormal noticeable, but three or four hours after leaving the chamber he was taken with nausea and vomiting. Since he had left the laboratory further observations were not obtained. Experiment VI.—H. D. A., November 2-3. A two-day fasting experiment had been planned, and the subject entered thechamber at 3 p.m., Novem- ber 2. The observations were begun at 5 p.m. The subject retired as usual at rz Pp. M. and reported the next morning, when he was called at 7 A. M., that he did-not sleep very well, perspiring freely during the night. On getting up in the morning a cough developed, his pulse rate increased, and the temperature continued to rise. The subject was taken out of the chamber at 2.52 Pp. M. On leaving the chamber a violent cough developed. 194 Thorne M.C arpenter and Francis G. Benedict. The subject was much nauseated, but did not vomit. The respiration was very painful, especially when taking a long breath. The next morn- ing, after a comfortable night at his room, the subject still felt soreness in the chest and difficulty in taking a full breath. Time. Pulse. Respiration. Ra Nov. 2-3. Rate per min. Rate per min. oC. 5.00 P. M. ea an 36.96 7.00 P. M. a ae 37.13 9.00 P. M. a = 36.70 11.00 P. u. 65 16 36.54 1.00 a. mM. 75 15 36.55 2.00 A. M. 71 20 36.75 3.00 A. M. 88 22 37.05 4.00 a. M. 96 22 37.28 5.00 A. M. 88 20 37.40 6.00 A. M. 90 21 37.99 7.00 A. M. 94 21 37.46 8.00 A. Mt. 83 22 37.26 9.00 A. mM. 85 24 37.41 10.00 a. u. 91 24 37.60 11.00 a. u. oe me 37.70 12.00 A. M. 94 28 37.90 1.00 P. m. es 5 38.27 1.15 P. a. i - 37.943 1.30 P. M. 107 se pete 5.30 P. M. 103 ae 39.00 # 7.00 P. M. 100 ao 38.56! 1 Sublingual. Mercurial Poisoning of Men. 195 Experiment VII. — Subject, F. E. S., November 4-5. This experiment was designed to study the effect of passing the oxygen used in the system through a combustion tube and thereby destroy any possible organic poisonous substance. The subject entered the chamber at 8 P. M., ob- servations began at 9 p.M. The subject slept until 2 A. Mm. and woke up with a dull ache in the lower part of the lungs. Went to sleep again, but was called at 7 A.M. On rising he coughed a little and was breathing heavily. He had no appetite, eating but 7 gm. of prepared cereal, 9 gm. of bread, and 31 gm. of cream. As the temperature, pulse, and respiration all continued to increase, the subject was taken out of the chamber at 3 p. M. before more uncomfortable symptoms appeared. Body Pulse. temperature. Respiration. Rate per min. 74 Rate per min. 17 °C. 37.30 37.04 36.70 36.55 5.00 A. M 73 18 36.70 7.00 A. M 99 18 37.15 9.00 A. M : 37.41 i (=) So > B oO WwW i) oo Song 1.00 P.M es i 38.19 2.00 P. M. 102 26 38.37 3.00 P. M. 115 31 38.30 Experiment VIIT.— Subject, G. V. S., November 6-7. While the results of the preceding days were distinctly discouraging to the consummation of a two-day fasting experiment, this subject had planned for such an experi- ment and hence it was begun. The subject entered the chamber at 8 P. M., November 6, observations began at 9 Pp. M., and the subject reported at 3 A. M. a nausea and a cough every time he drew a long breath. He slept very well after 3 a.m. After rising at 7 A. M. he felt better at first, but on moving around felt very weak. Owing to the great discomfort experi- enced by previous subjects in these experiments, it was decided to take him out of the chamber at 7 A. Mm. On coming out of the chamber he was 196 Thorne M. Carpenter and Francis G. Benedict. very much nauseated and vomited. He went to his room and vomited again at 10.30 A.M. He remained in bed all day, and his lungs felt very sore. When he was lying down he could breathe fairly comfortably, but when sitting or standing he felt great pain and immediately began coughing. Body Pulse. Respiration. temperature. Rate per min, Rate per min, oC) 74 19 3159 53 19 37.12 49 16 36.34 52 21 36.43 66 22 36.59 74 24 36.75 72 31 37.04 37.26 37.90 + ? Sublingual. Experiment IX. — Subject, A. H. M., November 8-9. This experiment was planned to attempt to eliminate the possible source of contamination to the air by using oxygen prepared from sodium peroxide. The subject entered at 8 p. M., observations began at 9 p.m. Upon rising at 6 A. M., November g, the subject reported he had slept well. At 10 A.M. there was a sublingual temperature of 38°.56. The subject said he felt well. The temperature as indicated by the electrical resistance thermometer was still rising rapidly. The calorimeter window was opened at 12 noon, much to the surprise of the subject. When he stood up to come out of the chamber, he coughed severely and at the suggestion of one of the physicians he was allowed to lie inside on the cot. The window was opened, although the ventilation system was not running. He had a violent nausea and was coughing most of the time. He came out of the chamber at 4 p.m. Although the exact data regarding his subsequent condition are lacking, he felt well enough to make another experiment four days later. Mercurial Poisoning of Men. 197 Body temperature. Pulse. Respiration. Rate per min. Rate per min. . °C. 36.31 36.32 36.06 5.78" 35.94 36.23 37.15 Si (epl 38.04 38.46 39.27 39.25 39.08 38.76 rho) ly 1 Sublingual. Variations in susceptibility and the personal equation. — During the spring of 1905, when Experiments I-III cited above were being made, there was a continuous series of experiments made with a large number of subjects in which no ill effects either during or after the experiments were observed. In the fall of 1905 there were likewise certain experiments interspersed between those here re- ported in which individuals also appeared immune. The most striking instances are those experiments with C. R. Y., the sub- ject of Experiment I. On September 13 an experiment of twenty- four hours’ duration was made with him with no difficulty whatever. 198 Thorne M. Carpenter and Francis G. Benedict. On October 4 and 5 the same person was used for the subject of an experiment on the digestibility of cheese, but, as he was unable to partake of the diet as planned, the experiment was stopped at eight o’clock the following morning. There were no appearances whatever of any thermal or respiratory disturbances. On October 26-29 a two-day fasting experiment was made with him without difficulty.* An examination of the dates of these experiments shows that while C. R. Y. was made distinctly unwell by an experiment in March, he was not affected by several sojourns in the chamber in the fall, in spite of the fact that, judging from the frequency of the appearance of toxic symptoms in different individuals, the cause had not been removed. On October 12-15, a two-day fasting experiment was made with another subject, H. E. S., and no abnormal results were apparent.® On October 23 an experiment was made with one of us, a rest experiment of four hours’ duration, with no untoward results. On the afternoon of October 24, immediately following Experiment V cited above, an experiment was made with another subject who likewise performed some typewriting. This experiment, which lasted some six hours, was without abnormal results. It is thus apparent that individuals varied considerably in regard to their susceptibility to the toxic influences obtaining in the chamber. In at least one case, C. R. Y., it appeared that he had been affected in one experiment, but had gone through several experiments sub- sequently without discomfort. Suspected sources of toxic effect.— Since this apparatus had been in constant use with almost no difficulties for several years, the appearance of these toxic effects was indeed most puzzling. To our knowledge there had been no material alteration in any step of procedure of the manipulation, and we were entirely at a loss to account for these symptoms. Among the many suspected sources of difficulty were the possi- bilities of arsenic compounds in the acid and the possibility of im- purities in the oxygen. With regard to the acid experiments were made with acid from different sources, and while we had been commonly accustomed to using the highest grades of commercial * F. G. Benepict: Publication No. 77, Carnegie Institution of Washington, 1907, pp. 222-273. ° BENEDICT: Loc. cit., pp. 222-273. [oe “ur Seay Mercurial Poisoning of Men. 199 acid, we made a number of experiments with chemically pure acid, with the same result. The oxygen supply had been obtained from the S. S. White Dental Mfg. Co., perhaps the largest manufacturers of oxygen for medical use in the United States. A visit to their works by one of us resulted in a most careful examination of their method of manu- facture, and no possible sources of contamination were found there. As an added precaution, the oxygen in certain experiments before being admitted to the chamber was passed through a heated glass tube containing copper oxide, and in one experiment oxygen pre- pared over sodium peroxide was used. In all the experiments the same results were obtained. Among the large number of experts consulted on this problem, it was suggested a number of times that there might have been some bacterial infection which, on account of the closed connections of the chamber, might have been communicated from subject to subject. To prevent this as far as possible, the chamber was fumigated with formaldehyde candles repeatedly without avail. Furthermore, it was impossible to conceive of air passing through strong sulphuric acid twice and then through a filter of cotton containing sodium bicarbonate and not have bacteria practically removed. While the possible presence of mercury vapor in the system had always been recognized, it was felt that the surface of mercury exposed to the air current was so small as to almost preclude any possible material volatilization of the mercury, but when other sources of difficulty failed to be detected, it was finally decided to remove the mercury valves at the rear of the system and simultane- ously replace the old piping which conducted the air from the valve to the chamber with new galvanized iron pipe of like size. Temporary changes were made by substituting rubber hose in place of the galvanized iron piping formerly used. The mercury valves were replaced with brass valves commonly used in steam and water piping. The mercury valves at the entrance end of the absorber system were not removed, because here the pressure to be guarded against was very large, greater than at the exit end, and it was assumed that any mercury vapor would be completely removed from the air cur- rent by passing twice through the acid in the porcelain absorbers. 200 Thorne M. Carpenter and Francis G. Benedict. EXPERIMENTS FOLLOWING THE CHANGES IN THE VALVE SYSTEM OF THE RESPIRATION CHAMBER. It is exceedingly fortunate for the success of these experiments that two of the gentlemen who had been seriously affected by their previous sojourn in the respiration apparatus willingly volunteered to subject themselves to the further experiments with the apparatus as altered. It is a great pleasure here to express our appreciation of the high scientific interest that actuated both these gentlemen, Mr. A. H. Middlemass and Mr. H. D. Allen, to expose them- selves to the possibilities of another experience as distressing as they had formerly gone through. Fortunately both these experi- ments were without unpleasant results. Experiment X.— Subject, A. H. M., November 13-14. The subject of the experiment entered the chamber at 2 p. m., November 13; the experiment proper began at 5 P. m., and as no unpleasant conditions developed, the next day, at 12.15 P.M., the experiment ended. During this time the pulse, respiration, and temperature all indicated nothing abnormal. Experiment XI. — Subject, H. D. A., November 14-15. This experiment began two hours after Experiment X ended. ‘The experiment proper began at 3 Pp. M., and the next morning, as no untoward results appeared, the experiment was concluded. Subsequent experiments. — Following Experiments X and XI, the respiration calorimeter was used for a large number of metabolism experiments during the winters of 1905-1906 and 1906-1907. Some 10g experiments with 14 individuals were made, and in no in- stance did the sojourn in the chamber produce any uncomfortable results. The complete disappearance of all feelings of discomfort with the removal of the mercury valves and old piping seems to prove conclusively that these were the chief causes of the toxic symp- toms noticed in the earlier experiments. While we were far from being able to carry out a careful toxicological investigation of this problem, we utilized as much time as we could spare from our other metabolism experiments to attempt to throw more light upon the exact cause of toxic symptoms. Experiments on dogs. — Two experiments were made with dogs in which the dogs were confined in a metal box through which air Mercurial Poisoning of Men. 201 was drawn, and the air before entering the chamber was passed through the suspected mercury valve and piping which had been re- moved from the respiration chamber. In one of these experiments the dog used was very old, somewhat over twenty years, and while he refused to eat during the experiment, he was so very feeble at the conclusion it was decided most humane to chloroform him. It was impossible to draw any deductions from this experiment. A much younger and more active dog lived in this small metal box with air passing through these pipes most of the time for two weeks without any ill effects. It should be stated, however, that the immunity of this particular dog to mercurial poisoning must have been very marked, as for several days it was arranged to have him sleep on a wire screen over a large number of dishes containing mercury. Even under these conditions the dog exhibited absolutely no indications of mercurial poisoning. Experiments on man using the original mercury valve and piping. — A second attempt was made in December, 1906, to secure more definite evidence regarding this case by repeating an experiment with man in the respiration chamber using the mercury valve and piping that had formerly been removed and using one of the subjects, A. H. M., who formerly exhibited toxic symptoms. The subject volunteered to make this experiment. The experiment con- tinued for three days, and, as a matter of fact, on the early morning of the second day there was a marked temperature rise. He did not, however, reach any actual febrile state, and as it did not rise any farther and he himself made no complaint and there being no respiratory disturbances, the experiment continued for some thirty hours after the temperature rise was noticed. The pulse also showed no increase. Unfortunately again this experiment does not leave any clear evidence regarding the case. Test for mercury on the inner walls of the piping. — Since all the experiments pointed towards a gradual absorption of mercury vapor by the layer of zinc inside of the galvanized iron pipe, an attempt was made to demonstrate the presence of mercury. A section of the pipe was placed in the lathe and one chip was taken off for some distance. This material was subjected to tests for mercury, which, however, were unsatisfactory. There was no vis- ible appearance of amalgamation, and the tests for mercury were by no means satisfactory, so that we could not state definitely whether there was mercury present or not. The important point was that there was no appearance of amalgamation on the zinc itself. 202 Thorne M. Carpenter and Francis G. Benedict. CoNCLUSION. While there lacks a definite scientific demonstration that the causes of toxic poisoning during these experiments were due to poisoning by mercurial vapor, there seems to be no other possible conclusion. The examination of the literature shows but very little evidence regarding cases of mercurial poisoning presenting symptoms of the type noticed here, although Rubner ® states that in mercurial poisoning sleep is disturbed and there is fever which can be more or less pronounced. Of special interest, however, in considering the question of poisoning inside the respiration chamber is the discussion of mercurial poisoning brought out by Krogh.” Krogh employed mercury in his admirable apparatus used in his experimental researches on the expiration of free nitrogen from the body. In these researches he found that there was a great mortality in the experiments with eggs, and he also is inclined to believe that the classic experiments of Seegen and Nowak were also complicated by mercurial poisoning. Bing § reports nine cases of poisoning of patients in a hospital due to the escape of steam through a mercury reduction valve. The patients exhibited a temporary rise of temperature, increase of respiration, increase of pulse rate, cyanosis, vomiting, and diarrhea. Two of them (infants) died, but the others recovered in a few days. 8 RuBNER: Lehrbuch der Hygiene, p. 76r. 7 Krocu: Skandinavisches Archiv fiir Physiologie, 1906, xviii, p. 399. 8 Binc: Archiv fiir Hygiene, 1903, xlvi, pp. 200-223 & Piel ne PRELIMINARY OBSERVATIONS ON METABOLISM DURING FEVER. By THORNE M. CARPENTER anp FRANCIS G. BENEDICT. [Nutrition Laboratory, Carnegie Institution of Washington, Boston, Massachusetts.] INTRODUCTION. HE typical picture of fever, presenting as it does high tem- perature, flushed skin, rapid respiration and pulse rate, has led to the assumption that during febrile process there is a marked increase in heat elimination. The temperature rise is fundamentally to be considered as a disturbance of the delicate adjustment between thermogenesis and thermolysis ordinarily obtaining, and accordingly such disturbance can be accounted for in several ways. There may be an increased thermogenesis, or there may be a decreased thermolysis, and finally both factors may exert an influence. In considering the methods used to study metabolism during fever, it is seen that they have been confined to two, namely, direct calorimetry, in which the heat elimination from the whole or a part of the body has been measured, and indirect calorimetry, in which the heat production has been computed from a study of the gaseous exchange. Assuming that the gaseous exchange is a direct measure of the heat production in the body, an assumption that in the light of more recent investigations is very much to be doubted, it is seen that this method can be used only in studying cases in which the temperature rise can be ascribed solely to increased thermogenesis. If the production of heat remains constant and there is a decreased loss of heat, a rise in temperature will occur, but this will not be shown in the method of indirect calorimetry. Obviously, then, if the temperature rise is a resultant of both of the above-mentioned factors, indirect calorimetry can throw but little light on such complicated processes. Since we must rely upon : 203 204 Thorne M. Carpenter and Francis G. Benedict. evidence furnished by direct calorimetry to solve this important problem, any observations involving direct calorimetry which con- tribute to our knowledge of metabolism in the febrile state can reasonably be considered as of value. In connection with an extensive series of investigations with the respiration calorimeter in Wesleyan University, Middletown, Conn., an accidental toxic condition of the ventilating air in the respiration apparatus resulted in a number of marked cases of respiratory disturbance accompanied by fever. All the subjects were men, and usually the experiments were designed to study some special problem in metabolism, such as the influence of inanition and the influence of the ingestion of various kinds of food. The toxicological features of these cases have been collected and pre- sented in the preceding article Since at the time we were using every possible means to eliminate the toxic conditions, but little time was spared for studying the metabolism in the febrile condition, and hence the observations we have to report are more or less fragmentary, and pending a more comprehensive investigation of a number of typical fevers by means of the respiration calorimeter, they are to be looked upon as distinctly preliminary. Inasmuch as these investigations are of so incomplete a nature, it hardly seems desirable in this place to enter into a long review of the literature of the subject of metabolism during fever. Such reviews have been worked out admirably by Kraus,? Krehl,? and especially has the literature been well reviewed by Likhatscheff and Avroroff.4 This last review is unfortunately in S053 and hence not. accessible to most readers.® 1 T. M. CARPENTER and F. G.-BENEpIcT: This journal, 1909, xxiv, p. 187. ? Kraus: Von Noorpen’s Metabolism, Physiology and Pathology, 1907, ii, p. 90, W. T. Keener & Co., Chicago. * KreEHL: Zeitschrift fiir allgemeine Physiologie, 1902, i, p. 20. - * LrkHATSCHEFF and AvrororFr : Investigations of gas and heat exchange in fevers, Reports of the Imperial Military Academy, St. Petersburg, 1902, v, Nos. 3 and 4. * A short account of the original experiments of these authors on a case of malarial fever has been published by them: LikHaTscHEFF et AVROROFF, Comptes rendus de XIII Congress International de Medicin, Paris, 1900. Section, Pathologie générale et pathologie expérimentelle. The original Russian monograph has been translated for use in the Nutrition Laboratory, where the original and the translation are always accessible. Observations on Metabolism during Fever. 205 EXPERIMENTAL PART. The apparatus used in these experiments was a respiration cal- orimeter of the closed circuit type which permitted the simultaneous determination of the carbon dioxide production, water vapor elimi- nation, oxygen absorption, and heat production. While in the course of many years’ experimenting with this apparatus, there had been various personal impressions recorded by the subjects as to their well being while inside the chamber, some having complained of slight headaches, others of temperature changes and minor body discomforts, there were no serious. in- terruptions to the experiments until the spring of 1905, when one or two men were taken with violent nausea after the experiment had ceased. In the fall of 1905 a number of subjects in a short space of time were taken with violent nausea, respiratory disturb- ances, and marked temperature rise while inside of the respira- tion chamber, the discomfort being so great that in most instances it was necessary to remove the subject from the chamber. The subjective impressions of these men, together with a discussion of the toxicological condition in all probability causing this disturb- ance, are to be found in the preceding paper.® Suffice here to state that the cause of the disturbance was in all probability the use of mercury valves in the ventilating system. On the removal of these valves all disturbance ceased, and we have evidently here to do with a rather remarkable instance of mercurial poisoning which presented a very striking picture. Some ten men were thus affected by the mercurial poisoning. Of these, however, the febrile con- dition was reached while inside the chamber with only six, and hence these six alone can be used to furnish fragmentary evidence regarding the metabolism during a febrile state. In order to study the metabolism during fever with these men, it is highly desirable that control experiments should have been made with the same subject under as near like conditions of bodily activity as possible. Unfortunately this could not be done, as other more pressing work had been delayed by reason of the experience with the mercurial poisoning. It transpired, however, that in a few instances we had other experiments with these same subjects which permitted of at least a reasonable comparison, and we also ® T. M. Carpenter and F. G. Benepict: This journal, 1909, xxiv, p. 187. 206 Thorne M. Carpenter and Francis G. Benedict. have from a large number of unpublished experiments selected one or two which permit of reasonable comparison with the other experi- ments, although with different individuals. In making these selec- tions it was our effort in so far as possible to select individuals of the same body weight and general build. In the last analysis, however, we must lay the greatest stress in our comparison of the experiments themselves on the metabolism before and during the state of fever. Some of these experiments were designed especially to aid in solving the problem of the source of the toxic influences in the chamber, no other plans being involved in their arrangement, and consequently the data secured were not obtained with the frequency that would have been desired had it been planned at that time to study metabolism during fever with special reference to the time relations. In all the experiments the subjects were inside the chamber one or two hours before the experiment proper began to become accus- tomed to the surroundings and prepare for the longer experiment. Provisions were made inside the chamber to permit a comfortable night’s sleep, and the subject usually undressed, with the exception of a union suit of underclothing, covered himself with a blanket, and retired at II P.M. Some experiments were made after fasting and some after food, the febrile condition being only an incident and not a part of the plan itself. Usually the subject slept until 7 A. M., when he was called to collect the urine, and the body weight was taken at this time. Body temperature.— The body temperature, which is of the ut- most importance in the study of fever, was determined in all of these experiments by means of an electrical resistance thermom- eter. This thermometer permits the measurement of temperature deep in the rectum to 0.01° C. and, indeed, without any discomfort to the subject, so that continuous observations during the night can be obtained. The apparatus is in constant use in the laboratory and has proven eminently satisfactory. In certain of these experiments recourse was had to the usual clinical thermometer and the sublingual temperature recorded. In connection with these sublingual temperatures it should be stated that the subjects had not engaged in any muscular work and that the temperature of the surrounding air in the chamber remained constant during the whole experiment. There is then no doubt / Observations on Metabolism during Fever. 207 but that the sublingual temperatures recorded were reasonably accurate. Respiration and pulse rate.— In certain of the experiments the subjects were asked to take their own pulse, counting it for two minutes and making their own records. The respiration rate was obtained by means of a pneumograph which was connected, through a metal tube passing the wall of the chamber, with a tambour. This tambour could be used to draw a curve on the kymograph paper, and in some cases the vibrations of the tambour were strik- ing enough without drawing the curve. The respiration movements were easily obtained, and while in some experiments it was possible to obtain the pulse rate by counting the minor vibrations of the tambour, in other instances it could not be obtained with sufficient accuracy. When obtained, it was with great accuracy and consider- able frequency. These observations were likewise made without the knowledge of the subject. Fever Experiment I. — Subject, C. R. Y. The subject, whose body weight without clothing was 67.8 kilos, entered the chamber the evening before and went to bed at the usual time. He woke up two or three times during the night, which was not at all unusual during the first experience in ab- normal surroundings. He arose at 7 A.M. and ate breakfast consisting of 200 gm. of milk, shredded wheat, graham crackers, gluten, cereal and 80 gm. of sugar. The total diet furnished about 4.6 gm. of nitrogen and 502 calories of energy. He lay down again at 7.48 A. M., as the experi- ment was primarily designed to study the influence of the metabolism while the subject was lying down as compared with sitting up. About 1 Pp. M. he drank a little water, but immediately vomited. He was removed from the chamber at 3 P.M.” During this experiment it was possible to + study in two-hour periods throughout practically the whole experiment the metabolism so far as indicated by the carbon dioxide production, water vapor elimination, oxygen consumption, and heat production. The results are given in Table I. The relatively large carbon dioxide elimination and oxygen ab- sorption between 7 A.M. and g A.M. has a natural explanation in that during this period there was unusual muscular activity inci- dental to rising from bed, dressing, collecting the urine, and minor preparations for the day. It was impracticable to secure the pulse or respiration rate. 7 For further details regarding this experiment, see this journal, 1909, XXiv, p. 19I. 208 Thorne M. Carpenter and Francis G. Benedict. LABEL oe METABOLISM OF C. R. Y. DURING FEVER. EXPERIMENT I. (QUANTITIES PER Hour.) Carbon | Water | Oxygen | Respi- Heat Heat Body Period. dioxide | vapor- con- ratory | elimi- pro- | temper- exhaled.| ized. | sumed. | quotient.| nated. | duced. | ature. March 30, 1905. gm. gm. gm. Cal. Cal, oC: 36.41 la.M—3aA.M. .| - 29.4 58.2 22.8 94 78.7 79.6 36.47 Samo 5a .| 229 | 462 | 204. | 0820 ola.) 74mm eee Nee Bey) 44.2 23.6 Hill 73.4 79.2 36.75 Tk. OVALM. 12] SIS 38.0 36.9 74 92.4 O74 36.90 9a.mM—10 A.M. . 90.8 108.0 37.18 31.2 42.9 26.3 87 10 a.mM.-ll a.m. . 92.8 103.8 37.38 11 A. m.—12 m. ! 89.9 111.4 Sy216 j 33.4 40.1 31.3 Li] 2M. -1P.M. . 84.6 112.4 38.23 1 peu = 3PM. «| “S8sl 44.8 31.4 88 79.6 108.0 3989 Fever Experiment IT. — Subject, H. E.B. The subject, whose body weight without clothing was 56.7 kilos, entered the chamber at 7 P. M., October 19, and the observations began at g p.m. After a reasonably comfortable night’s sleep, in which, however, he reports he was awake more or less and felt warm and perspiration was free, he arose at 7 A.M. On rising he became dizzy and unconscious and fainted for afew moments. Nausea continued for about one hour, and although he was removed from the calorimeter chamber, he remained in the laboratory until 3 o’clock in the afternoon. The results of the metabolism are shown in Table II. During the experiment proper the pulse rate could not be ob- tained, but after coming out of the chamber the pulse rate was found to be, at 11 a.M., 89, and, at 2.30 P. M., QO. The data as here tabulated require certain explanations, partic- ularly with reference to the heat elimination and production, which is apparently very large during the period from II P. M. to I2 P.M. when the subject was asleep, while during the next period from 12 to.1 A.M. it falls off enormously. This apparently large heat pro- duction is in part false in that it has been found necessary to make a correction for the amount of heat required to warm up the bed Observations on Metabolism during Fever. 209 and bedding after the subject goes to sleep. At present the very unsatisfactory usage is to assume that 30 calories of heat were required in warming up the bed and bedding to the temperature at which they were during the greater part of the night, conse- quently to the heat measured by the calorimeter we have added 30 calories. This correction was determined on a large number TABLE II. METABOLISM OF H. E. B. DURING FEVER. EXPERIMENT II. (QUANTITIES PER HOvR.) Carbon| Water |Oxygen Heat | Heat Period. dioxide | vapor- | con- elimi- | pro- exhaled.) ized. | sumed. nated. | duced. | Oct. 19-20, 1905. gm. : : : Cal. 9p.mM-—-ll P.M... ll p.m.-12 P.M. . 12p.M-1l1aA.M.. la.M-— 2 A.M... 2A.M— 3 A. 3 aA.M— 4A. 4 A.M 5 A. 5 A.M— 64. 6 A.M— 7 A.M. . of experiments where men completely undressed and had heavier clothing than was used by this man, and consequently the correc- tion is at best tentative. On the other hand, during the period between 7 A. M. and 8 A. M. in the morning there is likewise an addition of heat to that pro- duced during this hour by the amount of heat lost by bed and bedding as they are cooled by the surrounding temperature to that of the room, and it has been our custom to subtract 30 calories from the heat as measured by the calorimeter in the period from 74. M.to8a.M. When the total metabolism for twenty-four hours is involved, this correction simply balances in that there is an addi- tion of heat to the measured amount between I1 Pp. M. and 12 P. M. and there is a deduction from the amount measured between 7 A. M. 210 Thorne M. Carpenter and Francis G. Benedict. and 8 A.M., the correction being only to aid in subdividing the day into periods. In this experiment the heat production from II P.M. to 12 midnight is very much larger than that during the following hours of the night. In explanation of this large heat TABLE III. METABOLISM OF H. D. A. DURING FEVER. EXPERIMENT III. (QUANTITIES PER Hour.) i | Pulse . Carbon| Water! Oxy- | Respi- Heat | Heat Body rate Respi- Period. ae ; gen | ratory] 9). - tem- |. ration dioxide | vapor- ese ae elimi- | pro- are | Del exhaled.| ized. q nated. |duced.| P | min- te DE isumed.| tient. ture. | te, | Minute. Noy, 2-3, 1905, gm. gm. gm, Cal. | Cal. oc. 36.70 9 p.mM—ll P.M. 29:3 | 49:3 | 22:0 | 97 111.8 | 105.4 | 36.54 65 15 llp.m— la.. 29.3 | 49.1 | 26.7 | .80 86.7 | 86.2 | 36.55] 70 14 La.M— 2 A.M. ( 79.7 | 90.3 | 36.75 | 74 18 A.M.— 3 A.M. ! 37.05 21 37.28 23 -M.- 4 A.M. .M— 5 A.M. 37.40 23 37.55 19 37.46 A.M.— 8 A. 37.26 .M—- 9A. 37.41 .M.—10 A.M. 37.60 10 A.M.—11 a. 37.70 11 A.m.-12 M. 37.90 | | et l J l ) 12m. -— 1pP.M. 38.27 1 p.M.-1.52 P.M. Be aA a 38.46 production it should be stated that in this experiment the subject did not make up his bed and undress until some time after II P. M., and hence the extra heat may have been because of this extraneous muscular activity during a period in which subjects are usually resting quietly in bed. Fever Experiment ITT. — Subject, H.D. A. The subject, whose body weight without clothing was 65.4 kilos, entered the chamber on the afternoon of Observations on Metabolism during Fever. 211 November 2. The observations extended from 9 Pp. m., November 2, to 1.52 Pp. M., November 3. Shortly after entering the chamber the subject defecated after an enema. He spent the afternoon and evening in sitting quietly reading, although between 9 and 11 p. M. he made a dozen tests with the hand dynamometer. He went to bed shortly after rz p.m. and did not spend a particularly quiet night, as he felt somewhat warm and woke up rather early in the morning. After getting up at 7 A. M. he spent most of the forenoon lying down, reading and sleeping a part of the time. The subject was much nauseated during the day, but did not vomit. The results of his metabolism are given in Table III. In addition to the pulse records above noted it was found at 5-30 p. M., November 3, on coming out of the chamber that the pulse was 103, and, at 7 P. M., 100. The large heat production from 9 P.M. to II P.M. is probably in great part accounted for by making strength tests with the hand dynamometer. TABLE IV. METABOLISM'OF F. E. S. DURING FEVER. EXPERIMENT IV. (QUANTITIES PER HOvR.) } ; Respi- Carbon | Water| ~* | Heat | Heat ration Period. dioxide | vapor-) ’| elim- | pro- rate ‘exhaled. ized. | JNO" |inated.| duced. ato per | ; ; ; minute. Nov. 4-5, 1905. | gm. gm. | 3 i5 ; oC, 37.30 9 p.m.—ll P.M. é .2 | 37.04 17 llp.m—la.m. | : .9 | 36.70 18 1aA.M— 3 A.M. : a) 36:55 16 Seu San. |( 0| 64.1 | 36.70 17 5 A.M.— 7 A.M. : 37.5 18 7 AM— 9 A.M. | : 8 | 37.41 18 9a.mM—l1l A.M. , 4 | 38.19 11 A.a— 1 Pa. 9 | 108.4 | 38.19 lp. 2PM. rel s ‘5 | 38.37 2P.M—3P.M. | : 9 | 38.30 | 109 Fever Experiment IV.— Subject, F. E. S. The subject, whose body weight without clothing was 56.6 kilos, entered the chamber during the evening 212 Thorne M. Carpenter and Francis G. Benedict. of November 4, and the observations continued from g P. M., November 4, until 3 p.m., November 5. After the preliminary weighing of the man with his bed clothing, etc., he sat reading from 9.30 until bedtime at 11 p.M. He awoke at 2 A. M., but went to sleep again. He arose at 7 A. M. and ate a light breakfast consisting of cream, cereal, bread and butter, containing 0.4 gm. nitrogen and 143 calories. Most of the forenoon was spent in reading until 12.43, after which he lay down and was probably asleep until 2.30 Pp. M., when he arose and sat reading until the end of the experiment. The results of the metabolism are given in Table IV. The respiratory gases were studied in this experiment in long periods, since the special object of the experiment was to test the effect of the passing of the oxygen used in the experiment over heated copper oxide and thus oxidize any possible poisonous ingre- dients. The records of the heat measurements, however, are taken so frequently that it is possible to apportion the heat elimination and heat production with reasonable accuracy into two-hour periods, although it is necessary in this apportionment to assume that the amounts of vaporized water per hour remained constant throughout the whole period. Thus in the long period from 9 Pp. M., Novem- ber 4, to g A. M., November 5, there were 42.5 gm. of water vapor- ized per hour. The amount of water vaporized was taken into consideration in computing the heat elimination, as the evaporation of 1 gm. of water at 20° required 0.586 calories. Since the experi- ment was in two long periods and there was a very close agreement in amounts of water vaporized per hour, this assumption is prob- ably not far out of the way. Fever Experiment V.— Subject, G. V. S. The subject, whose body weight without clothing was 56.0 kilos, entered the chamber at 8 Pp. m., Novem- ber 6, and the observations were continued from g Pp. M., November 6, to 6.52 A.M., November 7, when it was necessary to take him out of the chamber. He reported that he slept well until 3 A. M., but experienced nausea with coughing afterwards and had but little sleep during the rest of his sojourn in the calorimeter. The absence of a marked temperature rise in this experiment may throw doubt on the legitimacy of including it in this discussion of fever, but the rapidly increasing pulse and respiration rates justify the assumption that the body temperature was ‘inclined to rise, as at 11 A.M., November 7, the temperature sublingual was 37-90. Observations on Metabolism during Fever. 213 TABLE V. METABOLISM OF G. V. S. DURING FEVER. EXPERIMENT V. (QUANTITIES PER HovR.) Body tem- pera- ture. Carbon! Water| ~*)” Heat | Heat Period. dioxide elimi- | pro- exhaled.| ized. nated. |duced. Nov, 6-7, 1905. a : : Cal. Cal. oC, 37.59 9p.m—ll p.m. 110.9 | 99.3 | 37.12 ll p.m- la.M. 83.9 | 63.4 | 36.34 1A.M.— 3 A.M. 62.9 | 64.5 | 36.43 3A.M.- 44.M. F : : ; 73.4 | 80.7 | 36.59 4a.M-— 5 A.M. 57.3 | 64:6 | 36.75 5A.M— 6A.M. 82.5 | 96.2 | 37.04 6 A.M.— 6.52 A.M. 65.2| 74.8 | 37.26 | The pulse rate at 11 aA. m., November 7, four hours after coming out of the chamber, was 90. Fever Experiment VI. — Subject, A. H. M. The subject, whose body weight without clothing was 62.8 kilos, entered the chamber at 8 p. m., Novem- ber 8, and observations continued from g Pp. M. until 12 noon on Novem- ber 9. The subject reported an excellent night’s sleep, but was very drowsy on the evening before. He ate breakfast consisting of cream, ce- real, and milk with an estimated nitrogen content of about 3 gm. and 790 calories of energy. Most of the forenoon was spent lying down on the bed, reading or sleeping. The results of his metabolism are given in Table VI. In this experiment, as in the preceding one, the special object being to test the toxic influences inside the chamber, the gaseous exchange was studied only in long periods, and the same method of apportionment of the heat of vaporization of water in the com- putation of the heat production and elimination was used in this experiment as in Experiment V. EXPERIMENTS FOR COMPARISON. Two of these subjects, H. D. A. and A. H. M., who were made wretchedly ill by mercurial vapor in the air current, had sufficient 214 Thorne M. Carpenter and Francis G. Benedict. interest in the success of the respiration chamber to volunteer to be the subjects of experiments made after the removal of the mercury valve, although at that time it was by no means certain that the removal of the valves would insure the absence of the METABOLISM OF A. H. M. DURING FEVER. EXPERIMENT VI. Period. Nov. 8-9, 1905. 9 pm Ill p.m. 1 1 ata SMS 3 am-5 5.48 Am.— 7 7 ame 8 8 am 9 12 ae 1 pm-2 2 pm 3 3 pm 4 A.M. 5 AM 5.48 A.M.| A.M. 9 am-l10 Am. 10 am -ll aA. 1k AM.-12 M. Carbon dioxide ex- haled. gm. (em. SCAB TEES Vale Oxy- gen con- sumed. gm- - | Respi-| Heat | Heat ratory] j-- | ane elimi- | pro- q nated. duced. tient. | Cal. Cal. 102.3 | 102.0 72.6 YES) 65.0 49.2 59/47) 63.3 67.1) 86.4 83.2 | 124.6 86.6| 94.8 76.2 | 115.5 68.6 | 91.0 9 62.3 | 105.9 89.7 | ¢ 2 . 88.1 Resp. (QUANTITIES PER Hour.) rate per min. toxicological symptoms. Consequently, with these two experiments we have direct control. One of the other subjects, C. R. Y., was also used in another experiment inside the chamber, and the results of this experiment can be used with a certain amount of accuracy The remaining experiments must unfortunately be compared with experiments made on other individuals of like body weight and general build. This last method of comparison is ad- for comparison. Observations on Metabolism during Fever. 215 mittedly very unsatisfactory, but it is possible to predict with consid- erable accuracy the metabolism from people of similar age, weight, and general build, provided there is like muscular activity in both instances. Unfortunately, again, in these comparative experiments it was in most instances impossible to exactly duplicate the mus- cular activity in the febrile experiments. On the other hand, since the fever subjects were lying down most of the time, the muscular work in the control experiments was almost invariably somewhat greater than in the fever experiments, and this fact is of unusual interest, as will be seen in the general discussion. CoNTROL EXPERIMENTS. Control Experiment I. — Subject, C. R. Y., October 27, 1905. In this experi- ment the subject was undergoing a two-days fast, which he completed successfully with no abnormal indications. As a matter of fact, for a few days before this experiment and for two or three days after this experi- ment, the respiration calorimeter was used for experiments with other subjects in which they showed marked toxic symptoms as a result of this mercurial poisoning; but this subject who six months before had also given strong evidence of mercurial poisoning, here passed through the experiment without the slightest difficulty. For purposes of comparison, only that portion of the experiment which covers the same time of day as is covered in Experiment I is here presented. The metabolism during this experiment was studied in considerable detail, and the experiment as a whole is described completely elsewhere.* The subject went to sleep at the usual time, but said he did not sleep very well, waking up occasionally; rose at 7 A. M. and dressed and spent practically the rest of the morning sitting quietly reading. About 12 noon, he lay down on the bed and read lying down. Then in the afternoon until 3 o’clock he was lying down on the bed reading and probably was asleep from 2 o’clock until 3. The body weight without clothing at the time of this experiment was 67.1 kilos. So much of the data regarding the metab- olism studied with this man as are required for comparison with the fever experiments are given in Table VII herewith. These data may be taken as indicating the metabolism of this man under normal conditions. The last food was eaten at 6 P. M. on October 26. 8 F. G. Benepicr: Publication No. 77 of the Carnegie Institution of Washington. 216 Thorne M. Carpenter and Francis G. Benedict. The usual increase in heat production attendant upon rising, ad- justing bed and clothes for the day is noticed between the hours OI 77h. Mi tor AYRE: The time covered by the preliminary night and a portion of the day’s fast agrees with the time covered by the experiment showing TABLE VII. METABOLISM OF C. R. Y. CONTROL EXPERIMENT I. (QUANTITIES PER HOvR.) Waves | Respi- Carbon | V | es Heat | Heat | j ration Period. dioxide | v. | elimi- | pro- sons rate exhaled.| ized. nated. duced. P per 3 | | ture. s | | min. Oct. 27, 1905, : : Sarl | Cal. Soe | 36.44 Urea ara : Dil 74.5 72.4 | 36.39 eye 5 | 75.0| 79.6 | 36.57 5 A.M-— 7 A.M. ; ee =e 56.0 | 63.7 | 36.86 7 AM— 9 A.M. 3 | 35. ; 104.8 }107.2 | 37.03 | © 36.89 36.98 9 a.M.—10 A.M. 87.6 85.8 10 aA.m.—1l A.M. dt Ana 10 ye ee | (, 36.97 | 77.4| 71.14 | {| 36.78 12m. —1pP.mM.. 1pM- 3PM. 9. | 197.) 67.1| 66.7 | 36.83 fever on March 30, 1905. The activity for the night was very much the same in the two experiments compared. The activity for the following day was different for the two experiments, but it is thought that the difference in activity will be offset by the fact that during the experiment showing fever the man ate a certain amount of food (energy of this food was 502.3 calories and the nitrogen, 4.16 gm.). Following the eating of this food in the experiment showing fever the man was lying down most of the time, whereas during the day of October 27, which is used for comparison, the man was sitting up and there was some activity until he lay down later in the day. Control Experiments IT and IIIT. — Subject, H. R. D., December 4-5, 1905, and May 9-10, 1906. Data for portions of these two experiments have Observations on Metabolism during Fever. 257 been selected for comparison with the fever experiment with H. E. B. made on October 19-20, 1905. The two men were of about the same body weight. H. E. B. weighed 60.7 kilos with clothing. The weight of H. R. D. on December 4-5 was about 60 kilos, and, on May 9-10, 61.1 kilos. The data for all three experiments are for the night, when the men were asleep most of the time. Neither subject slept all night, sleep for TABLE VIII. METABOLISM OF H. R. D. Controt ExPERIMENT II. (QUANTITIES PER Hovr.) : Respi- Carbon| W pa ration Period. dioxide |v ? | elimi rate exhaled.| ized. q ; per ; ; min. Dec. 4-5, 1905. gm. 1l p.m.-12 p.m. 24.7 12 p.mM— 1a.M. la.M-— 2 A.M 18.3 2A.M—- 3 A.M 3 A.M 4 A.M. 20.1 4a.M— 5 A.M. 5a.M— 6 A.M 21.3 6 A.M— 7 A.M. H. R. D. on December 4-5 being perhaps somewhat better than for his other night and than that obtained by H. E. B. The two nights, Octo- ber 19-20 with H. E. B. and December 4-5 with H. R. D., were in preparation for a regular fast for a day or two, while the night of May 9-10 with H. R. D. was a continuation of an experiment already under way, the subject having eaten at 9.10 A. M. on May 9g about 100 gm. of gluten bread and 220 gm. of skim milk, the energy being 621 calories and the nitrogen 15.39 gm. Except for the quality of sleep the conditions of the three nights were about the same with the usual preparations at bed- time. : In Control Experiment II the subject went to bed at rr Pp. a., December 4, and slept very well, waking for a few minutes at 4 A.M. At 6.30 he woke up but lay quietly in bed until called at 7 A. M. The metabolism during this experiment is given in Table VIII herewith. In Control Experiment III the subject went to bed at 9.45 P. M., May 9, 218 Thorne M. Carpenter and Francis G. Benedict. 1906. He did not sleep as well as in Control Experiment II, owing to a sense of discomfort attending the inserting of the rectal thermometer. Usually the subjects do not notice the presence of the thermometer five minutes after it has been inserted. The data for the metabolism are given in Table IX. TABLE IX. METABOLISM OF H. R. D. CoNnTROL EXPERIMENT III. (QUANTITIES PER Hour.) Car- | . ' bon |Water etek ae Heat | Heat Period. dioxide}| vapor-| © | ) acre | ly. | con--| quo- ex- ized. haled. | | | Body | Pulse oar | tem- rate elimi- | pro- | ata~ lanes : | nated.) duced.) P iP sumed.) tient. | ture. | min. | May 9-10, 1906. . | | Wem: 9.30 p.at.-11.30 P.a. we sé 9| 53.1 | 36.30| 52 11.30 p.m.—12.30 p.m. | | 36. | 51 12.30 pat 1.30 a.m. Re PU sesmaiess 1.30 a.M.— 2.30 A.M. | : 52 2.30 a.at.— 3.30 a.m. | ae 41) 53 3.30 a.m— 4.30 A.M. | | 36.36| 53 4.30 am.— 5.30 A.M. | = .39| 54 5.30 A.m.— 6.30 A.M. | 36.65 | 56 6.30 a.M.— 7.30 A.M.| loos | 36. | 66 This experiment began on the half hour, so the subject did not get up until called at 7.30 a. M., while in the other experiments the subjects usually rose at 7 A.M. Obviously the metabolism is in no wise affected, so far as its use for comparison is concerned, by this schedule. These two experiments are also to be compared with the fever experiment with G. V. S., November 6-7, 1905. H. R. D. and G. V. S. were men of about the same body weight. G. V. S. weighed 60 kilos, and H. R. D. on December 4-5 weighed about 60 kilos and, on May g—10, 61.1 kilos. The experiment with G. V. S. was a night of sleep in preparation for a fast. He slept well until 3 A. M., when he had trouble with his stomach and coughed and did not sleep much afterwards. The experiments, December 4-5, 1905, and May 9-10, 1906, with H. R. D. show somewhat the same activity Observations on Metabolism during Fever. 219 conditions, the one for December 4—5 perhaps a little nearer to the conditions of the experiment with G. V. S., November 6-7, in that H. R. D. on that night awoke about the same time in the early morn- ing that G. V. S. did and his sleep was likewise broken afterwards. Control Experiment IV. — Subject, H. R. D., April 20-21, 1906. The data given here are from an experiment made on April 20-21, 1906, when the subject fasted during the day of April 20, made his preparations for bed TABLE X. METABOLISM OF H. R. D. ConTRoL EXPERIMENT IV. (QUANTITIES PER Hovr.) Car- bon | Water Oxy- | Body | Pulse Period. dioxide|vapor-| 8°" elimi- | pro- | ase pees 3 con- era- er ex- ized. nated. |duced.| JES sumed. ture. | min. haled. Heat | Heat April 20-21, 1906. gm gm. ; Cal. ; oG. | | 36.08 | 9.00 p.ar.—11 p.m. 22.7 | 22:8 F 74.7 a5 56 11.00r.a.- las. | 21.0) 22.2 Fi N G46 )| || ieee oe 100am-3am. | 20.9 | 21.3 SM naa a CY 3.00am— Sam. | 21.7| 218 HB) eh ORO: 60 5.00 A.M.— 7 A.M. DR Sy Ni OPA : 67.9 | 75.9 67 7.00 Am.-8.10 a.m. | 36.9 | 26.9 les 94.7 |123.7 | 85 8.10 a.x-10.10 axc| 31.1 | 26.4 ike 92.6 90.2 79 10.10 aA.mw.—12.10 p.m.| 34.0 | 27.4 .; 92.4 | 92.9 82 12.10 p.s.— 1.10 p.m. | 83 77 ( 78 29.1 | 24.6 : : : ee a 3222) | 2616 : 4 89.7 | 84.8 1.10 p.m.— 2.10 p.m. 2.10 p.m.— 3.10 p.m. | 3.10 p.m— 4.10 p.m at the usual time, and slept during the night until about 6.30 A.M. on April 21. After 7 A. M. he ate 1171 gm. of bananas and 103 gm. of sugar, the energy being 1581 calories and the nitrogen 2.10 gm. After eating his breakfast he sat idle or reading during the remainder of the time. These data are used for comparison with data obtained in the experiment with H. E. S. on November 4-5, 1905, in which the subject went to bed at the usual time. The subject reported that he was awake at 2 A.M., but went to sleep again. In the morning he got up at 7 o’clock and in the next 220 Thorne M. Carpenter and Francis G. Benedict. hour ate 308 gm. of cream, 7 gm. of breakfast cereal, 9 gm. of bread, and 4 gm. of butter, the energy being about 143 calories and the nitrogen about ; 0.40 gm. After breakfast the subject was weighed, as were also the chair and clothes. He then walked from to.12 until 10.52 A.M. He was.more ’ or less active until about 12.45 P. M., when he lay down and must have \ slept for an hour or so. The experiment of April 20-21 with H. R. D. has been chosen for comparison with the experiment on November 4-5 ’ with H. E. S., because the men were of about the same body weight. | H. E. S. weighed 60.7 kilos and H. R. D. weighed 62.5. Both weights were with clothing. H.R. D. ate considerably more food, and the energy , and nitrogen of the food were considerably greater than eaten in the H. E.S. | experiment. H.R. D. was not nearly so active immediately following the eating of the food as was H. E. S., who weighed himself and chair and clothes and was otherwise more or less active, besides walking about for 40 minutes during the two hours succeeding the breakfast period. The data for the metabolism are given in Table X herewith. Owing to a defective connection, the electrical rectal thermometer did not give readings during the early part of the night, and hence the heat production cannot be accurately computed. It is evident that the changes in body temperature must have been small during this time, and hence the heat elimination is probably a very close index of the true heat production. Control Experiment V.— Subject, H. D. A., November 14-15, 1905. The data included under this head are for the afternoon and night of Novem- ber 14 and the morning of November 15 to compare with the correspond- ing period of a fever experiment made on November 2-3 with the same subject. The subject ate supper at about the accustomed time in the calorimeter on November 2. After supper he sat reading during the even- ing and then prepared to retire at the usual time and slept until morning, but not very well. The morning following the night of sleep was spent in a certain amount of activity for a couple of hours in the midst of which he slept perhaps 20 minutes. The remainder of the forenoon he was lying down mostly and slept into the early afternoon. At 6 P.M. on November 14 the subject ate supper consisting of 60 gm. of shredded wheat and 245.4 gm. of cream; the energy being about 739 calories and the nitrogen about 2.00 gm. Following supper, he spent the evening sitting reading, writing, and some studying. Then he prepared for bed at the usual time and retired and, so far as known, slept well until morning. The forenoon following his night of sleep was spent with a certain amount Observations on Metabolism during Fever. 221 of activity for an hour or so, and then the subject lay on the bed most of the remainder of the time. The metabolism in the control experiment is given in Table XI herewith. TABLE XI. METABOLISM OF H. D. A. ConTROL EXPERIMENT V. (QUANTITIES PER Hour.) f ee Respi- Oxy- | Respi- Heat ration en ratory 8 > pro- rate con- uo- q duced. per sumed.| tient. Bat Carbon |Water Period. dioxide | vapor- exhaled.} ized. Nov, 14-15, 1905- : 5 gm. 3 p.M— 5 P.M. 5 p.M— 7 P.M. 7 P.M— 9 P.M. 9 p.m-ll P.M. ll p.m- 144. LlamM- 2A. 2A.M-— 34. 3 a.M-— 4A. 4a.M—5 A.M 5 A.M. 64. 6 A.M.— 7 A. 7AM — 8 A. 8a.M- 9A. 9 A.M—10 A. 10 A.m.-ll a. The pneumograph failed to operate satisfactorily after 5 A. M., so the pulse and respiration rates are wanting. Control Experiment VI. — Subject, A. H. M., November 13-14, 1905. The subject entered the chamber at 4 P. M. and at 6 ate supper consisting of bread, 190.2 gm.; cheese, 40.3 gm.; milk, 256 gm.; and butter, 12.7 gm.; the energy being about 1031 calories and the nitrogen about 5.78 gm. Following his supper the subject spent the evening reading or sitting idly, and becoming drowsy he prepared for bed and probably went to sleep i) i) bo Thorne M. Carpenter and Francis G. Benedict. immediately after 11 o’clock and had a good night’s sleep. In the morn- ing he arose at 7 o’clock and ate 101 gm. of bananas, 226.5 gm. of milk, and 86.8 gm. of bread for his breakfast. The energy of the food was about 522.1 calories and the nitrogen about 2.69 gm. After breakfast he was lying down until noontime except that he got up occasionally. TABLE XII. METABOLISM OF A. H. M. CONTROL EXPERIMENT VI. (QUANTITIES PER HOUR.) Respi- ration q Se rate ' elimi- ae rate nated. | Ws per. ‘ min. : ' min. . | | ] Carbon) Water| Ore ISes0F Heat wie: gen- | ratory Period. | dioxide | vapor- : con- | quo- exhaled.) ized. | : isumed.| tient. | Nov. 1415, 1905. te (| : ° Cal. 5p.M.— 7 P.M. : 104.7 : 59 7P.M.— 9 P.M. : 101.7 9 p.m.—11 P.M. : 87.7 llp.m.— 1 A.M. : 75.4 la.m.— 3 A.M. 3 54.5 3AM- 5 AM. 5 AM.— 6AM. 6AM.— 7 AM. 7AM— 8 A.M. 8 AM.— 9 A.M. 9 A.M—10 A.M. 10 A.m.—1l A.M. 11 a.m.—12.15 p.m. This experiment is used for comparison with an experiment made with the same subject on November 8-9, 1905, during which the subject spent the evening of November 8 in reading or lying on the bed following supper (amount eaten unknown). He adjusted his bed to lie down about half an hour before bedtime, then prepared for bed and retired at 11 o’clock. He said he slept well all night. Following his breakfast he lay down and was lying down most of the time until he was taken from the chamber along in the afternoon. He slept some of the time, was up occasionally, and later he coughed a great deal. The activity during the two experi- Observations on Metabolism during Fever. 223 ments was much the same, except that during the forenoon of November 14 the subject did not sleep, but on the other hand he was not quite so active as he was when awake during the morning of November 9. The metabolism in this experiment is given in Table XII herewith. In this experiment of November 13-14, unfortunately the rectal thermometer did not operate properly, and it was impossible to get accurate temperature observations between 7 Pp. M. and 9 A. M., and hence the heat production cannot be satisfactorily computed. The probable difference between heat production and heat elimination may, however, be reasonably estimated when the regular curve for body temperature is taken into consideration. Usually there is a marked fall in temperature immediately after the subject goes to bed, persisting for an hour or more. The result would be that during this period there would be a loss of heat from the body and apparently a greater heat elimination, and the heat production would be some- what smaller than the heat elimination. On the other hand, begin- ning with about 5 o'clock in the morning, there is a noticeable tem- perature rise, which increases rapidly after the subject gets out of bed, so that during the periods from 5 to 8 the heat elimination is somewhat less than the heat production, as part of the produced heat is used to warm the body. Under the conditions obtaining inside the respiration chamber when the muscular activity is very slight, these temperature fluctuations are not very noticeable, and hence for purposes of comparison it is reasonable to assume that the heat production does not differ greatly from the heat elimina- tion. Asa matter of fact, in two other experiments made with this same subject, covering the same periods of time, the heat pro- duction did not differ from the heat elimination more than 4 calories in any period, thus substantiating this view. DISCUSSION OF RESULTS. The most striking feature regarding these experiments is the marked and rapid temperature rise in certain of them. This was in almost every instance accompanied by a marked increase in the respiration rate. On the other hand, in some experiments, namely, with G. V. S., where the temperature rise was not very marked, the respiration rate increased noticeably, thus indicating a respira- tory disturbance. 224 Thorne M. Carpenter and Francis G. Benedict. -The pulse rate also increased during the febrile stage. Un- fortunately the pulse rate could not be obtained in the first and second experiments, as the technique was not perfected at that time for securing this information. Carbon dioxide production. —The carbon dioxide production is commonly considered as a general index of the total metabolism, although obviously with variations in the proportions of fat and carbohydrates burned, the energy accompanying this combustion will vary per gram of carbon dioxide. However, as a general index of metabolism, the carbon dioxide may still be taken as fairly accurate. In these experiments the carbon dioxide determi- nations in both fever and control experiments have been combined in Table XIII for purpose of comparison. Bearing in mind the diffi- culties of comparing experiments of this nature, particularly when the control experiments must be made with a subject other than that used in the fever experiments, the evidence is still very strik- ing to show that the carbon dioxide per hour during fever is some- what greater than that during the control. With C. R. Y. this is shown in practically all periods in which there was a marked tem- perature increase. Unfortunately the long periods of observation in the experiments with F. E. S., G. V. S., and A. H. M> couldtnee permit of the careful apportionment over the short periods as do the others, but nevertheless the carbon dioxide excretion is appar- ently greater during the fever than during the control. Oxygen consumption. — While there are variations in the amount of energy per gram of carbon dioxide depending upon whether the combustion is of fats or carbohydrates, these differences disappear in large part when measurements of oxygen are taken into consid- eration, as the number of calories per gram of oxygen is not widely different whether the substance burned be fat, carbohydrate, or protein. Accordingly we would expect to find in experiments of this type the oxygen consumption as especially indicative of the total metabolism. Before considering the figures for these ex- periments in both fever and control, it is necessary to bear in mind that the determinations of oxygen with the respiration apparatus here used, while extremely accurate for experiments of twenty-four hours’ duration, are by no means so accurate for shorter periods. The large volume of residual air, some 4500 litres, is subject to considerable temperature variations, particularly when there are differences in the bodily activity at the beginning and end of the ISS Observations on Metabolism during Fever. 225 experimental period. In determining the oxygen according to this method, the absolute volume of gas inside the respiration chamber TABLE XIII. CARBON DIOXIDE PER: HOUR DURING FEVER AND CONTROL. La.m.| 3 A.M. Subject. to to 3 A.M.] 5 A.M. 9 A.M. to 11 A.M. 11 A.M.) 1 P. a.) 3 P.M. to fo to | lep.m.|/3P.M./5 P.M. gm. gm. Gan Y-rever lt <>.) 29:4 22.9 CeRoy. Control i =i 23:8 25a! ob. BFever2...| 26.8 30.2 H.R.D. Control II .) 18.3 20.1 gm. 31.2 29.4 gm gm gm. 33.4 | 38.1 27.6 | 24.5 meDrAaKeverS <. -| 27.7 30.0 ie DSA Control If.) 25.3 27.4 SFO F.E.S. Fever4 .. 28.2 H. R. D. Conirol IV. “aad Dies | 2335 SSS —— GaVeS. Fever5 < .- 25.6 Ee): Control IL =) 18:3 | 20.1 | 21.3 SSS ee A: H.M. Fever 6. .- 25.9 ———_———— A. H. M. Control VI 17.4 20.1 22.8 35.9 IWoyerzie = ~< 17.4 19.0 20.0 MiaregZssur 1. Sieh 30.3 ? 29.6 26.55 34.0 24.3% 25.4 17.00 a.m.—8.10 a.m. The remaining figures for this experiment are for two-hour periods following 8.10 A.M. 2 9.00-12.00 noon. 3 11.00 a.m.—12.15 P.M. J * Control experiment made with same subject and introduced for further comparison. 5 8.30 a.m.—9.30 A.M. - must be known with great accuracy, and hence the temperature fluc- tuations can influence the accuracy of this calculation in a marked degree. This feature of the determination of oxygen consump- 226 Thorne M. Carpenter and Francis G. Benedict. TTA BIE: SSlivic OxyYGEN CONSUMED PER HourR IN FEVER AND CONTROL. Subject. to to to to to to to to gm CAR i yetever le fe E2228 20.4 | 23.6 36.9 26.3 31-32) 3184s eae Cun. YaiControl as =|)! 20:2 19.6 16.5 35.4 22.0 23.2 | (19s sa H. H. EE. FE. H. G. Ee H. A. A. E. B. Fever2 . .| 20.5 21.8 | 24.6 BPS ae Boe PAP ates R. D. Control II .| 16.2 ZAROY | elOss sae ee See bore alec R. D. Control III) 21.1 | 19.4 | 1724 eee ay et ; | ae D. A.-Fever3.-.| 24.8 | 27.4 | 264 | 38:8 | 28:9 | 20.¢.q)\eeee iD) AniControleVe 4|00 230m: aoe | 27:8 | 400 | 300 | ...) "eee E.S. Fever4 . .| 2202. 34.1 R.D.ControlIv | “Te4 ) 210) 240) 410:| 274) 283 / 230 | 216 | =. SS = VAS sbever san] 22.6 | ew Re DsControliil.|) 6:2 21.0 16.3 aa ae rhs ey SE Sad RD. Tl eit | oa!) Iza) 2 6) a : a | H. M. Fever6. .| 22.9 27.9? —_——— OO "| H.M.Control VI | 20.2 19.8 25.3 36.6 26.3 ZAM ne = Sec Nov. 214 ‘ ‘< 14.1 15.5 16.1 Bere =< sare SOC sé Mari 23s “ Be a sete sare 23.21) @20!0 Meier | S50 1 7.00 A.m.-8.10 a.m. The remaining figures for this experiment are for two-hour periods following 8.10 a.m. 2 9.00-12 noon. 3 11 a.m.—12.15 p.m. * Control experiment made with same subject and introduced for further comparison. 5 8.30 a.m.—9.30 A.M. 1 A. M. SiN NeUItS Aaa: 7A.M.|9aA.mM.}/11 A.M.) 1 P.u.| 3 P. wu 3 A.M. S A.M.) 7 A.M. 9Qa.m.|1la.m.)] 1 p.m. |3P.M.)5 P. um. Observations on Metabolism during Fever. 227 tion with this apparatus has been discussed at considerable length elsewhere.® In the particular experiments here under discussion, however, the bodily activity aside from going to bed at night and getting up in the morning was in the majority of the periods reasonably constant. It is possible, therefore, in these experiments to use the data for the oxygen with a reasonable feeling of confidence that they are fairly typical of the exact oxygen consumption during the periods under discussion. The data for oxygen absorption both during fever and control have been compiled in Table XIV for comparison. The oxygen consumption during fever is in practically all cases noticeably greater than during control. The only exception to this is in the case of the experiment with H. D. A. Respiratory quotient. — It is conceivable that during fever there may be a greater draft upon previously stored glycogen, and hence one might expect a variation in the respiratory quotient to corre- spond to this increase in carbohydrate burned during the febrile state. As a matter of fact, the respiratory quotients as determined in these experiments have not been tabulated, especially as it is believed that the possible errors in the oxygen determination would render any deductions from them liable to error. They have been given in connection with each experiment, and while the data show a slight tendency for the respiratory quotient to increase during fever, the complications attending the ingestion of food, variations in muscular activity, and errors in oxygen determination do not warrant any sweeping deductions from these data. Water vaporized. — The special significance of the determination of water vapor in these experiments was to add to the heat brought away by the water current of the respiration calorimeter the heat of vaporization of water, as this amounts to 0.586 calories per gram of water. Hence in these experiments no particular study was made of water vaporized other than for this purpose. The data have been collected and summarized in Table XV herewith. The figures show that in general there was an increase in the water of vaporization during fever over that during the control period. Since, however, the control experiments showed marked ® F. G. Benepicr: Publication No. 77 of the Carnegie Institution of Washing- ton, p. 451, and F. G. Benepict and R. D. Mizner: Bulletin 175 of the Office of Experiment Stations, U. S. Department of Agriculture, pp. 28-30. 228 Thorne M. Carpenter and Francis G. Benedict. variations when compared with the fever experiments during peri- ods when there was no appreciable fever, it is obvious that here again we cannot draw any sweeping deductions regarding this point. TABLE XV. WATER VAPORIZED PER HOUR IN FEVER AND CONTROL. 9a.m.|1l a.m. Subject. to ||", to ll a.m.) 1 P. u.- GSR. Y. Fever 122 ‘ : : 42.9 40.1 CARY, Control 1. : F : 24.5 42.8 Hae S Bo Bever2.5 = HERS D> Control: 1S Re IDE Co Lie HH. DVA. Fever3 2 = H. D. A. Control V F. E.S. Fever 4 SS SS) Oe H.R. D. ControlIV| 21.3 | 21.8 | 22.4 26.91) 26.4 | $27.4 | 26.6 Gi VeS= Fevers . - 49.8 =H { FH H.R. D. Control II 36.5 3102 30.2 EeeReD 2 = 2228 222, DAR SS SS A.H.M. Fever6 . 36.7 a A.H.M. Control VI] 32.5 | 30.7 31.1 33.0 Nov. 21 4 ay Pie || 30:0 27.8 27.0 March 23 4 e« Bee zee Fes ee. 227 Hears * 7.00 A.M-8.10 A.M. The remaining figures for this experiment are for two-hour periods following 8.10 A.M. 2 g.00-12 noon. 3 Il A.M.—I2.15 P.M. * Control experiment made with same subject and introduced for further comparison. 5 8.30 A.M.-9.30 A.M. \ Observations on Metabolism during Fever. 229 An increased water of vaporization output might be expected as a result of the increased respiration rate during fever in that more air would be drawn into the lungs, there saturated with water and expelled again. This is, however, on the assumption that the total ventilation of the lungs is greater during these fever experiments than during the control. It is greatly to be questioned, however, whether or not the increased respiration rate was not in part com- pensated by a diminished volume, for, as noted elsewhere,!° a num- ber of the subjects complained of difficulty in taking long breaths and the respiration was of a decidedly shallow type. Hence we are not sure that there was a greater total ventilation during the febrile period. The mere fact that the body temperature is increased some I or 2 degrees C. would not in itself lead us to expect any marked increase in the amount of water vaporized from the surface of the body, but obviously more elaborate experiments to study this par- ticular point should be made before any accurate deductions can be drawn. The present data indicate that a greater amount of water is vaporized from the lungs and skin during fever than under normal conditions. Heat eliminated. — These experiments permit of an interesting comparison of the heat elimination during fever and during con- trol. The results have been tabulated for comparison in Table XVI herewith. An examination of the results shows an interesting contradiction. In all experiments but that with A. H. M. there is a noticeable increase in the heat elimination during the fever periods. In the experiment with A. H. M., on the other hand, there is in some in- stances an actual decrease. This can, however, in part at least, be accounted for by the fact that during the fever experiments this subject spent most of the forenoon lying on the bed reading or sleeping, and hence the conditions are not ideal for comparison. If we consider particularly those periods during the night when the subjects were in bed and presumably asleep, the evidence is all in favor of the assumption that there was a considerable increase in the heat elimination, the only exception to this being the experi- ment with F. E. S., controlled by the experiment with H. R. D., another subject. Here the contrary is indicated, as the heat elimina- tion was somewhat larger in the control period. 10 -T. M. Carpenter and F. G. Benepict: This journal, 1909, xxiv, p. 194. ‘uostedwio9 19y JAN} 1OF poonpojur pue yoofqns owes YIM opeul syuoutedxa JomyUOD KH, + — Te —/ ies |e oh saa eke adh kee +6L 9°S8 hae ee eta wa ae ley wears era”, MOLE SE tS a ca a Sa eet Ste aaa Ne g'Es Vrs 9°2S i pan an TS AON S weir — = Nate | rua €S8 78 s9S css SS IA [97100 “WH 'V % a I) SS cre he aes ee AOS AS) VEE) WGRV A WARS Me grates 42) v'6S ELS ~ 9 loAoyT “WH 'V ea) ae ——_ | a4" : a ieee a a ak aes 5 ii sie he: 0'€9 819 $09 Ill 5 c@eeclivEn O pases eran | Se ne earn. eel SS pee b'6S L'19 T'+9 II lou0D “Gd “AH “XK | —— nH eter fe tee er alia wa me fe RRR Srl ae gecQn Cee eG ey 6°29 " * g3asaq “S*A‘D = + Se ee Oe Oo ce ee oS8 £68 b°C6 9°26 Lv6 6°19 6°S9 vs9 AT JORUOS) Gi O'ZOT | OTL 6 80T 9TIT 9°SOT £°6S 019 Sas9. WEES eG chat - Been eae Ale api arse em at fie 6'8L S'b6 9'b8 6'+8 9'EL Ajonu0g “-V'd"H aS aa Sal ee OO, Wi ZO, SOIT | 8°26 SOL SL yOees Teri) 79O0Le8 8 S062 e268 S8L> ZiGL i Ke tesou VCE ne Paes BS me a rise, aan) fii puis ei. 0°¢9 8°19 S109 III», Eats Eas sc [SES i ceo eae a Mc a $'6S £19 T'+9 II 199u0D “GU "H LC ees ie RS oe ean = ar ate ee | SOLA SL: 8 VL Oss Oi8S SiS Ns eGtohore Sa OE = a1 TL9 VLL 9°18 8°r0T 0°9S OSL SPL TOU IOS A te). © ; a Ie : I ae = oa 9°6L 9'r8 6°68 8°26 8°06 b°76 Vel eel L8L The CEA TS Wize be) S ——_ = Es Sy | Soe A ———“——_ ———_ —_—_— —" = oe 76) “Teo Te “1890 Te) “Teo Te) ize) “Te ze) “Te0 Te) ‘Teo “Te0 eS Wdp|Wed¢|NW'd ZZ] Nd | NV ZL WV IT) VOT) V6/N V8(W'V Z| NV O/ WV CNV yy] WV E | ‘Nv Z 0} 0} 0} oO} 0} 0} 0} 0} 0} oO} oO} 0} 0} 0} 0} *pofqng fo) Wade | Wd Z| Wd TW ZT | WV ITN VOT] NV 6 TN VSI NV ZW VO] WV G/N VP) Ve] WV Z| NVI Se) 2 “IOUINOD) AUNV UWHADT NI ANOR Add CGALVNIWNITY LVAY TAX WIdvib Observations on Metabolism during Fever. 231 In view of the necessarily tentative nature of all deductions made from these experiments, it has not been deemed advisable to at- tempt to discuss the influence of fever on the various paths of heat elimination. As an extensive investigation into the total metabolism and heat production of the various types of fevers is soon to be undertaken, such discussion can profitably be left until greater and more accurate and positive data have been accumulated. Heat production. — The marked temperature rise in many of these experiments results in the storage of a considerable amount of heat within the body, and accordingly the heat elimination as measured by the heat given off from the body supplemented by the heat required to vaporize the water does not give a correct impression of the heat actually produced. It is necessary, therefore, to correct the heat elimination for changes in body temperature.’? The results for these fever and control experiments have been computed and compared in Table XVII herewith. Whatever doubt may exist with regard to the increase of carbon dioxide production, oxygen consumption, water vaporization, and heat elimination, there can be no doubt that during these experi- ments there was a marked increase in heat production. In prac- tically every instance we find, during the periods when fever was at its highest, a very noticeable increase in the heat production. In considering this table it is important to distinguish between those periods which are averaged together. Thus, in the period from I P. M. to 2 Pp. M. with F. E. S. the heat produced was 79.5 calories, while apparently in the control period with H. R. D. it was 84.8 calories. However, if we average the period from 12 noon to I Pp. M., and I P.M. to 2 Pp. M., we obtain evidence of distinct heat increase. This is but another indication of the disadvantage of comparing two experiments on two different individuals. When we compare the experiments made with C. R. Y., H. D. A., and A. H. M., in which the control was made with the same person, the figures show conclusively that during the febrile stage there is a marked increase in the total heat production. Unfortunately the data do not throw any light upon the heat production during the period when the body temperature remains constant nor during defervescence of the fever. To be sure, in the 1 For a complete discussion of this point, see Publication No. 77 of the Carnegie Institution of Washington, pp. 46-50. SS Sy SS SS Coa base a ato wk kein ce eB Ggge (ee | eeu eee | ey eel et) eevee nee, LQ ian ke oan ae Nala ie a hee ay 6LS Ss 94S » ” 1LG AON eae baaes a leiaee elle as ae |e (Para 268 ae cae ie Bie IA 190009 “WH 'V S Si oa eR “< T88 6 SOL O16 CSIT| 8°46 | 9'PZI | 798 e7E9 cf 6 " * QI0A0T “WH 'V aa) is ee) SF cS hat i Ane om ad aad en ees a 10 POL 0O'19 S69 Ts 3 Giese ea at et he eek ate San 2) = ‘ales BS +69 £°6S TLS II ]oUOs) “d “Ua ‘H a) "FF | MT Sele cer ate raiment te metre ager [crc yb aoeee| FA Tee eg | 796 J 949 | Log S99 |? * Gaeton SA" SS ae = = —-——7 ow o 78 878 6°76 © 06 L:SG0- 6SL hee ie AI [94009 “d‘U'H SS aete 626 | S6L Pb Sor Pp OfT SILL OOL T v9 T19 ee PicoAote Site sl rar cee eras |e ag Sy 128 9°b8 Z'6L Sb8 91, |ITIenuD -y'a‘H ~ A A A A A a Mae a! 8601} £06 8'O0ZT 9°16 GSILT|] O'OOT|) SLIT] S46 + +°06 $96 916 TS6 £06 |° © €49A07 “VY 'C‘H coe || aN AT VO es ce (ee eel es $9L 019 s69 = | ITT osW0D CH SE ee a aa en a oa ama 469 £6 Ts |" Ir1enu0p ‘QA seri tect eer cate alee |e! ocala al co omeeenee ora panel Se iee aceme ae aaa fess SS ibe 199 Tels 8°S8 c LOL Le9 9'6L VOL * JT JorMOD “AAO oe ae 0°01 yell | PIII T | 0°80t 'L6 Z6L bbl 96, |" - Tteeg AU" S ein. 7 ay a —~“. ———““—. ——nN~. ———— = "Teo “T8) ‘yep mie) 70) ‘T80 ‘Te a ie) “T20 ‘v9 Bie) "Teo mse) i 78) “Teo RK, [Nd pl weag| wa Z| Wd juoouZy | WV IT] "WV OT|N'V6/NvVg/ WV Z\Wvglnwvs|nv yl Wwive| ‘WV Z 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} 0} ‘poafqns ‘Wd €/W'd Z| ‘Wd [\Woou ZT/'W'V TT |W 'VOT| WV6 |Wvg|Nvz|Wivol;Wvs|Wvy| HK VeE| WV Z| cM VT “IOUINOD GNV UAATY NI UNOP Utd Aaonaoug Ava TIAX AZTAVL Observations on Metabolism during Fever. 238 case of F. E. S., the body temperature remained practically con- stant from I2 noon to 3 P.M. and there was an indication of a noticeable falling off in the heat production. Similarly in the case of A. H. M. the temperature had reached a maximum at II A.M. and from 11 to 12 noon, the heat production decreased noticeably. From these few observations we might infer that the heat produc- tion after the body temperature had ceased rising was consider- ably less than during the period of temperature rise, but further experiments on fever will be planned to include observations on this point. THE VARIATIONS IN THE ENZYME CONCENTRATION WITH THE VARIATION IN THE BLOOD SUPPES TO) THE SECRETING: 'GEAND: BY jG: (RYAN: [From the Hull Physiological Laboratory of the University of Chicago.] I. INTRODUCTORY. ‘a ee experiments reported in this paper are a continuation of the studies of Carlson, Greer, and Becht, Carlson and McLean,’ on the variations in the composition of the saliva that follow the variation in the blood supply to the glands. The work was undertaken under Professor Carlson’s direction and assistance. In some of the experiments I was aided by Mr. A. B. Luckhardt. The main aim of the work was the determination of the rela- tive concentration of the ferments in the saliva produced by stimu- lation of Jacobson’s nerve, or reflexly, and that produced by the stimulation of the cervical sympathetic; and, secondly, to ascer- tain whether these differences can be duplicated by variations in the blood supply to the active gland. The studies of Carlson have shown that the concentration of the organic solids in the saliva increases pari passu with the diminution in the blood supply to the gland. If it should be found that the enzyme concentration runs parallel with that of the organic constituents, this might afford a clew to the structure of the enzyme itself. The mechanism of secretion of ferments and enzymes, their chemical composition and relation to organic substances, particu- larly the proteids, have long been matters of interest and research. Quite a number of the well-known enzymes, e. g., ptyalin? and lipase,* although not yet isolated in absolutely pure form, have been found not to give any of the proteid reactions; so it seems ? CARLSON, GREER, and Becut: This journal, 1907, xix, pp. 360, 408, xx, p. 180; CaRLsoNn and McLean: This journal, 1908, xx, p. 457, xxii, p. 279. ? CoHNHEIM: VircHow’s Archiv fiir pathologische Anatomie, 1863, xxviii, p. 241. 3 NicHoL: Personal communication. 234 Variations in the Enzyme Concentration. 235 that, in case of part of the enzymes at least, the proteid reactions formerly obtained represented impurities and not the true chemi- cal nature of the enzymes themselves. There is some evidence in favor of the idea that enzymes are derivatives of the substances upon which they act; e. g., yeast grow- ing on galactose does not ferment it at first, but will after some time has elapsed, — the conditions during the growth of the yeast having become such as to favor the formation of an enzyme or ferment from the galactose itself. This view, however, could not be true in the strict sense in the case of ptyalin, since the ptyalin is produced within the salivary gland cells, and in the ab- sence of the substance upon which it exerts its action, — starch. II. LITERATURE. Becher and Ludwig‘? found that the submaxillary corda saliva grows poorer in organic solids as the gland approaches fatigue. This was confirmed by Heidenhain,® who also showed that stimu- lation of Jacobson’s nerve in the dog gives watery saliva from the parotid, while stimulation of the cervical sympathetic nerve gives a viscid saliva. Langley, by microscopic examination of the parotid of the rabbit, observed that a more rapid change takes place in the gland cells during stimulation of the cervical sympathetic than during stimulation of Jacobson’s nerve. None of these observers, how- ever, took occasion to test the amylolytic activity of the saliva collected under the different conditions of secretion. Hofbauer* demonstrated that the activity of human mixed saliva fluctuates during the course of twenty-four hours, being greater before breakfast than after, —that is to say, the amylo- lytic activity of the saliva was diminished after a period of activity of the glands. He did not attempt to show any relation between the amylolytic power and the chemical composition of the saliva. Chittenden and Ely® carried out a series of experiments on 4 Becuer and Lupwic: Zeitschrift fiir rationelle Medizin, 1851, i, p. 278. 5 Hemenuarn: Archiv fiir die gesammte Physiologie, 1878, xvii, p. 28; HER- MANN’S Handbuch, v, p. 55. ® LANGLEY: Journal of physiology, 1880, ii, p. 26r. 7 Horsaver: Archiv fiir die gesammte Physiologie, 1897, Ixv, p. 503. 8 CHITTENDEN and Ey: American chemical journal, 1883, iv, p. 329. 236 J. G. Ryan. human mixed saliva collected between 9 and Io A. M. in order to show the relation of alkalinity (the alkaline salts contained in the saliva) to the amylolytic power. They obtained fairly constant results for the same individual, but the results for different in- dividuals fluctuated greatly, showing no corresponding differences nor relations between the amylolytic power and the alkalinity. Langley ® states that the alkalinity of mixed human saliva is least when fasting (as before breakfast), and increases, reach- ing its maximum during or after eating. On the other hand, Chittenden and Richards,'® while studying the variations in amylolytic power and chemical composition of human saliva, found the alkalinity, acidity, and amylolytic power to be greater before breakfast than after breakfast, — the same being true for dinner, but to a less marked degree. This suggests some degree of rela- tionship between the amylolytic power and the alkaline salts con- tained in the saliva. However, in view of the fact that ptyalin acts best in neutral solutions,** we would expect a greater amy- lolytic power after breakfast, when the alkalinity and acidity are reduced. Chittenden and Richards call attention to these points, and their results show that the differences in acidity and alkalinity before and after breakfast bear no close parallelism with the amylolytic power. They are forced to the conclusion that the greater activity of saliva before breakfast means simply a higher concentration of ptyalin. In the same series of experiments they also determined the relation of the amylolytic power to the or- ganic and inorganic solids in the saliva. Their results, although variable for different individuals, showed, on the whole, more organic and inorganic matter in the saliva before meals, and the stronger amylolytic power nearly always corresponded to the greater amount of organic and inorganic solids. Ill. ExprertmMENTAL METHODS. Large and medium sized rabbits were used in all experiments, and the various samples of saliva collected under different condi- tions. of secretion were tested comparatively as to their relative ® LANGLEY: Text-book of physiology, 1898, i, p. 504. 10 CHITTENDEN and RicHarps: This journal, 1808, i, p. 461. 1! LANGLEY and Eves: Journal of physiology, 1883, iv, p. 18; CHITTENDEN and SMITH: Studies in physiological chemistry, Yale Univ., 1885, i, p. 8. Variations in the Enzyme Concentration. 237 amylolytic powers. The concentration of the ptyalin was deter- mined (1) by the rate of clearing of the starch solution and (2) by the complete disappearance of erythrodextrin. A 1 per cent solution of arrowroot starch was used, and the saliva and starch were always mixed in the proportion of 1 cc. to 250 cc. Ina majority of the experiments where series of samples were col- lected it was impossible to obtain more than 0.5 to 1 c.c. in each sample, thus making it necessary to use small amounts of saliva in order that several tests could be made from each sample to check up the results and eliminate experimental errors. There- fore, to maintain the ratio of I c.c. to 250 c.c. as mentioned above, 0.1 c.c. of pure undiluted saliva from the various samples was carefully measured out into very small specially prepared glass receptacles of uniform size and depth and dropped simultaneously into narrow-necked 100 c.c. flasks containing 25 c.c. of starch, and after shaking the mixtures for the same length of time and with the same intensity, they were allowed to stand at room tempera- ture, — the rate of clearing and the disappearance of the erythro- dextrin being taken as evidence of the rate of action. After the first general comparative tests were completed, a second test was always made to obtain more accurately the relative percentages of ptyalin in the different samples collected under different condi- tions of secretion. At first this was tried by diluting the highly active samples with physiological salt solution to various strengths and comparing these dilutions with the less active ones. But find- ing it extremely difficult to get uniform mixtures with such small amounts of saliva, and to obviate the possibility of any retarding or accelerating effect that the salt solution might have upon the rate of action of the ptyalin, this source of error was eliminated by taking a certain amount (0.1 c.c.) of the highly active samples and comparing it with 0.2, 0.3, 0.4, 0.5, and 0.6 c.c. of the less active ones. By this method it could be determined how much of the weaker saliva was necessary to equal 0.1 c.c. of the stronger, and by applying Schiitz’s Law their relative percentages of ptyalin could be quite accurately measured. The methods of obtaining the saliva were as follows: (1) under ether anzesthesia and (2) without anzesthesia. (1) During ether anesthesia. (a) Submaxillary saliva. —In the first few experiments cannulas were placed in Wharton’s ducts, and samples of 0.5 c.c. each of submaxillary saliva collected by alternate 238 J. G. Ryan. stimulation of the chorda and cervical sympathetic, and lastly by injection of pilocarpin when the glands would no longer respond to electrical stimulation of the nerves. (b) Parotid saliva. — Parotid saliva was obtained by placing cannulas in Stenson’s ducts and collecting saliva under the fol- lowing conditions: (1) By stimulation of Jacobson’s nerve, the electrodes being placed in the middle ear after rupture of the tympanic membrane and filling the tympanum with physiological saline solution. (2) By simultaneous stimulation of.the cervical sympathetic and Jacobson’s nerve alternating with stimulation of Jacobson’s nerve alone. (3) By stimulating of Jacobson’s nerve, alternating the samples with and without anzmia of the gland pro- duced by clamping of both common carotid and both vertebral arteries. In the majority of cases the anzemia after ligation was so complete as to necessitate artificial respiration, and in most instances the carotid artery on the same side as the gland had to be released occasionally in order to get any flow at all, even during stimula- tion of Jacobson’s nerve. In ligating the vertebral arteries care had to be taken to keep the animal warm and avoid too much manipulation of the brachial plexus, otherwise the animal would invariably die of shock. Several of the animals seemed to be particularly susceptible to the shock produced by placing the elec- trodes in the middle ear, and for this reason in part of the ex- periments pilocarpine was substituted in place of electrical stimu- lation of Jacobson’s nerve. (2) Without anesthesia. (a) Parotid saliva. —In this series of experiments without anzsthesia a permanent fistula was established in each cheek of a large white rabbit, and with proper care the one animal served to complete the entire series. Saliva was obtained from these fistulas as follows: (1) The animal fasted over night and then normal or reflex saliva was collected by allowing the rab- bit to eat cabbage, carrots, etc., several samples of 0.5 c.c. each being obtained at each feeding. (2) Two samples were collected by feed- ing cabbage. Then the cervical sympathetic nerves were isolated (by local anzesthesia with ethyl chloride) and stimulated for ten minutes, during which time there was never any appreciable flow from the gland except at one time I c.c. was obtained. After this period of stimulation of the sympathetic, the animal was allowed to eat again and the saliva collected as before. (3) Owing to the difficulty of keeping the animal alive after clamping of the carotid LS Variations in the Enzyme Concentration. 239 and vertebral arteries, anemia of the glands was produced as follows: (a) Two samples of reflex saliva were obtained and then 30 c.c. of blood drawn from the ear. After allowing the animal to rest for a few minutes it was fed more cabbage and the saliva collected as before. If the animal refused to eat after bleed- ing, the saliva was obtained by injection of pilocarpin. (b) The animal was bled every fifth day from the ear for three experi- ments, and then after a week’s rest a fourth test was made by drawing 25 c.c. of blood directly from the heart with a sterile needle. (4) Numbers 1, 2, and 3 were all repeated, using pilo- carpine instead of letting the rabbit eat cabbage. The methods of collecting saliva without anzsthesia were used to check up our previous results by approaching as nearly as pos- sible to pure physiological conditions, thus obviating the errors that might be introduced by changes in blood pressure, heart action, osmotic pressure of the blood,” etc., produced by the anesthetic. TV. Resurrs, I. During anesthesia. (1) Submavillary saliva. — The submaxil- lary as obtained in the first few experiments was found to be entirely devoid of amylolytic power. After testing a number of samples, some of them standing several days at thermostat temperature, we were convinced that the submaxillary saliva of the rabbit exerts no appreciable diastatic action on starch, and consequently it was not considered in the remainder of the experiments. (2) Parotid saliva. — The parotid saliva, on the other hand, is highly active, equal to and in some instances greater than pure parotid human saliva collected by inserting a cannula into Sten- son’s duct at its opening into the vestibule of the mouth. In one case, however, in a perfectly healthy rabbit the parotid saliva showed absolutely no action on starch, and the serum from the same animal was also found to be entirely free from diastase. We offer no explanation of this peculiar phenomenon. (a) Gradual decrease in the ptyalin during the secretion of the gland. — Table I is a typical example of the results of eight ex- periments, in four of which series of samples of I c.c. each were obtained by stimulation of Jacobson’s nerve till the gland became 22 Caritson and LucxHarpr: This journal, 1908, xxi, p. 162. 240 J. G. Ryan. fatigued, and in the other four by the use of pilocarpine. The saliva of different animals varied considerably in the rate of action, but the same essential features are always present. From TABLE I. THe DECREASE IN THE PTYALIN CONCENTRATION OF THE RABBIT’S PAROTID SALIVA DURING THE PERIOD OF ACTIVITY. RESULTS OF EXPERIMENTS 1 (STIMULATION OF JAcoBson’s NERVE) AND 5 (HyPODERMIC INJECTION OF PILOCARPINE). Results of Experiment 1. Results of Experiment 5. Saliva samples of 1 c.c. each Rate of Disappearance Rate of Disappearance clearing. of erythrodex- clearing. of erythrodex- trin. trin, min. sec. i min. sec. min. sec. 0 15 0 12 2 00 0 30 0 40 6 00 0 37 0 45 7 30 0 42 0 52 9 00 0 60 1 8 14 00 1 30 2 00 22 00 4 00 5 00 50 00 15 00 2 hours 12 00 2 hours Scarcely any No action action za after 1 hour this table we see that the amylolytic power decreases rapidly at first, then gradually, and as the gland approaches fatigue there appears to be another sudden decrease, and finally the ptyalin al- most entirely disappears. It would thus seem that the concentra- tion of the ptyalin during a period of activity runs the same course as that of the organic solids. (b) The increase in ptyalin produced by stimulation of cervical sympathetic.—In six experiments upon the effect produced by simultaneous stimulation of Jacobson’s nerve and the cervical sym- pathetic, alternating with stimulation of Jacobson’s nerve alone, uniform results were obtained, as shown in Table II, which repre- sents one typical experiment. Variations in the Enzyme Concentration. 241 (c) The increase in the ptyalin produced by clamping the ar- teries supplying the gland.— The six experiments were carried out using anemia produced by clamping of both common caro- tids and both vertebral arteries instead of stimulation of the cervical sympathetic, and the results were practically identical with those on the sympathetic saliva. This shows that the con- TABLE II. DETAIL OF EXPERIMENT 2 IN THE SERIES SHOWING THE INCREASE IN THE PTYALIN CONCENTRATION PRODUCED BY THE STIMULATION OF THE CERVICAL SYMPATHETIC. Action. Relative concentra- Comparative | . tion of ptyalin cal- action. culated according to Schiitz’s Law. Saliva samples of 1 c.c. each. Disappear- ance of erythro- dextrin. Rate of clearing. min. (1) Jacobson’s N. 5 0.1 c.c. of (2) Bysym. & J. N. 1 0.6 c.c. of (1) 0.1 c.c. of (2) & (3) J. N. na core 3) 0.1 c.c. of (4) = (4) Sym. & J.N. 0.4 c.c. of (3) (5) Jou: No test made 0.1 c.c. of (6) (6) Sym. & J. N. 0.2 c.c. of (5) ditions (vaso-constriction during stimulation of the sympathetic and clamping the arteries) producing a reduction in oxygen sup- ply to the parotid gland cause an increase in the concentration of ptyalin in the parotid saliva. A corresponding change in the percentage of organic solids is produced by these same conditions of secretion, but to a less marked degree. The work of Heidenhain and Carlson, Greer and Becht, and Carlson and McLean on the salivary glands of the dog, cat, and rabbit shows the parotid sympathetic saliva to be much richer in organic solids than saliva obtained by stimulation of Jacobson’s nerve or by injection of pilocarpine, but in no instance did they find the great difference shown in the relative amylolytic powers of the two salivas. This marked difference as shown by Schiitz’s Law seems almost incred- ible, but, on the other hand, even if Schiitz’s Law would not hold good, there is still a greater difference than has been obtained 242 J. G. Ryan. in the relative percentages of the other organic constituents. In these experiments it was not possible to measure the degree of anemia during sympathetic stimulation or compression of the arteries to the gland, and for this reason the same differences in concentration of ptyalin were not always obtained. In two in- stances the changes were only slight, probably due to faulty stimu- lation or to an inactive sympathetic nerve. The rate of secretion during anemia of the gland (however produced) is always very slow, but in these two cases the slowing was scarcely. perceptible, and we were led to believe that marked diminution in the blood supply had not been secured. 5 io Ill rS2° O.5Ort 627008 200° o.4or “% Nee To IV i cc : “ nA 0.70 10.29 Wotale . 4.2. 3S. Ae 2 eee ee 242.18 gm. The undistilled residue weighed 56.00 gm. Fraction I. — This fraction was saponified by boiling with ten volumes of water for ten hours, the solution evaporated to dryness under diminished pressure at 50°, and proline extracted from the 254 T.B. Osborne, D. B. Jones, and C. S. Leavenworth. residue by boiling with alcohol. The part insoluble in alcohol, which weighed 32.44 gm., was separated into thirteen fractions by sys- tematic crystallization, and the thirteenth fraction, which contained all of the remaining substance, weighed only 3.02 gm. Analysis showed this to contain 38.77 per cent of carbon and 7.52 per cent of hydrogen, indicating the presence of very little glycocoll if any. By further systematic crystallization of these thirteen fractions, 13 gm. leucine, 9.91 gm. of valine, and 8.81 gm. of alanine were obtained. The leucine, when recrystallized once, gave the following analysis: Carbon and hydrogen, 0.1437 gm. subst., gave 0.2891 gm. CO, and 0.1279 gm. HO: Calculated for C,H,,0,N = C 54.96; H 9.92 per cent. oun diene =C 54.87; H 9.89 “c 6c cd The valine crystallized in homogeneous characteristic plates which gave the following analysis: Carbon and hydrogen, 0.1207 gm. subst., gave 0.2280 gm. CO, and 0.1022 gm. HO! Calculated for C;H,,O,N = C 51.28; H 9.40 per cent. Pounds ute =C E525 H 9.40 “ TZ: When dissolved in 20 per cent HCl, it showed a specific rotation 20° : of (a) eee + 23.10°. For further identification the valine was racemized by heating for twenty-four hours with baryta in an autoclave at 170°-175° and converted into the phenylhydantoic acid derivative which crystallized from water in characteristic hexagonal plates melting sharply at 159°. Carbon and hydrogen, 0.1308 gm. subst., gave 0.2936 gm. CO, and 0.0816 gm. HO: Calculated for C,,H,,O,;N, = C 60.96; H 6.83 per cent. HOUMGME hi. sre) ce = == C161.22; 6:98 “Ss -% The alanine, when recrystallized from water, formed large, dense prisms which decomposed at 290°. Carbon and hydrogen, 0.1798 gm. subst., gave 0.2667 gm. CO, and 0.1283 gm. HO: Calculated for C,H,O,N = C 40.45; H 7.86 per cent. Pound. --220.5%2 4% = Cir4ods si 7.93. 459 4" Hydrolysis of Crystallized Albumin. 255 Fraction 11. — This fraction was saponified by boiling with water, evaporated to dryness, and the proline extracted from the residue with absolute alcohol. The amino-acids insoluble in alcohol weighed 38.79 gm., of which 27.77 gm. were leucine. When recrystallized once from water, the leucine gave the following analysis: . Carbon and hydrogen, 0.1375 gm. subst., gave 0.2752 gm. CO, and 0.1212 gm. EO: Calculated for C,H,,0,N = C 54.96; H 9.92 per cent. Bound) aes se 2) 2 = 'C 54.505 EH 9:79) From the filtrate from the leucine were isolated 2.65 gm. copper aspartate and 4.02 gm. of the copper salt of leucine. The copper aspartate crystallized in the characteristic sheaves of fine needles which gave the following analysis: Copper, 0.1297 gm. air-dried subst., gave 0.0372 gm. CuO. Calculated for C,H,0,NCu 44 H,O = Cu 23.07 per cent. UNG UE A a OE BOE lle eas OD ae The leucine copper salt on analysis gave the following results: Copper, 0.1190 gm. air-dried subst., gave 0.0288 gm. CuO. Calculated for C,,H,,0,N,Cu = Cu 19.50 per cent. BOWE wg co Rha me = Curto:34 = -* From the alcoholic extracts of Fractions I and II the proline was separated in the usual way. By converting the proline into its copper salt and separating the levo from the racemic salt by solu- tion in alcohol, a quantity of each of these forms was obtained which corresponded to 13.65 gm. of /-proline and 0.46 gm. of r-proline. The phenylhydantoine of the J/-proline crystallized ‘in characteristic prisms which melted sharply at 142°. . The air-dried racemic proline copper salt gave the following analysis: Water, 0.1371 gm. subst., lost 0.0155 gm. H,O. Calculated for C,,H,,O,N,Cu 2 H,O = H,O 10.99 per cent. De Cat¢ IM ie ay ioe ae ee VE Oner.2o Copper, 0.1206 gm. subst., dried at 110°, gave 0.0330 gm. CuO. Calculated for C,,H,,O,N,Cu = Cu 21.81 per cent. Mound) 3.4.4) 646) eh Sie Eoo. v!: oY 256 T. B. Osborne, D. B. Jones, and C. S. Leavenworth. _ Fraction I111.— The phenylalanine was extracted with ether and converted into the hydrochloride, which weighed 9.37 gm. The free phenylalanine gave the following analysis: Carbon and hydrogen, 0.1222 gm. subst., gave 0.2930 gm. CO, and 0.0748 gm. HO: Calculated for C,H,,O,.N = C 65.45; H 6.66 per cent. Found ytca.se @ = © O5744- 7 E110:800 ee The aqueous layer was saponified with baryta, and 3.05 gm. aspartic acid obtained in the form of the barium salt. The free aspartic acid reddened without decomposing at about 300°. Carbon and hydrogen, 0.2271 gm. subst., gave 0.2997 gm. CO, and 0.1099 gm. HO: Calculated for C,H,O,N = C 36.09; H 5.26 per cent. Found a fe oa So = G 35-99; H 5-39 73 73 The filtrate from the barium aspartate yielded 6.32 gm. glu- taminic acid hydrochloride and 7.33 gm. copper aspartate. The copper aspartate gave the following analysis: Copper, 0.1496 gm. subst., gave 0.0433 gm. CuO. Calculated for C,H,O,NCu 44 H,O = Cu 23.07 per cent. Hound! ...02e ss cee ee ge —NCN2 arse ne The filtrate from the copper aspartate was freed from copper with hydrogen sulphide and an attempt made to isolate serine. From the mixture of substances contained in the solution 0.3 gm. of aspartic acid and 0.29 gm. of phenylalanine were isolated, but no serine was obtained even after long-continued efforts to bring it to crystallization. Fraction Iv.— This fraction, when treated in the same way as described for Fraction III, yielded 11.48 gm. of phenylalanine, hydrochloride. By freeing the mother liquor from this from the excess of hydrochloric acid, decolorizing with animal charcoal, and neutralizing with ammonia, 3 gm. of the free acid were isolated which decomposed at 273°. The total phenylalanine thus obtained from the esters was equal to 20.08 gm. of free phenylalanine, or 5.7 per cent of the ovalbumin. The aqueous solution from which the phenylalanine had been removed by shaking with ether yielded 29.33 gm. of glutaminic acid hydrochloride and 1.2 gm. of copper Hydrolysis of Crystallized Albumin. 257 aspartate. The free glutaminic acid decomposed with effervescence at 202°. Carbon and hydrogen, 0.1649 gm. subst., gave 0.2463 gm. CO, and o.ogor gm. HO: Calculated for C,JH,O,N = C 40.81; H 6.12 per cent. ; Found aw Ce wl € a (S 40.74; H 6.07 “ “cc THE RESIDUE AFTER DISTILLATION. This yielded 1.60 gm. of glutaminic acid hydrochloride, which decomposed at 200°. From the total products of this hydrolysis were isolated 29.85 gm. glutaminic acid, or 7.53 per cent of the ovalbumin, or about 82 per cent of the quantity obtained by Osborne and Gilbert ® by a direct determination, namely, 9.10 per cent. CYSTINE. Cystine was not obtained, although a persistent effort was made to separate it from a solution of the products of hydrochloric acid hydrolysis of 100 gm. of the ovalbumin. Under similar conditions Abderhalden and Pregl isolated 0.2 gm. of cystine, which shows the presence of this amino-acid in the albumin. We have no doubt that our failure to obtain cystine was entirely due to our inability to establish favorable conditions for its separation, for, as Moerner °® says, it is an accident if the best possible output is obtained. TYROSINE. A quantity of ovalbumin equivalent to 43.87 gm. ash and moisture free substance was hydrolyzed by boiling for twenty-four hours in an oil bath with a mixture of 150 gm. of sulphuric acid and 300 c.c. of water. The sulphuric acid was then removed from the diluted solution with an equivalent quantity of baryta, and after thoroughly washing the barium sulphate by repeatedly boiling with water, the filtrate and washings were concentrated to crystallization. The sub- 5 OsporNE and GILBERT: This journal, 1906, xv, p. 333. 8 MoernerR, K. A. H.: Zeitschrift fiir physiologische Chemie, 1902, xxxiv, p. 215. 258 T. B. Osborne, D. B. Jones, and C. S. Leavenworth. stance which separated after twenty-four hours was dissolved in water, the solution decolorized by boiling with bone black, and the tyrosine crystallized by concentrating and cooling. The tyrosine, thus obtained, weighed 0.7805 gm. equal to I. 77 per cent, and gave the following result on analysis: Nitrogen, 0.2183 gm. subst., required 1.68 c.c. # N—HCI. Calculated for C,H,,0,N =N 7.73 per cent. Hound wey eet eae =N 7.69 “ “ The filtrate and washings of the tyrosine were used for deter- minations of the basic amino-acids according to the method of Kossel and Patten. HISTIDINE. The solution of the histidine = 500 c.c. Nitrogen, 50 c.c. sol. required 2.04 c.c. # N—-HCI = 0.2040 gm. N in 500 c.c. = 0.7500 gm. histidine = 1.71 per cent. The histidine was converted into the dichloride for identification. Chlorine, 0.1153 gm. subst., gave 0.1453 gm. AgCl. Calculated for C,H,,O,N,Cl, = Cl 31.14 per cent. HOMME cet ean ae rated CZ tril 6) ae maa ARGININE. The solution of the arginine = 1000 c.c. Nitrogen, 50 c.c. sol. required 3.3 c.c. 4 N—HCl = 0.6600 gm. N in 1000 c.c- = 2.0506 gm. arginine + 0.1026 gm. = 2.1532 gm. = 4.91 per cent. The arginine was converted into the copper-nitrate double salt for identification. Copper, 0.1154 gm. subst., air dry, gave 0.0160 gm. CuO. Calculated for C,,H,,0,N,Cu(NO,), 3 H,O = Cu 10.79 per cent. Found: Aichi eee Cee eee = Cuomios' Fy Hydrolysis of Crystallized Albumin. 259 LYSINE. The lysine picrate weighed 4.2407 gm. = 1.6509 gm. lysine == 3-76 per.cent. The lysine picrate gave the following analysis: Nitrogen, 0.3000 gm. subst., dried at 100°, required 5.62 c.c. # N—HCL. Calculated for C,H,,O.N, - C,H,0,N, = N 18.67 per cent. OUIGbes eae wee teeta torn en iar at ate = Nitonya Say es CARBOHYDRATE. Very conflicting statements are to be found in the literature re- specting not only the amount, but even the existence of carbohy- drate in ovalbumin. Much of the confusion has unquestionably , arisen from a lack of purity of the ovalbumin used in many of the experiments, for some of those who were careful to separate their ovalbumin from ovomucoid overlooked the presence of ovomucin. Since Eichholz has shown that ovomucin yields an osazone, a part of the contradictory data that are on record may be explained by the presence of this substance in some of the preparations of oval- bumin which have been examined. Abderhalden* has found that the glucosamine content of once recrystallized ovalbumin is 7 per cent; of thrice recrystallized, 4 per cent; and seven times recrystallized is 2.5 per cent, as indicated by the weight of the crude osazone. The latter proportion is in close agreement with the quantity similarly estimated by Osborne and Campbell § from the weight of the crude osazone obtained from care- fully purified preparations of ovalbumin. From the six times recrystallized ovalbumin used for this hy- drolysis, in one experiment we obtained a similar quantity of crude osazone, but on recrystallizing this we found that it contained a not inconsiderable quantity of tyrosine. After separating the tyro- sine too little of the osazone remained to permit of its purification by further recrystallization. In another experiment 10 gm. of the ovalbumin were boiled with 7 ABDERHALDEN: Lehrbuch der physiologischen Chemie, 1909, p. 217. ® OsBorNE and CAMPBELL: Journal of the American Chemical Society, 1900, Xxli, p. 422. 260 T. B. Osborne, D. B. Jones, and C. S. Leavenworth. 200 c.c. of 5 per cent sulphuric acid for three hours, and after removing the sulphuric acid the solution was concentrated, under diminished pressure at a low temperature, to about 150 cc. A mixture of 6 c.c. of phenylhydrazine and 6 c.c. of 80 per cent acetic acid and 20 gm. of sodium acetate was then added, and the solution heated on the water bath for three hours. After twenty-four hours the substance which separated was filtered out, washed with water and with alcohol, and dried at 100°. By concentrating and cooling the filtrate, from this first separation, a further small quantity was obtained which was added to the first. The crude product, which weighed 1.1033 gm. was dissolved, as far as possible, in boiling absolute alcohol, the filtered solution concentrated, and water added until a precipitate formed. After heating until all redissolved, the solution was allowed to cool over night. The substance which separated in balls of needles was again recrystallized, as before, and obtained in the form characteristic of glucosazone. When dried at 100°, this weighed 0.0860 gm. By concentrating the filtrate a sec-+ ond crop, weighing 0.0500 gm., was obtained, making the total pure glucosazone 0.1360 gm., which melted at 202°. From 0.5000 gm. of pure crystallized glucose, by similar treatment, 0.6470 gm. of osazone was obtained. If the osazone was yielded by the glu- cosamine from the ovalbumin in the same proportion as by the glucose, this quantity would correspond to 0.1136 gm. of glucosa- mine, or 1.23 per cent of the moisture and ash-free ovalbumin. It is not probable that this result gives any fair measure of the glu- cosamine which ovalbumin yields on hydrolysis, for we have no evidence that the brief hydrolysis employed is sufficient to liberate the whole of this substance. It is possible that much of the glu- cosamine may have been removed with the barium sulphate, for Miller ® found that his solutions lost much in reducing power after removing sulphuric acid with baryta. He also found that glu- cosazone separates from a solution of glucosamine more slowly than from a solution of glucose, and from 3 gm. of glucosamine hy- drochloride he obtained only 1.0240 gm. of glucosazone. An attempt was made to isolate the glucosamine as the phenyliso- cyanate derivative according to the method proposed by Steudel,?° but without success. Attempts to estimate the carbohydrate from the copper oxide reducing power of the hydrolysis solutions also ® Miter: Zeitschrift fiir Biologie, rgor1, xlii, p. 468. 1 STEUDEL: Zeitschrift fiir physiologische Chemie, 1902, xxxiv, p. 353. Hydrolysis of Crystallized Albumin. 261 failed, for, in harmony with the previous experience of Osborne and Campbell, no reduction could be obtained, even after repeated and persistent efforts. The amount of glucosamine which we have found can therefore be considered to be of only qualitative value, for approximately quantitative determinations evidently cannot yet be made by any of the methods now available. The results of the several hydrolyses of ovalbumin which have been made are as follows: Osborne and Abderholden Hougounenq Jones. and Pregl. and Morel. per cent. per cent, per cent. EIVCOCOMMS VS coy ear ce ke Teele 0.00 0.00 0.00 PRHINMIA A Ares WSR ence Sal ae attic 2.22 2.10 8.40 VLE eae Age rn ae an 2.50 ? sea WPeUGCME ene. Sle we es 10.71 6.10 15.20 COMME Fels 5 a ai etee on os 3.56 D5 1.10 ipbenylalamime 02/28/06. 5.07. - 4.40 5-20 PNSPALLICTACIG o-oo sy io> et'et 2.20 1.50 1.70 Guutaminic acids | a oe eae 9.10 8.00 3.50 SS e) en ae Sg ? ? Sag MAFEOSINOUS Sens, Sot Sus 577 1.10 0.99 pete fe alae ts apse! arse? = ? 0.20 28 PUISCIMTS 5) Sil) eR oe) sha pony eas rae reine sti eee, % wey ena 4.91 Sar rer LSI Co Sine aaa el 3.76 Lake 0.27 JEN AtV0 (eo eg 1.34 a ais Gracosamine, — 20%. 2 es 1:23 axe soe Sicyotophane) 2 si. =. = -<) + 4s present sare sae Mb otalig omen ny ey Seo, 2 50.08 Hougounenq’s analysis was made on the products of hydrolysis with baryta, and is therefore not comparable with those of Abder- halden and Pregl or the authors’. In a note published later by Abderhalden!? he gives as corrected figures for leucine 7.1 and for alanine 8.1 per cent, but gives no data concerning the way in which this result was reached. It is to be noted that the sum of the leucine, valine, and alanine in the hydrolysis by Abderhalden 11 OsBORNE and CAMPBELL: Journal of the American Chemical Society, 1900, Xxli, Pp. 422. 12 ABDERHALDEN: Zeitschrift fiir physiologische Chemie, 1906, xviii, p. 518. 262 T.B. Osborne, D. B. Jones, and C. S. Leavenworth. and Pregl, namely, 15.2 per cent, agrees closely with the sum of the leucine, valine, and alanine which we have found, namely, 15.43 per cent. As we weighed only practically pure substances which were subjected to a strict chemical identification, it would seem probable that Abderhalden and Pregl’s earlier figure for alanine was more nearly correct than the revised figure given by Abderhalden later. In other respects these two independent hydrolyses are in good agreement, and show that essentially the same results can be obtained by experienced workers if sufficient care is taken in sep- arating the different amino-acids. The low summation shown by our hydrolysis is not due to any defect in carrying out the processes incident to the isolation of the amino-acids, for throughout the entire analysis the separations were effected with unusually small losses. The unusual deficiency, in our opinion, is rather to be attributed to the presence of some non- protein complex which, in combination with protein, constitutes this albumin. Possibly some complex similar to chondroitin-sulphuric acid may here occur, which is suggested by the fact that the amount of sulphide sulphur to be obtained from ovalbumin indicates that one half of its total sulphur belongs to some other complex than cystine. Further evidence of this is given by the fact that when boiled with dilute sulphuric acid a volatile acid is yielded in small quantity, which Seemann ?* found to be acetic acid. We, also, ob- served the presence of a volatile acid, but its amount was too small to permit of its identification. 13 SEEMANN: Archiv fiir Verdauungsheit, 1898, iv, p. 275. ii EEPECTS OF CHLORIDE, SULPHATE, NITRATE, AND NITRIC RADICLES OF SOME COMMON BASES, ONSUHE FROGS: HEART, BY bh. CaCOOK. [From the Laboratory of Physiology of the George Washington University Medical School, Washington, D. C.] HE question of the relation of the inorganic salts of the blood and lymph to the contractility of the heart and other muscles has been under discussion for many years. Investigators have con- cerned themselves chiefly with the action of the chlorides of sodium, potassium, and calcium, but the actions of other salts of these bases and of the salts of other bases have been but sparingly studied. A review of the literature on this subject up to the year 1898 is given by Howell,! who concludes that the calcium in calcium chlo- ride is the stimulus responsible for heart contraction, while the rhyth- mic contractions and relaxations are brought about by the addition of a certain proportion of potassium. In mixed solutions, such as Ringer’s, sodium chloride seems to be essential, chiefly for preserv- ing the osmotic relations between the tissues and the surrounding liquid. Similar work was continued by Greene,” who dealt with the relations of the inorganic salts of blood to the automatic activity of strips of ventricular muscle. Greene also gives a review of the literature on this subject. Among more recent investigators in this country may be men- tioned Loeb,? Meltzer and Auer,? Howell,® Guenther,® Carlson,‘ Matthews,’ Denis,? and Hunt.1° The latter investigator studied the 1 HowELt: This journal, 1808, ii, p. 47. 2 GREENE: I[bid., 1898, ii, p. 82. 3 Lors: Studies in general physiology, University of Chicago, 1905. 4 MeLTzER and AUER: This journal, 1908, xxi, pp. 400 and 449. 5 HowELL: Ibid., 1901, vi, p. 191. 6 GUENTHER: Ibid., 1905, xiv, p. 73. 7 Cartson: Ibid., 1904, xii, pp. 55, 67, 471. 8 MattHews: Ibid., 1907, xix, pp. 5, 20, and 323 9 Dents: Ibid., 1906, xvii, p. 35. 10 Hunt: Ibid., 1899, ii, p. 395. 263 264 FNGz (Goor: nervous mechanism of the heart. The earlier work of Ringer, Locke, and others is well known. The object of this work was to study the actions of dilute solu- tions (I per cent) of the chlorides, sulphates, nitrates, and nitrites of sodium, ammonium, strontium, magnesium, copper and iron, as well as those of the dilute acids themselves, on the frog’s heart. It was considered quite probable that the acid radicle might have an important bearing in determining the influence of the base on the frog’s heart. Another object was to determine the difference, if any, in the relative actions of the nitrates and nitrites on the heart. The extensive use of nitrates in the preservation of meat is well known. At the present time the general action of nitrate and nitrite radicles on the body as a whole and on special physiological functions is of considerable interest and must soon be carefully investigated. In the experiments which are reported in this paper the procedure was as follows: A frog was pithed and tied to a frog board. No curare or other drugs were used. Care was taken not to injure more structures than necessary and to prevent unnecessary loss of blood. All curves were made with the heart im situ and with nerves, etc., intact. A delicate wire attached the apex of the heart to a fine counterbalanced heart lever, and the beats were recorded on a slowly moving kymograph on which the time curve was recorded simultaneously from a contact metronome beating seconds. No rec- ords of the heart beats were made until several minutes after pithing the frog in order to allow time for recovery from “ shock.” Twenty to thirty seconds were allowed for the recording of normal heart curves. Then 2 c.c. of solution were allowed to flow over the heart from a finely drawn-out pipette. The time of flow was approxi- mately one minute. Immediately after the application of the dilute solutions of the salts or acids, the excess was removed with filter paper and the heart was bathed with normal saline solution (0.6 per cent) from a pipette in the same manner as the testing solutions. One per cent solutions of the salts and twenty-five hundredths of one per cent solutions of the acids were used. The curves for each experiment were divided into sections of twenty-second periods, and the number and strength of the beats were determined for these periods before, during, and after the treatment with each solution. The results of the experiments are recorded in the tables below. The effects of the solutions on the rate of the heart will be found The Effects of Chloride, etc., on the Frog’s Heart. 265 in Table I, and those on the force of the beat in Table II. In each case the results indicate the specific action as determined by counts TABLE I. THE ACTION OF SOME COMMON SALTS ON THE RATE OF THE BEAT OF THE FROG’S HEART. No, Inc. (brief) Inc. Dec. Inc. Inc. (slight) Inc. (irreg.) Dec. (slight) Inc. Inc. (slight) Inc. TABLE I. So, Inc. (slight) Inc. Inc. No change Inc. (irreg.) Inc. Dec. (irreg.) Dec. Cl Inc. (slight) Inc. Inc. Inc. Inc. Dec. & inc. No change Inc. (slight) THE ACTION OF SOME CoMMON SALTS ON THE FORCE OF THE BEAT OF THE FROG’S HEART. No change Inc. Inc. (slight) Inc. Inc. Inc. (slight) Dec. Dec. (strong) and measurements. Inc. (slight) Dec. & inc. Inc. Inc. (slight) Inc. (slight) Inc. Dec. & inc. No change Inc. Inc. Dec. No change Inc. (slight) Inc. (slight) Dec. (slight) Inc. (slight) Dec. & inc. Since the heart rate varied so greatly in the different frogs, no comparisons of the average rates for the differ- ent salts could be made. 266 FOG: iGoake: All the salts of copper which were studied acted as stimulants by increasing both the rate and the force of the heart beat under the conditions of the experiment. In the case of the nitrate this action was not so marked on the rate, and on the force there was so slight a change in all the experiments that it could not be considered of any value. That dilute solutions of copper sulphate act as stimulants to the heart is well known, and this stimulating action is shown here both on the rate and on the force of the action. With the exception of the chloride, the salts of strontium in- creased both the rate and the force, while in the case of the chloride the rate was increased, the force being decreased. These results, it will be noted, are not strictly in accord with the statements of Wood,!! who says that the strontium salts produce at first a stimula- tion of the heart muscle followed by a slowing and paralysis. These primary and secondary actions were not found in my experiments, but it is, of course, possible that the long-continued action of strontium is accompanied by the later depressing effects noted by that author. The salts of magnesium showed a varied action. The sulphate and chloride increased the rate, the sulphate increased the force slightly. With the nitrate the rate was decreased, the force being slightly increased. The experiments of Macnider and Matthews! showed that the chloride or sulphate of magnesium injected into the circulation of the dog had a depressing effect on the heart, slowed the heart rhythm, and decreased the contractions. Many other investigators have reported similar results, but in their experiments there was a much greater effect in these directions, possibly because of the more intimate relation of the magnesium salt with the heart muscle and nervous tissues through the feeding of the heart by the magnesium- loaded blood. From the accounts of various experimenters the salts of am- monium are considered to be cardiac stimulants, but the results are discordant. It is said that large doses produce a fall of blood pressure, that they arrest the heart in diastole, and that ammonia acts on the heart directly. All the results in the present study have been in the direction of stimulation, although no appreciable change * Woop: Therapeutics, its principles and practice, 14th ed., Philadelphia, 1908. 1” Macniper and Martruews: Loc. cit. The Effects of Chioride, etc., on the Frog’s Heart. 267 in rate was noted when ammonium sulphate was used. The im- portance of the ammonium salts of the blood in regulating or affecting the heart beat is worth noting, for so far as the author is aware little has been done to indicate the value or function of these important blood salts in regulating the rhythm of the heart. In these experiments the salts of sodium had a slight stimulat- ing action, increasing both the rate and the force of the heart. This increase was not marked on either the rate or the force, as will be observed from an examination of the tables. Sodium is the largest single mineral constituent of the blood, and it is doubtful if small doses have any appreciable effect on the normal tissue of animals. It certainly does not depress the system as potassium is said to do. The salts of potassium, as noted in the tables above, are rather irregular in their action. The nitrite markedly increased the rate, but at first decreased and later increased the force. The other salts were irregular in respect to increasing the rates, while their actions on the force were varied. There was often found an initial decrease followed by a secondary increased action. The results obtained in these experiments are in accord with the statements of Wood !* that there is some evidence to show that potassium has a slight stimulating effect on the heart when given in small doses. Wood concludes that such an action is not proven, but it is difficult to explain the results otherwise. On the other hand, we know, how- ever, that in full doses the action of potassium is depressive, both on the heart and the blood vessels. The salts of iron have been little studied in respect to their action on the heart, although in this case there is also the probability, noted above in the case of ammonium, that this constituent of the blood may have an important bearing upon the heart action. In the cases of two of the salts here studied the rate was decreased, and in the case of the chloride so slight a change was noted that it was within the limit of experimental error. With the nitrate a decided decrease in the force was found, while with the chloride a slight increase was noted. The sulphate had no effect upon the strength of the beat. ; The dilute acids showed a diversity of action. The nitric acid increased the number of beats and decreased the force. The sul- phuric acid decreased the number of beats but increased the force. 13 Woop: Loc. cit. 268 Es C.4Goor: The hydrochloric acid slightly increased the number of beats; the force showed an initial decrease followed by an increase. SUMMARY. The action of nitrites on the heart has been sparingly studied, and their action is not well established. With the small doses which were used a slight stimulating action was noted. Large doses, however, are said to depress the cardiac muscles as well as the vaso-motor system. From the limited number of experiments here recorded the in- dications are that nitrates, with the exception of magnesium and iron, increase the rate of the heart beats, and that the force of the heart beat is increased save in the case of iron and hydrogen. All the sulphates, with the exception of iron and hydrogen, in- creased the rate, and in no case was a decreased force noted. Of the chlorides all increased the rate, although potassium was found to be irregular in action. All but strontium and potassium increased the force; in these two cases there was a decrease. This work was begun in conjunction with Mr. E. W. Boughton, but the results recorded in this paper were obtained after the pres- sure of other duties had compelled his withdrawal. The writer wishes to express his indebtedness to Professor Shepherd Ivory Franz for suggestions and assistance throughout the progress of the work. PROUAN TITATIVE STUDY OF FARADIC: STIMULA-— TION. —IIl. THE MEASUREMENT OF “MAKE” SHOCKS! By E. G. MARTIN. [From the Laboratory of Physiology in the Harvard Medical School.] LL the first paper of this series? I set about the task of developing a plan whereby the physiological efficiencies of faradic stimuli might be expressed in terms of stimulation units. In that paper were stated the various factors which determine the physiological intensities of induction shocks. The factors of primary importance were shown to be the construction of the inductorium, the position of the secondary coil with respect to the primary, and the intensity of the primary current. Inasmuch as Helmholtz? had shown that the relationship of these factors to “ break ”’ shocks is a compara- tively simple one, the first step was the development of a scheme of calibration for “break” stimuli which should take into account the influence of these three fundamental factors. This calibration was presented in the second paper of the series.* The present paper contains the second step in the general plan, a method for measuring the influence of these same three factors on “ make” shocks. From the outset I desired to base the method of measuring “make” stimuli, if possible, upon the calibration which had been proposed for “break” shocks. From observations reported in the first paper of the series * it was clear that that calibration would not express directly the relationships existing between ‘‘ make ” shocks as it does those between “ breaks.” A general consideration of the physical principles underlying the production of the two sorts of induced currents led me, however, to believe that a simple mathe- matical relationship between the intensities of “ break” stimuli and 1 Martin: This journal, 1908, xxii, p. 61. ? HELMHOLTZ: PoccENDorr’s Annalen der Physik und Chemie, 1851, Ixxxiii, Pp. 505. 3 Martin: Loc. cit., p. 116. “ Martin: Ibid., p. 68. 269 270 E. G. Martin. of “ make” stimuli might be shown to exist, and a formula deduced for expressing the latter in terms of the calibration already proposed for the former. Finding myself unable to make any progress toward such a formula through study of the principles of electro- magnetic induction, I turned to experimentation in the hope that an empirical formula could be established which should fulfil the purpose desired. The method of experimentation was essentially the same as that used in the study of “ break” shocks,® the index of the value of the stimulus being the minimal contraction of a frog’s gastrocnemius muscle uncurarized. With the onset of hot weather a change in the method of applying the stimulus to the muscle was adopted in the ‘hope of improving the ability of isolated muscles to retain uniform irritability. Instead of stimulating by means of platinum needles thrust directly through the stripped muscle a specially arranged nerve-muscle preparation was used. The entire leg with skin intact was removed from the body. The gastrocnemius was separated in its skin according to the usual laboratory method. A longitudinal incision was then made through the skin and muscles of the ventral surface of the thigh exposing the femur, which was then cut through as high as possible and drawn outward through the skin incision to serve, when placed in a clamp, as support for the entire preparation. A little blunt dissecting in the cavity left by the withdrawal of the bone revealed the sciatic nerve. This was slipped across the ter- minals of a pair of small shielded electrodes. The thigh muscles were then returned as nearly as possible to place, care being taken that they should protect the nerve completely from exposure to air. The leg with shielded electrodes in place was supported in a moist chamber in such fashion as to allow the femur to be placed in a clamp and the gastrocnemius to be attached to a recording lever. This preparation when made with care could usually be depended on to maintain uniform irritability for two or three hours in the hottest weather. In most instances the threshold value of the stimulus applied thus through the nerve was lower than the average obtained from direct stimulation of the muscle. The difference, however, was not so marked as might have been expected. The results given by these preparations were perfectly concordant with those obtained from muscles stimulated directly. 5 Martin: Loc. cit., p. 117. A Quantitative Study of Faradic Stimulation. 271 APPLICATION OF THE “ BREAK”? CALIBRATION TO “MAKE” STIMULI. The problem before me was to determine by experiment whether under a given set of conditions for generating induced currents a definite mathematical relationship exists between the physiological intensities of the “break” and “make” shocks. In a former paper © it was shown that for equal “ break” stimuli the product 1 depends on the position of the secondary coil with respect to the primary, and / being the intensity of the primary current in amperes. This constant represents the value of the stimulus and will be called Z,* the general formula for “ break” shocks then being: M MPs. eee ee Lor — X I is constant, L being the “ calibration number ” whose value M f Z, = Ee 5 {1) My first procedure was to obtain a series of equal “‘ make ” stimuli with the secondary coil at various distances from the primary. The “calibration number” for each secondary position was then multi- plied by the intensity of primary current employed at that position, and the products for each experiment were set down,in a table. Three such experiments are quoted in Table I. For the inner positions of the secondary coil, positions which have relatively large values of Z the product < I is nearly constant; as the secondary coil is moved out into the parts of the field where the values of a are small, the product ” x I is progressively larger the farther out the secondary coil is pushed, and consequently the smaller are the values M Ae sous of —. Numerous repetitions of the experiment gave precisely simi- 1B lar results. These experiments indicated quite clearly the existence of a com- paratively simple relationship between “ make” and “ break”’ stimuli, and also suggested a method for expressing the relationship mathe- F ® Mart: Loc. cit., p. 132. 7 To distinguish between “break” stimuli and ‘‘make”’ stimuli the former will be represented by Z;, the latter by Zm. 272 E. G. Marti. matically in the simplest possible fashion, namely, through the intro- duction of a single factor into the “break” shock formula (1), which when introduced would cause it to give equal values for Z for equal “make” stimuli. It is-obvious, from a study of Table I, TABER VALUES OBTAINED WHEN THE PRIMARY CURRENTS GIVING EQuatL ‘‘MAKE” STIMULI ARE MULTIPLIED BY THEIR CORRESPONDING CALIBRATION NUMBERS. COIL B. Exp. of April 18, Exp. of Nov. 5, Exp. of Nov. 18, fi 19072 EMD ap: 1907. EM. D. BP. (1907. Ea Mopars Position I of primary current= | of primary current= | of primary current? of sec- y 8.8 volts. 4 volts. = 5 volts. ondary er Val. of I. | M Val. of I. Val. of I. | inamperes.| ZL x I. in amperes. * |inamperes.| 7 Kole 0.00053 0.00185 0.00250 0.00603 0.0138 0.0297 0.110 0.335 0.475 that the factor to be introduced must be relatively larger the smaller is the value of a and must tend to diminish M A constant number has this effect if it is subtracted from a Formula (1) modified in accordance with this idea becomes Zn = (F- K) i +@) It was found that in practically every experiment of a large series some number could be selected to be substituted for K in formula (2) with a fairly constant value of Z resulting. For each experi- ment the value of K had to be determined empirically, and it was A Quantitative Study of Faradic Stimulation. 273 found to vary widely in different experiments. In all the earlier experiments the values of K were negligibly small in comparison with the values of - for secondary positions of 12 cm. or less. Therefore, in order to save time, most of the later experiments were begun with the secondary coil at 12 cm. Some typical experiments illustrating the application of formula (2) to equal “ make” stimuli are given in Table II. TABLE II. EXPERIMENTS SHOWING THAT Equa ‘‘MAKE” STIMULI GIVE EQUAL VALUES OF Zn, M WHEN THE LATTER IS COMPUTED ACCORDING TO THE FORMULA Z,, = Ges K) Te Coir B. Position of sec- ondary in cm. Exp. of Nov. 5, 1907. Pri. voltage =20. K = 5.5. Val. of I in am- peres. 0.00205 0.0086 Exp. of Oct. 14, 1908. Pri. voltage = 4. K=18. Val. of I in am- peres. 0.00125 0.0059 Exp. of Oct. 29, 1908. Pri. voltage = 10. K=12. Val. of I in am- peres. 0.0023 Exp. of Jan. 4, 1909. Pri. voltage = 2. K=22. Val. of I in am- peres. 0.0012 0.0053 0.0108 0.0185 0.0285 0.048 0.080 DEVELOPMENT OF A GENERAL FORMULA FOR “ MAKE” SHOCKS. The discovery of formula (2) is a decided step toward the ulti- mate solution of the problem of measuring “ make ’”’ shocks, but it is not a complete solution, since it offers no means of determining 274 E. G. Martin. in advance what the value of K will be under any given set of con- ditions. The next thing done was to study a large series of experi- ments with reference to the conditions upon which the walues of K depend. That the voltage of the primary current has great influence upon the stimulating values of “ make” shocks was stated in a former paper * and is demonstrated in Table III, where it is shown that to TABLE III. EXPERIMENTS SHOWING THAT TO OBTAIN EQuaL ‘‘MAKE” STIMULI WITH VARYING PRIMARY VOLTAGES CONSIDERABLE COMPENSATORY CHANGES IN PRIMARY AMPER- AGE ARE NECESSARY. EXPERIMENT OF Nov. 19, 1907. SEcoNDARY Coit aT 16 cM. = = 250. Primary E. M. D. P. in volts. Primary current in amperes.| 0.07 0.0199 0.0188 EXPERIMENT OF Nov. 2, 1908. SECONDARY AT 22 cM. 4 = 64. Primary E. M. D. P. in volts. Primary current in amperes.| 0.097 0.048 0.042 0.038 0.0365 ’ obtain equal “make” stimuli with varying primary voltages con- siderable compensatory changes in primary current intensities are necessary. In order to eliminate for the moment the voltage factor, all the experiments with “make” shocks were divided into groups, each group containing all the experiments at any single primary voltage. The values of K for the different experiments of any group still differed widely, but it was now noticeable that wherever the value of K was large the value of Z was also large and vice versa. This suggested at once a possible dependence of the value of K upon that of Z. To test this possibility the experiments of each group were plotted, values of K against values of Z. The resulting curve in each case is a straight line having the simple equation K= dad. (3) 8 Martin: Loc. cit., p. 72. A Quantitative Study of Faradic Stimulation. 275 <— Fig. 1 gives the curve for coil B obtained by plotting the experiments at 2 volts. The value of a given by this curve is 18. Substituting in equation (2) the value of K given in equation (3), we have al Sees a Ge aie ie 20 Ficure II.— Curve obtained by plotting 0 20-40 Ficure J. — Curve obtained when the val- ues of Z given by the application of the formula Z = i - K) I to the experi- ments performed with a primary voltage of 2 are plotted against the values of K used in these experiments. Ordinates represent values of K; abscisse repre- against the different primary voltages used in these experiments the values of a obtained from curves plotted as in Fig. 1. The equation for this curve is EXa=36. Ordinates represent values of a; abscisse represent primary voltages. sent values of Z. The equation for this curve is K=18 Z. This solved for Z,, and simplified gives M Zn = = : (5) pe ° bP) an equation which enables us to determine the value GL i make stimuli at any given primary voltage, for which the value of a is known. 276 E. G.. Martin. There remains now for the completion of the “make” shock formula only the establishment of a definite relationship between the values of a at various primary voltages and the voltages them- selves. To determine whether such a relationship exists another curve was plotted, primary voltages against values of a previously determined. This curve is represented in Fig. 2. It has the simple equation E-a=C, (6) in which C represents a constant. Substituting in equation (5) the value of a given by equation (6), we have ‘ which is the general equation for “make”’ induction shocks. The value of C is fixed for each inductorium. For the one with which this equation was developed, coil B, its value is 36. The equation here presented is based upon eighty-five concordant experiments in which were used 16 different primary voltages rang- ing from 1 to 49 volts. Some representative experiments are given in Tables IV and V. Application of the general formula for ‘‘ make” stimuli to another in- -ductorium. — As an additional assurance of the correctness of the general formula above presented and also to learn how much experi- mentation is required for the empirical determination of the value of C, some experiments were carried on with an inductorium which in this work will be known as coil A. This inductorium is a recent one, made by Bischhausen Brothers of Berne. It is provided with the Kronecker graduation, which, however, is unfortunately quite useless for quantitative purposes, because the makers of the induc- torium have taken pains to secure the iron core of the primary coil firmly in place. In a former paper ® it was emphasized that under this circumstance the relationships indicated by the Kronecker gradu- ation do not hold. For that matter the graduation was not intended by its originator to apply to this condition.1° This inductorium I calibrated for “ break ”’ shocks according to the method previously ® Martin: Loc. cit., p. 64. 0 Cyon: Methodik der physiologischen Experimente, Giessen, 1876, p. 38r. ea ) 87 9¢'0 Sgr iPr Ses peal apneic Sees lpg a teal) ake? a ne ag S| ee reamed ST €€ ha ata €9'0 €£0'0 aa Ses a i fe | aa hae boa he aS <2: ce a a rae IZ og S LUE SST0 z9°0 $200 anes a 60° SLT0| 18'0 Tr0'0 BOs dee LZ 8Z ~S aS ol lie | ata ; $90 OZOID 9 amc! | tenes as cedie esata 2x) i052} 98Z0'0 BAG ae S¢ 92 = IZ€ 080°0 €9°0 S+10'0 267 680°0 £0°Z 790'0| 08°70 020°0 LOZ 8EZ0 | OF +2 ~S a) ats ees z9°0 OT0'0 98° 9S50°0 ge PI heated ey oho €>10'0 Te'% €01'0 | +9 2 % = t7€ 8£0'0 pA | oi ds 89°7 6££0°0 90° LOO'O\s ake ao eae soz 8s00 | 16 0Z S S ae heey +9'0 $+00'0 aie eerie PEGS gals-5 os eller GOr() $900'0 712 8610'0 Sb SI ne 90°€ SET0'0 19°0 L200'0 CLG +1100 €l'Z GOODE aca made S02 9600°0 | 0sz 91 Lm ee ‘ = Ce eT elie eae | it aaa S67 LS00'0 pais Pecek cell aes otal ance S77 94000 | oEs al > , = % Determining the Position of the Centre of Gravity. 293 Aside from the purely scientific aspect of the work in its bearing on physiology and animal mechanics, it is our belief that it will prove of practical use, especially in the fields of orthopedic surgery and gynecology. We acknowledge most gratefully our obligation to Professor Ira N. Hollis, of Harvard University, for his advice at various stages of our investigation. A NOTE ON THE ABSOREMON VOR EAde By R. H. WHITEHEAD. [From the Anatomical Laboratory, University of Virginia.] T was generally believed for a long time that the fats in foods were absorbed in the form of a fine emulsion. As the result, however, of increasing knowledge concerning the bile and the pan- creatic and intestinal secretions a rival theory soon appeared. It was shown repeatedly that much of the fat was split into the cor- responding fatty acids and glycerol by the lipase of the pancreatic juice, and that these acids united with the alkalis of the bile and intestinal juice to form soluble soaps. Accordingly it was taught by many that ingested fat was absorbed both in the form of an emulsion and in the water-soluble form of soap, neither view ex- cluding the other. The latter view, however, has prevailed in recent years, having been strengthened by various investigations, particu- larly those of Kastle and Loevenhart,! who demonstrated the almost universal presence of lipase in the tissues, and showed that this ferment was able by reversibility of action to synthesize fats as well as to split them; thus furnishing an explanation of the phe- nomena which had led many to believe that the fats after having been decomposed in the intestinal canal are built up again in the villi. Still, although practically all physiologists accept the view that fat is absorbed in the form of soluble soaps, many are reluctant to deny that it may also be absorbed, in part, as an emulsion; be- cause there seem to be no a priori reasons why fine emulsions may not be absorbed, and because the intestinal contents are such as to favor the formation of emulsions. The histological evidence in favor of the absorption of emulsions has been obtained, for the most part, by staining sections of the small intestine with osmic acid a few hours after a meal of fat. In such cases globules of fat ? KastLE and LorvENHART: American chemical journal, 1900, xxiv; LOEVEN- HART: This journal, 1902, vi. 294 A Note on the Absorption of Fat. 295 colored brown-black by the osmic acid are found in large numbers in the epithelium and lacteals of the villi. It is clear, however, in the light of the work of Kastle and Loevenhart that the finding of fat in these situations and under these circumstances does not prove that the fat was absorbed as such — it may have been absorbed in the soluble forms of its constituents and then synthesized in the villi. It occurred to me to test the question by studying sections of the intestine after feeding fat which had been stained before feeding. For this purpose I chose Sudan III, the dye so much used in recent years for the staining of fat in histological examinations; it has been shown by Dr. S. H. Gage? that this dye can be fed to hens with impunity. Sudan II was dissolved in melted butter in suff- cient quantity to give the butter a dark red color. After cooling the butter was fed to a half-grown cat which had been without food for fourteen hours. The animal was killed at the expiration of four and a half hours, and pieces of the small intestine were fixed in 10 per cent solution of commercial formaldehyde. It was noticed that, while the mucous membrane was stained quite red, the lacteals in the mesentery were white. Sections of the intestine were made with a freezing microtome and examined on the slide in glycerine, some of the sections having been immersed for a few seconds in 50 per cent alcohol according to the usual technique. It was found that, while the sections which had not been immersed in alcohol had a diffuse rosy tinge, in both sorts of sections there was entire ab- sence of red globules in the villi, although masses of red butter still unsplit were evident in the lumen of the intestine. In the case of the sections which were dipped in the alcohol even the rosy hue was wanting. These findings indicated that either (1) the fat had not been absorbed at all or (2) it had not been absorbed as such. To test this point the same sections which had just been examined were removed from the slides, stained by the usual technique in a satu- rated solution of Sudan I[I]-in 80 per cent alcohol, and re-examined. Fat was now demonstrated in great abundance both in the lining epithelium and in the lacteals of the villi, in the form of minute red globules. Undoubtedly, then, the fat was absorbed, and it is 2 GacE: Proceedings of the Association of American Anatomists, Baltimore, 1908. 296 k. H. Wiutehead. also evident that it had not been absorbed as such, but had been taken up by the villi in water-soluble forms of its constituents. The presumption is that the fatty acids entered the villi in soaps. For oleic acid dissolves Sudan III almost as readily as does fat itself, and if the red color of the mucous membrane had been due to the presence of this acid stained by the dye, it would not have been decolorized so readily by the slight exposure to the action of weak alcohol. On the other hand, a solution of a soap made with oleic acid and sodium carbonate dissolved Sudan III very feebly. I would call attention, in conclusion, to the fact that the experi- ment is one that can easily be employed for class-work in the laboratory. No attempt has been made in this brief note to review the voluminous literature of the subject, but excellent reviews have been given by Oppel.® 3 OppEL: Lehrbuch der vergleichenden mikroskopischen Anatomie, 1897, ii; and in Ergebnisse der Anatomie und der Entwicklung, 1go2, xii. fae EPPECTS OF BONE ASH IN THE DIET ON THE GASTRO-INTESTINAL CONDITIONS OF DOGS. By ALFRED PEIRCE LOTHROP. [From the Laboratory of Biological Chemistry of Columbia University, at the College of Physicians and Surgeons, New York.| CONTENTS. Page Ja UNWU EU ROR eA RR See ae eS er 297 ite VF CABOUISM Ee XPHRIMENTS, «6 2 oe 50. 6 E (7 days). | eee : : ; Total. | Daily Total.| Daily = | otal Dey Total. _ | av. av. fe av. av. gm gm gm. gm. gm. gm gm. gm. 5.93 | 2.482 | 0.354 |} 4.316| 0.616] 6.80 | 0.971 | 0.112 | 0.016 IROre; area 41.5 | Dosage . .| 136.5 | 19.50 | 4.518 | 0.647 |74.530 |10.650| 7.07 | 1.010 | 0.105 | 0.015 After . . .| 28.5 | 4.07 | 1.810 | 0.258 | 4.090} 0.584) 3.01 | 0.430 | 0.049 | 0.007 Effects of Bone Ash in the Diet of Dogs. 305 3. Second metabolism experiment. — A female Scotch terrier was used in the second experiment and received a daily diet of 135 gm. of meat, 36 gm. of cracker meal, 27 gm. of lard, and 315 c.c. of TABLE IV. SECOND EXPERIMENT. Dairy RECORDS. Fore period. Normal conditions. Number of the day Body weight (kilos) 8.95 | 8.97 | 8.97 | 8.98 | 8.96 Urine: volume (c.c.) 465 | 350 | 365 | 365 | 360 Sulphur of ethereal sulphates (gm.) | 0.0415| 0.0340) 0.0334/ 0.0292) 0.0401) 0.0194|0.0329 Heces-acdryawelght (em.)) 3 3 = .| 0 0 9.0 0 35 AT 62283 II. Dosage period. 9 gm. of bone ash per diem. Number of the day Body weight (kilos) . 8.92 | 8.93 | 8.89 Urine: volume (c.c.) S15» |) 35h) fh S50 Sulphur of ethereal sulphates (gm.) | 0.0165, 0.0165] 0.0251) 0.0134] 0.0194 0.0141 0.0178 Feces: dry weight (gm.) .-...| 8 11 17252 | 'S35 26 BE5ieS!25 Ill. After period. Normal conditions. Number of the day 16 17 Body weight (kilos) 8.84 | 8.84 | 8.82 | 8.89 | 8.92 Urine: volume (c.c.) 398 | 305 | 260 | 355 Sulphur of ethereal sulphates (gm.) | 0.0115|0.0232 |0.0162 |0.0063 0.0181 Heces= dry weight (gm.) - - - .| 18 0 10 3 0 * The urine was lost by accident after it had been made up to 1000 c.c., so that the after period in the quantitative records was shortened to five days. 306 Alfred Peirce Lothrop. water. Five days sufficed to accustom the animal to the cage. The weight remained fairly constant throughout the experiment, being 8.95 kilos at the start and 8.93 kilos at the finish. In the fore period of six days defecation of black, foul-smelling stools occurred on the third, fifth, and sixth days, and the effect of the TABLE V. SECOND EXPERIMENT. URINE. ANALYTICAL TOTALS AND DatILy AVERAGES FOR EACH PERIOD. Volume. Ash. Nitrogen.| Sulphur. PO:;: CaO. Period. =— T’al. D’y av. D’1 av. Daily av. Tal.| PUY | Tal. ay. Total.| Diy bray, Daily | Total. | av. ay. c.c. | c.c. | gm. gm. gm. gm. gm. | gm. gm. gm. gm. gm. .| 2325 | 387 | 10.32] 1.720] 36.42) 6.07| 1.7958 0.2993) 4.49} 0.7485) 0.0918/0.0153 | -| 2107 | 351 | 10.68) 1.780] 36.06) 6.01) 1.9704 0.3284) 4.04/ 0.6732) 0.0792/0.0132 . -| 1963 ,327) 7.84} 1.568] 30.20) 6.04, 1.5350 0.3070) 3.68 0.7370) 0.0785|0.0157 TABLE VI. SECOND EXPERIMENT. FECES. ANALYTICAL TOTALS AND DAILY AVERAGES FOR EACH PERIOD. Weight. Nitrogen. Ash. ‘ Lecithin. Daily Total Daily Daily | Total. ay. ay. i av. gm. gm. gm. gm. gm. 0.197 | 2.295) 0.38 0.036 | 0.006 13.25 | 2.258 | 0.376 | 48.020} 8.00 0.024 | 0.004 0.344 | 12.410) 2.07 0.018 | 0.003 bone ash appeared immediately in the feces on the day after the first dose. During the dosage period there was daily defecation. The discontinuance of the bone ash seemed to have a disturbing effect, for on the first day of the after period there were two diarrheal movements, and on the third and fourth days the feces were still soft, and contained much mucous matter. The reaction of the daily urine was always acid to phenolthalin and alkaline to rosolic acid. It was acid to litmus on all but the last five days Effects of Bone Ash in the Diet of Dogs. 307 of the after period, when it was amphoteric to that indicator. The specific gravity of the urine fluctuated between ror5 and 1018. During the experiment the animal was pregnant, and two sturdy pups were born a month after the close of the experiment. The bone ash had no discernible effects on the mother or the pups. The daily records and analytical data of the second metabolism experiment are summarized in Tables 1V—VI. 4. Third metabolism experiment.— The animal, a small shaggy black and tan male dog, weighing 6.90 kilos, was given a daily diet consisting of 105 gm. of meat, 28 gm. of cracker meal, 21 gm. of lard, and 245 c.c. of water. The dog was under preliminary observation in a cage for twelve days, during which there was apparently no normal defecation. It was suspected, however, that the dog had been defecating but that he was disposing of the fecal matter, for on the fifth day a very small amount of feces (% gm.) was found on the pan. Accordingly, the animal was carefully watched. Meanwhile the daily records and excreta were kept. On the twelfth day a small amount of hard, black, and doughy material was passed. The dosage period was thereupon begun, and the urine samples of that and the five preceding days were taken to represent the fore period. Seven grams of bone ash were mixed with the food of the seventh day, and the characteristic bone ash feces appeared within twenty-four hours. The animal was kept under continuous observation during the daytime, and on the tenth day, about an hour after feeding, defecation occurred and it was noticed that the animal was licking his chops as if he had disposed of a portion of the feces. Although bone ash had been given on three successive days, the fecal matter was dark-colored and foul, and a small amount was purposely left on the wire to test the dog. It was soon ingested. On the following day fecal crumbs on the pan showed evidence of defecation, but there was no material on the wire. On the three succeeding days the excrements were dry and very hard, and were avoided by the animal. The dosage period was continued for eight days, and an after period of seven days followed, defecation occurring on the first, third, and last day of the period. In spite of the fact that there was evident ingestion of some of the feces, the animal did not seem to exhibit any symptoms what- ever. The output of ethereal sulphates showed an increasing amount of intestinal putrefaction up to the time when the fecal 308 matter became hard and dry and was avoided by the dog. Alfred Peirce Lothrop. How- ever, the putrefactive products seemed to produce no toxic effect, THIRD EXPERIMENT. TABLE D VII. AILY RECORDS. I. Fore period. Normal conditions. INumberiof theiday,” =) 2 25 5 = 1 2 3 4 5 6- Av. Body weight (kilos) ...... 6.74 | 6.75 | 6.81 | 6.74 | 6.74 | 6.81 Wxine volume (Cics)i 5. co eau 260° |. 230: | 17/O0| 310) 215. | 80 eZ Sulphur of ethereal sulphates (gm.), 0.0253) 0.0253) 0.0217) 0.030 | 0.0263, 0.0339) 0.0271 Feces: dry weight (gm.) sve 0 0 0 0 0 9.0 1.5 II. Dosage period. 7 gm. of bone ash per diem. Number of the day 7 8 9 10 11 12 13 14 | Av Body weight (kilos) 6.87| 6.92 | 6.96 | 6:97 | 6.95 | 6.97 | 6.92 | 6:90 Urine: volume (c.c.)| 150: | 190 | 170 | 215 | 200 | 190 | 225 | 240 | 201 Sulphur of ethereal sulphates (gm.) .| 0.0274) 0.0369) 0.0306 0.0438 0.0552) 0.0416 0.046 | 0.0425 0.0405} Feces: dry wt. (gm.)|} 0 7 0 14.5 0 17.5 | 15.0 | tae5eee III. After period. Normal conditions. Number of the day 15 16 17 18 19 20 21 | Ave Body weight (kilos) 6.81 | 6.84 | 6.85 | 6.88 | 6.88 | 6.91 | 6.91 Urine: volume'(c.c:) ... .| 235 |, 215 | 220. | 200 | 240) | (2150 > 235eie223 Sulphur of ethereal | sulphates (gm.)..... 0.0386) 0.0288 0.0247) 0.0328) 0.0412) 0.0325) 0.0318 0.0329 Feces: dry weight (gm.). .| 140| 0 | 35 | 0 | o | o | 35 | 3 for the animal was very lively and apparently normal throughout the experiment and never became at all lethargic. The reaction of the daily urine was uniformly acid to phenoltha- lin, amphoteric to litmus, and alkaline to rosolic acid. The specific twsl Effects of Bone Ash in the Diet of Dogs. 309 gravity of the urine fluctuated between 1026 (seventh day) and IO15 (twenty-first day) ; it was usually above 1016 and below 1022. The daily records and analytical data of the third metabolism ex- periment are summarized in Tables VII-IX. TABLE VIII. THIRD EXPERIMENT. URINE. ANALYTICAL TOTALS AND DAILY AVERAGES FOR EACH PERIOD. Volume. ; Nitrogen.} Sulphur. Period. ; Nl aia Diy aval av. 3 yal ropalsl aa ay. ay. c.c. c.c, | -| 1365 | 227 | 7.584] 1.264! 26.72 1.5222] 0.2537 0.0192 1610 | 201 | 9.888) 1.236 33.18 2.4500} 0.3064 0.0314 .| 1560 | 223 | 8.568) 1.224) 29.42 1.6569) 0.2367 0.0228 TABLE IX. THIRD EXPERIMENT. FECES. ANALYTICAL TOTALS AND DaAILy AVERAGES FOR EACH PERIOD. Nitrogen. Ash. Period. Daily ay. Total. Daily V. Total. gm. em, gm. gm. 0.473 | 0.0788 | 9.85 | 0.164 1.520 | 0.1900 | 37.27 | 4.660 | 3.12 0.619 | 0.0884 | 11.13 | 1.576 | 0.93 5. Fourth metabolism experiment. — The animal chosen for this experiment was a young dog of a white mongrel type. The daily diet was as follows: 180.gm. of meat, 48 gm. of cracker meal, 36 gm. of lard, and 420 c.c. of water. When the dog was first con- fined in the cage, his weight was 11.72 kilos, and during the first week it increased about 50 gm. per day, so that at the end of seven days it had risen to 12.12 kilos. During the subsequent week he continued to gain in weight, and it was evident that he was a grow- ing dog. It was finally decided to start the experiment on this 310 Alfred Peirce Lothrop. ascending scale of weight. The dog continued to increase from 12.37 kilos at the beginning of the experiment to 12.91 kilos at the end. During a fore period of four days defecation occurred daily, and the feces were black, diarrheal, and contained much mucous matter. On the fifth day 12 gm. of bone ash were mixed with the food. The feces on the two following days were still soft but yellow and TABLE, X. FouRTH EXPERIMENT. DAILY RECORDS. I. Fore period. Normal conditions. INumberiot theday = ss (==> 1 2 3 4 Average. Body weight (kilos)) V2.0 WAS) 12.34 12.47 12.47 sia Urine volumieg(G:c:) = ascmen colsee 300 320 250 280 287 Sulphur of ethereal sulphates (gm.)} 0.0329 | 0.0277 | 0.0193 | 0.0346 0.0286 Feces: dry weight; (gm)! 5 2. 7 8.5 2 12 7.4 II. Dosage period. 12 gm. of bone ash per diem. Number ofstherday, | ay an eee 6 [7 8 9 10 Av. | Body weight (kilos) .....: 12. 0 12.52 | 12.57 | 12.62 | 12.66 | 12.71| ... Wrines-volumey(E:cs)) eeu eee 275 95 | 320 2/0 | 260 | 275 283 | 2 Sulphur of ethereal sulphates (gm.)| 0.0255) 0. 0191 0.0339) 0.0379) 0.0307) 0.0272) 0.0280 17 16:5) 1/20 16.6 | Feces: dry weight (gm.) ... .| 18 | 19 | 9 III. After period. Normal conditions. Number of the day ...... 11 12 13 14 $/ 15 | Ay. Body weight (kilos). ...... 12.77 | 12.82 | 12.84 | 12.87 | 12.91 Urine stvolumes(c:c))saeaiewreweerr 300 320 330 360 300 322 Sulphur of ethereal sulphates (gm.)| 0.0263 | 0.0255 | 0.0211 | 0.0289 | 0.0293 | 0.0262 Feces: dry weight i(gm:)) vo. e-|) 23%5 0 10 0 4.5 5.6 Effects of Bone Ash in the Diet of Dogs. a0 free from mucus. On the seventh day the feces were formed, and on the ninth day they were hard and dry. Bone ash was admin- istered for six days. During the after period of five days defecation occurred on alternate days, and on the thirteenth day the feces were TABLE XI. FOURTH EXPERIMENT. URINE. ANALYTICAL TOTALS AND DAILY AVERAGES FOR EACH PERIOD. Volume. : Nitrogen. Sulphur. P,O;. CaO. Period. | )P’Y | Total. | Dally |r, Daily | Total, | av. | av. pp] | gm. Tal. | PY T’al. av. s | | gm. | gm gm. gm 3163) 3.09| 0.7727 0.0680, He |) Ge gm. gm. .| 1150 | 287 | 22.94, 5.73} 1.2652) 0. -| 1695 | 283 33.91 5.65] 1.9074 0.3179} 4.06| 0.6771 0.1152 | .| 1610 | 322 30.06, 6.01 Ee 0.3531) 3.47] 0.6949, 0.0720. | | TABLE XII. FouRTH EXPERIMENT. FECES. ANALYTICAL TOTALS AND DAILY AVERAGES FOR EACH PERIOD. Weight. Nitrogen. Daily Total. | Daily | av. av. gm gm. | gm. 7.4 | 2.407 | 0.602 16.6 | 3.283 | 0.547 5.6 | 1.540 | 0.308 again black and free from bone ash, but did not show the diarrheal characteristics of the feces of the fore period. The great utility of bone ash in preventing diarrhea was strik- ingly shown in this animal. The discharges -at first were very watery, and twice the experiment had to be begun anew on account of the mingling of urine and feces. The first dose of bone ash had a marked effect, and on the third day the diarrhea was completely stopped. The reaction of the daily urine was uniformly acid to phenol- 312 Alfred Peirce Lothrop. thalin, amphoteric to litmus, and alkaline to rosolic acid. The specific gravity of the urine fluctuated between 1017 (twelfth « .1 fourteenth days) and 1o21 (first day). The daily records and summaries of the fourth metabolism ex- periment are given in Tables X—XII. 6. General comparison of the results of the four metabolism experi- ments. — A study of the summaries in Tables XIII and XIV, to- gether with the data in Tables I and XII inclusive, warrants the following deductions: Nitrogen balance. — In the first experiment the slight — balance is probably accounted for by the fact that the prescribed diet, as mentioned on page 304, was not sufficient for that particu- lar animal. There was, therefore, a utilization of a small amount of body protein with a decrease in weight during the experiment. The weight was very constant during the after period of the first experiment, when nitrogenous equilibrium was practically estab. lished. In the fourth experiment there was a steady increase ein weight throughout the work, and the excessive plus balances show that a relatively large amount of nitrogen was utilized in building up body tissue. The third animal gained somewhat in body weight; slightly more in the dosage than in the other periods. The posi- tive balances in the different periods accord with that observation. With the second dog the weight was practically constant throughout the experiment (page 306), although there was a continuous nega- tive balance. It is obvious that the administration of bone ash had no partic- ular effect in these experiments on the total nitrogenous metabolism. Urine volume. — In all the experiments there was a decrease in urine volume during the periods of the administration of bone ash, probably because of greater elimination of water in the increased bulk of fecal matter. In all but one case the average daily volume’ in the after period was greater than that in the corresponding dosage period. In only one instance was the average volume in the after period greater than that in the corresponding fore period. In the second animal.the discontinuance of the dosage with bone ash was coincident with a tendency to diarrheal movements and a con- sequent diminution in the daily average urine volume for the after period. The large increase exhibited by the fourth animal in the after period was probably due to the fact that he was a growing dog, and, as he continued to gain in weight, less water was utilized —STe Effects of Bone Ash in the Diet of Dogs. 313 ; TABLE XIII. Tr SUMMARY OF COMPARATIVE DAILY AVERAGES OF THE ANALYTICAL DATA, PER PERIOD, OF THE FOUR METABOLISM EXPERIMENTS. Urine (volume in c.c.). Feces (weight in gm.). Pome eres a ORGS hg | og bog og | Gem av. av Bare. -| 397 || 387 | 227 |- 287 | 324 || 5.93 | 2.83 | 1.50 f 7.38 | 4.41 Dosage . .} 358 | 351 | 201 | 283 | 298 |/19.50 |13.25 | 8.20 | 16.58 14.38 Biter . ~ -| 368 | 327 | 223 | 322 | 310 || 4.07 | 7.00 |3.00 | 5.60 |} 4.92 Nitrogen. Nitrogen. Fore .. .| 6.48 |6.07 | 4.45 | 5.74 | 5.69 || 0.354] 0.197 | 0.079 | 0.602! 0.308 ; Dosage . - 6.16 | 6.01 | 4.14 | 5.65 | 5.49 || 0.647} 0.376) 0.190 | 0.547| 0.440 mecrees — .| O40) 604 | 4.20 | 6.01 5.68 | 0.258 | 0.344| 0.088 | 0.308) 0.249 Ash. Ash. Fore . . -| 1.88 | 1.72 | 1.26 | 1.42 | 1.57 || 0.616] 0.382 | 0.164 | 0.660] 0.455 Dosage . .| 1.65 | 1.78 | 1.23 | 1.49 | 1.53 |/10.650} 8.000| 4.660 | 9.660) 8.240 SeeeeeeeslvG WISZ 1122 | 144+) 149 |) 0.584) 2.070) 1.580} 1:770| 1.500 Sulphur, Fat (ether extract). Fore .. . .| 0.340 | 0.299 | 0.254 | 0.316 | 0.302 || 0.971} 0.493 | 0.283 | 0.905 | 0.663 Dosage . .| 0.327 | 0.328 | 0.306 | 0.318 | 0.319 || 1.010) 0.477 | 0.390 | 0.858 0.684 After . . .| 0.377 | 0.307 | 0.237 | 0.353 | 0.318 || 0.430] 0.687 | 0.133 | 0.610} 0.465 CaO. Lecithin (ether extract). Fore . . .| 0.026 | 0.015 | 0.019 | 0.017 | 0.019 || 0.016] 0.006 Dosage . .| 0.016 | 0.013 | 0.031 | 0.019 | 0.020 || 0.015} 0.004 After . . .| 0.022 | 0.016 | 0.023 | 0.014 | 0.019 || 0.007 | 0.003 Urine. P,05- Sulphur of ethereal sulphates. Fore . . .| 0.850 | 0.743 | 0.532 | 0.773 | 0.726 0.101 | 0.033 | 0.027 0.029, 0.047 Dosage 0.546 | 0.673 | 0.562 | 0.677 | 0.614 | 0.066} 0.018 | 0.040 | 0.028 0.038 After . . .| 0.765 | 0.738 | 0.530 | 0.695 | 9-682 0.054 0.015 | 0.033 0.026 0.032 314 XIV. TABLE After. Alfred Peirce Lothrop. Dosage. +8.38 }+13.40 |+10.55 After. +4.44 Fore (8 days.)| (7 days.)| (4 days).| (6 days.)|(5 days.) Dosage. +4.72 Fore. +2.33 After. — 2.06 Dosage. — 2.09 — 1.37 NITROGEN BALANCE IN EACH PERIOD OF THE FOUR METABOLISM EXPERIMENTS. Experiment. After. Dosage. Fore (7 days.)| (7 days.)| (7 days.)| (6 days.)| (6 days.)| (5 days.)) (6 days.) —0.20 — 0.66 Balance than at the start. Although receiv- ing 420 c.c. daily, the largest vol- ume eliminated on any one day was 360 c.c., on the day previous to the close of the experiment. Ash.— The amount of ash that was excreted in the urine was prac- tically constant; in the last two ex- periments it was almost identical for each period. This fact indicates that there was no appreciable absorption of the inorganic constituents of the administered bone ash, although such an absorption might occur without any corresponding increase in the amount of urinary ash. In two cases there was a decrease and in the two others a slight increase of urinary ash during the dosage period. The amount of fecal ash was, of course, very much greater for the periods during which bone ash was given. The large amount of fecal ash for the after period in three of the four experiments showed that considerable bone ash persisted in the intestinal tract and was not com- pletely eliminated for several days. With the first animal practically complete elimination of ingested bone ash occurred within twenty- four hours after the interruption of the treatment with bone ash. The daily average amount of fecal ash in the corresponding after period was 0.58 gm., practically the same as that of the fore period, — 0.62 gm. Nitrogen. —In each case there was a slight decrease in urinary 5 5 : { Effects of Bone Ash in the Diet of Dogs. 315 nitrogen, and in three cases corresponding increases in fecal nitro- gen, during the bone ash periods. Possibly there was slightly less assimilation of nitrogenous material during the dosage period, and probably, also, at the same time an entanglement of a slightly larger amount of mucous matter by the increased bulk of excre- mentitious material. The urinary nitrogen in each after period was always greater than that of the corresponding dosage period. The fourth dog, which was not in weight equilibrium, but utilizing nitrogenous matter in building up body tissue, did not exhibit the increase of fecal nitrogen shown by the other animals during the dosage periods. The high content of fecal nitrogen for this dog in the fore period confirmed the observation of the presence of much mucous matter in the diarrheal stools mentioned on page 310, the same cause accounting for the high figure in the feces of the after period of the experiment on the second dog (page 306). Sulphur. — (a) Ethereal sulphates. In the first two expert- ments there was a constant decrease in the output of ethereal sul- phates. A minimum was reached during the latter part of the dosage period, showing that bone ash, probably by increasing the bulk of the feces and so causing daily evacuation, kept down in- testinal putrefaction. In the third experiment, where there was evidence of the ingestion of fecal matter, the amounts of ethereal sulphates rose gradually until the middle of the dosage period, when the feces became hard and dry and were avoided by the animal as mentioned on p. 307. Then, as daily defecation was in- augurated and no feces were ingested, the quantitative output of ethereal sulphate gradually decreased again. In the fourth experi- ment the quantitative elimination was practically the same in each period. (b) Total sulphur. The amount of total sulphur showed no consistent regularity. In two experiments (II and III) it was highest during the dosage periods, and in the other two (I and IV) in the after periods. Probably the increase in these two periods was largely dependent on the elimination of the sulphur in the bone ash (page 302). Calcium. — The variations in the output of urinary calcium were hardly more than the normal fluctuations. The quantity in dog’s urine is small at best, and shows no noticeable increase during the ingestion of bone ash, notwithstanding the relatively large amount of calcium present in bone ash. In two cases there were slight de- creases and in two slight increases of calcium output in the urine 316 Alfred Peirce Lothrop. during the dosage periods. In the third experiment there was a perceptible increase not only in urinary calcium excretion, but also in the amount of urinary phosphorus during the dosage period. Probably the ingestion of the fecal material already noted (page 307) brought about conditions which allowed a somewhat greater absorption of calcium phosphate. In that experiment (III) the amount of urinary sulphur was also much higher. Phosphorus. — Except in the instance already mentioned in the discussion of the calcium content, there was in each experiment a marked decrease in urinary phosphorus during the dosage period with a subsequent increase during the after period, the latter amounts, however, in no case quite reaching the high mark of the fore period. The*third experiment shows considerable devia- tion from the other three. The results prove without doubt that there was no appreciable absorption of the calcium phosphate of the bone ash. The decrease in phosphorus output in the urine may be accounted for, as suggested by Steel and Gies, by the probability of a diminished absorption of alkali phosphate from the food. “Such a result might ensue from interaction, in the intestine par- ticularly, between calcium and phosphate with the production of less soluble or precipitated products. Possibly such interaction would occur especially between alkali phosphate and that portion of the available calcium that had been converted into chlorid from carbonate, and which chlorid would be prone, in the intestine, to combine with phosphate in increasing proportion as the mixture containing them became less acid or perhaps alkaline in reac- tion. . . . The total amount of phosphorus (phosphate) in the ex- creta would doubtless be unaffected, for the phosphate withheld from absorption would be passed, within a few hours, as calcium phosphate into the feces.” 1° Fat absorption. — The amounts of unabsorbed fat in the feces indicated only a normal variation. The differences were in no case great enough to warrant the statement that the bone ash ex- ercised any influence one way or the other upon the absorption of fats. In each of two experiments there was a slight increase and in the other two slight decreases in the amounts excreted. In three cases there was a more perceptible decrease during the after periods, suggesting, perhaps, a somewhat more complete assimilation, but 16 STEEL and Gres: This journal, 1907, xx, p. 351. 4279S es : | Effects of Bone Ash in the Diet of Dogs. 217 not more than that of half a gram at the most out of 30 gm. given in the diet. Fecal lecithin. — It seemed desirable to determine the approxi- mate amount of lecithin in the fecal fatty matter in order to obtain, if possible, a suggestion regarding the effect of bone ash on the secretion of bile. Phosphorus in the residue was determined, and the amount of lecithin calculated therefrom with the aid of the figure for the phosphorus content of distearyl lecithin: 3.57 per cent. The results showed that there was no effect of the bone ash on the elimination of lecithin and probably none on the secretion of bile. III. ARrtTIFIcIAL DIGESTION EXPERIMENTS. Although the experiments described on the foregoing pages made it seem quite improbable that bone ash affects unfavorably either the digestive or absorptive processes, the following experiments were performed in order to reveal any possible effects on the activ- ity of the gastro-intestinal enzymes. I. Salivary digestion. — Two cubic centimetres of saliva immedi- ately digested the starch in 18 c.c. of I per cent paste containing as much as 0.4 gm. of bone ash. Preliminary tests with saliva greatly diluted with water to prevent rapid digestion of starch made it evi- dent that ptyalin is markedly restrained by comparatively large pro- portions of bone ash. The following data are among the most significant that were obtained in this connection: A, Conditions: Starch paste, 1 per cent—18c.c. Saliva, 1 in 5 —2c.c. Ss 1 2 3 4 5 6 Bone’ashi(gm.)). . . - - 0 0.025 0.05 0.1 0.2 0.4 Achromic point attained in 4 min. 6 min. 10 min. aa Purple after 14 hrs. 1 hrs. B. Conditions: Starch paste, 1 per cent—13c.c. Saliva, 1 in 50—2 c.c. Jy or 1 2 3 4 5 6 Bone ash (gm.) ... . 0 0),025- 0:05 OzL 0.2 DES Achromic point attained in 1 hr., 10 min. 11 hrs. 24 hrs. Blue, 24 Eeastintee C. Conditions: Starch paste, 1 per cent—18c.c. Saliva, 1 in 20—2 c.c. hea ee eerie 1 2 3 4 5 6 Boncwsh (pm)... . . s 0 0.005 0.01 0.02 0.05 0.1 Achromic point attained in . 20min. 20min. 35 min. 35 min. 36hrs. 42 hrs. 318 Alfred Peirce Lothrop. Slight quantities of soluble calcium salts, such as calcium chlo- ride, seem to have beneficial effects in such experiments, but tri- basic calcium phosphate was found to behave like bone ash. Cal- cium carbonate, while not so effective as tribasic calcium phosphate, retarded markedly the action of the ptyalin in similar experiments under the same conditions. Analogous experiments with solutions of diastase gave similar results. , D. Conditions: Starch paste, 1 per cent—18c.c. Diastase sol., 0.1 per cent —2 c.c (Digestion started at 5 P. m.) INO pk Ws veluistues totter ney ie i! 2 3 4 5 6 iBonezashe(on) tenses. 0 0.01 0.02 0.5 O-L 0.5 Achromic point attained in . 3hrs. Complete Red at Purple Saas at9a.M. 9a.M. at9A.M. Still blue at 9A. M. E. Conditions: Starch paste, 1 per cent—15 c.c. Diastase sol., 0.1 per cent —5 c.c. INOS cn foiet tae, ae meee fet te I 2 3 4 5 iBonerashy(Gm3) ase eee 0 0.01 0.02 0.05 0.1 Achromic pointattainedin . . I1shrs. 2hrs. 2¢hrs. 4hrs. Red 24 hrs. later. Other insoluble substances, such as glass wool, cotton, sand, in- fusorial earth, and barium sulphate, were tried under the same con- ditions, but none of them hindered the action of either ptyalin or diastase. The effect of bone ash seemed to be due to the insoluble calcium salts contained in it. The water soluble material from 0.5 gm. of bone ash was obtained by shaking repeatedly with 20 c.c. of water. After filtration, 10 c.c. of the extract were mixed with an equal volume of starch paste, and 2 c.c. of diluted saliva (1:5) were added. The digestion was not appreciably affected. This seems to show that the interference of the bone ash is mechanical rather than chemical. 2. Peptic digestion. — Experiments with pepsin-HCl were car- ried out, in general, according to the method described by Berg and Gies,!* a brief outline of which follows: Weighed amounts of bone ash were transferred to wide-mouthed glass-stop- pered bottles having a capacity of about 150 c.c. One hundred cubic centimetres of a solution containing 0.2 per cent of hydrochloric acid and 0.1 per cent or 0.3 per cent of pepsin (“‘ Merck, Ph. G. IV’) were then added to each bottle, followed by 1 gm. of fibrin or elastin, which had been dried to constant weight at rr0° C. The bottles containing the mixtures ‘7 Berc and Gres: Journal of biological chemistry, 1907, ii, p. 497. Effects of Bone Ash in the Diet of Dogs. 319 were then placed in a water bath maintained at 40° C. throughout the experiment. Controls were also run: protein alone, and bone ash alone, in comparable volumes of the pepsin-acid solution. Each mixture in a series was subjected, of course, to the same general conditions as all the others. At the conclusion of the digestive period the mixtures were filtered on dry weighed papers which were subsequently desiccated and weighed, giving, by difference, the amount of undigested protein plus any undissolved bone ash. The amount of undissolved ash in the corresponding control subtracted from the total weight of undissolved protein and ash gave the amount of protein residue (undigested matter). In the tests with fibrin aliquot portions of the filtrates were used for the quan- titative determination of the metaprotein.'* To stop digestion at a definite time after filtration was started, the liquid was made alkaline with dilute potassium hydroxide solution. Finally, while hot, the alkaline filtrates were neutralized and then made very faintly acid to lacmoid with dilute hydrochloric acid, in order to keep in solution the calcium phosphate from the bone ash and also to favor the complete precipitation of the metapro- tein. After standing over night the solutions were filtered through weighed papers, the precipitates washed free from saline matters, and the papers dried and weighed. The combined weight of the neutralization precipitate (per 100 c.c.) plus that of the protein residue subtracted from 1 gm. gave the weight of combined fibrin proteoses and peptones. In the case of elastin no precipitable metaprotein is formed in peptic digestion, so that the weight of the protein residue subtracted from 1 gm. gave the weight of the corresponding elastin proteoses and peptones. Observations of the digestive process in these experiments in- dicated that with increase in the amount of bone ash there were also larger amounts of protein residue, due, possibly, merely to the de- crease in the amount of hydrochloric acid by the neutralizing effect of the bone ash. The quantitative data show that in 100 c.c. of the pepsin-acid mixtures, containing not more than 300 mg. of bone ash (which amounts were completely dissolved in the control), digestion of fibrin was not seriously interrupted and the amounts of combined proteoses and peptones that were formed did not vary appreciably. Larger amounts of bone ash, however, by greatly reducing the acidity, appreciably retarded the peptic diges- tive process in the case of fibrin, as was shown by the corre- 18 Hawk and Gigs: This journal, 1902, vii, p. 460; also Gres and collaborators: Biochemical researches, 1903, i, p. 615. 320 Alfred Peirce Lothrop. spondingly larger amounts of residue and by the production of much less material beyond the metaprotein stage. Even larger quantities of bone ash were required to appreciably affect the peptic digestion of elastin. Typical data in this connection are summarized below: A. Elastin. 1gm. 100c.c. of 0.2 per cent HCl —0.3 per cent pepsin solution. Digestive period: 10 hours. Noneiale Ehime ne. wee 1 2 3 4 5 6 Bone ash (mg.) .. 2... 0- 100 200 — .300-- -S00- =a Residue (mg.). ... 2... 849 861 869 862 870 903 B. FibrinI. 1gm. 100c.c. of 0.2 per cent HC]—0.1 per cent pepsin solution. Digestive period: 13 hours. PU Os fous wis ge io cae A racpgee oaNomae 1 2 3 4 5 Boneyashy Gng))\ik eos) aah gel le rome ae 0 100 200 300 500 Residten(mg a. a5 ccs. Ss nee 25 33 31 55 127 Neutralization precipitate (mg.) . .. 264 255 206 243 303 Combined proteoses and peptones (mg.) 711 712 763 702 570 C. Fibrin II. 1 gm. 100 c.c. of 0.2 per cent HCI—0.1 per cent pepsin solution. Digestive period: 1} hours. INGOs ach top. See Soe ees i 2 3 4 5 6 Bone wash: (mg): 1) 6.2 ee oe 0 100 200 300 400 700 Residues (ngs) yom eee ee 72 84 96 127 248 315 Neutralization precipitate (mg.) 149 186 158 206 229 301 Combined proteoses and pep- tones (es) wee eee eae 779 730 746 667 523 384 Twenty milligrams of pepsin, dissolved in about 2 c.c. of water, were added to 5 gm. of bone ash and the thin paste mixed at inter- vals for about four hours. The mixture was filtered and 1 c.c. added to 20 c.c. of 0.2 per cent HCl. A small piece of fibrin, kept in the mixture at 40° C. for an hour, was unaffected. Evidently the alkalinity of the bone ash was sufficient to destroy the pepsin in , the comparatively concentrated solution employed. In the stomach, however, there is normally sufficient acid to immediately neutralize the slight alkalinity of the bone ash, in amounts even greater than those of the doses given in the metabolism experiments. 5 The effect of bone ash on specially dilute pepsin solution was de- . termined undér the following conditions: 20 ¢c.c. of 0.2 per cent HCl containing 0.01 per cent of pepsin (Merck, “ Ph. Go iVege fibrin, 0.1 gm.; temperature, 40° C. The experiment was started at 9 P. M. : Effects of Bone Ash in the Diet of Dogs. 321 After three hours digestion was evident in 1 and 2; it was prac- tically complete in 1 and 2 at nine o'clock on the following morning. In each of 3 and 4 there was an appreciable residue, even at the end of twenty-four hours, but after standing for several days digestion in 3 and 4 was also practically complete, although free acid was absent from 4. 3. Tryptic digestion. —The general method of procedure in the tryptic experiments was the same as in the peptic digestions. A solution containing 0.25 per cent of Na,CO, and o.1 per cent of trypsin (Merck) was used, with fibrin or elastin as the indicator. The weight of the soluble matter in t gm. of bone ash was determined, and from this the corresponding amounts in the various quanti- ties used were calculated. Observations of the digestion as it pro- ceeded suggested that the amounts of residue were practically the same in all the bottles. The quantitative data indicate that arti- ficial tryptic digestion is not materially interfered with by bone ash under the conditions of these experiments. The amounts of elastin residue decreased as the amounts of bone ash were in- creased in a series, but, in the case of fibrin, there was a gradual rise in the corresponding amounts of metaprotein, and a consequent decrease in the quantities of combined proteoses and peptones. The increase in the amount of metaprotein was gradual and fairly con- stant, amounting to an increase of 100 mg. in the presence of 2 gm. of the ash. Typical data are subjoined: A. Fibrin. 1 gm. 100 c.c. of 0.25 per cent Na,CO,—0.1 per cent trypsin solution. Digestive period: 2 hours. KEG. 2 peed ee pa 1 z, 3 4 5 6 aa MesaSH, (UALS) ssa ws o's = 0 100 200 500 1000 2000 BRCCICME THO) is i sx '* | 53 66 46 57 56 72 Neutralization precipitate (mg.) 177 191 201 215 271 275 Proteoses and peptones (mg.) . 770 743 753 728 673 653 B. Elastin. 1 gm. 100c.c. of 0.25 per cent Na,CO,—0.1 per cent trypsin solution. Diges- tive period: 1} hours. INGE Se Seca RE eee 1 2 3 4 5 6 BOnerashe(Mps) lee) ey =e - 0 100 200 500 1000 2000 : ldygon dh 5 6 136 117 92 78 99 79 pee (mg.) | Exp.2.. 116 51 45 44 62 71 4. Pancreatic amylolytic digestion. — For digestive experiments with amylopsin a glycerol extract of sheep’s pancreas was em- ployed. The acidity of the extract, due to acid phosphates, was 322 Alfred Peirce Lothrop. neutralized with 0.5 per cent Na,CO,, which was then added in excess until the reaction became faintly alkaline to litmus. A I per cent starch paste solution was employed, and the observa- tions conducted under conditions similar to those of the salivary digestion experiments (page 317). Ten drops of the extract and 20 c.c. of starch paste were mixed with different amounts of bone ash. The results were very similar to those obtained in the salivary experiments. Small amounts of bone ash had a decided retarding effect, and larger amounts prevented digestion to the achromic point for many hours. The bone ash decidedly interferes with the action of the amylases, seemingly in a mechanical way. The effect on the amylopsin was less marked than that on the ptyalin, prob- ably because of the greater concentration of the former under the prevailing conditions. 5. Lipolytic digestion. — A portion of the glycerol pancreatic ex- tract that had been prepared for use in the amylolytic experiments was also employed for the lipolytic experiments. The acid reac- tion of fresh milk was neutralized with Na,COs, and then the milk was colored strongly blue with litmus solution. Three cubic centi- metres of the pancreatic extract were added to 25 c.c. of milk in sev- eral bottles containing different amounts of bone ash. The bottles were kept in water at 40° C. and the colors compared from time to time. If under these conditions an extract displays lipolytic power, there is a gradual change from blue to red, caused by the hydrolytic production of fatty acid from milk fats, the intensity of the red coloration indicating, the relative speed of the reaction. It was very difficult to distinguish between the shades of color in the control tubes and those containing 0.1, 0.2, or 0.4 gm. of bone ash. The tubes containing 0.8 and 1.6 gm showed a distinct bluish color. The tube containing 1.6 gm. remained as blue as the original milk for a long time. On long standing there was a slow change of color in all the samples. The action of the lipase was not appreciably interfered with by moderate amounts of bone ash. It is a question whether the larger amounts of bone ash in- terfered chemically by neutralizing the fatty acids as they were formed, or merely mechanically, as appeared to be the case with the amylases. 6. General comparison of the results of the artificial digestion ex- periments. — The action of ptyalin, diastase, and amylopsin in dilute solutions was seriously impaired even by small quantities of bone Effects of Bone Ash in the Diet of Dogs. 323 ash. In comparatively large proportions bone ash retarded the action of each of these enzymes for many hours or inhibited it in each case altogether. Peptic digestion was retarded somewhat by large amounts of bone ash, apparently because of consequent decrease in the amount of acid. Small quantities of bone ash had very little in- fluence on the peptic process. Tryptic digestion was not disturbed, but, on the other hand, seemed to be favored somewhat by the presence of bone ash, probably as a result of the increased alkalin- ity imparted by the bone ash. Lipase hydrated milk fats readily in the presence of bone ash. Even relatively large amounts of bone ash did not appreciably retard the lipolysis. IV. SuMMARY OF GENERAL CONCLUSIONS. Moderate quantities of bone ash mixed with the food of dogs failed to affect perceptibly the reaction or the specific gravity of the urine, but increased the bulk, and frequency of elimination, of the feces. Such doses of bone ash appeared to reduce the vol- ume of the urine by about the volume of the extra amount of water that was eliminated at the same time in the correspondingly greater bulk of feces. Diarrhea was immediately greatly reduced or stopped entirely by dosage with bone ash. Moderate doses of bone ash failed to affect appreciably the elimination of total urinary inorganic matter, but the ash from the feces was correspondingly increased in amount. Calcium elimina- tion by the kidneys was unaffected by the bone ash treatment, but the excretion of phosphate in the urine was slightly diminished thereby. General protein metabolism, as registered by the nitrogen bal- ances for successive periods, was not materially affected by the amounts of bone ash administered in these experiments. The data for urinary sulphur and phosphorus were practically in accord with this conclusion. Urinary nitrogen and phosphorus were decreased somewhat during the dosage periods, whereas the amounts of these elements in the feces were increased in equivalent absolute amounts during the same periods. Although protein assimilation was perhaps slightly diminished, the reduction, on the most radical assumption the data would per- mit, must have been trivial. Absorption of both fat and protein 324 Alfred Peirce Lothrop. seemed to be unaffected and the excretion of bile unmodified. The elimination of mucus appeared to be increased slightly by me- chanical influences. Intestinal putrefaction was perceptibly dimin- ished by the bone ash dosage, doubtless as a result of the more frequent evacuations of the bowels that were induced. Although the action of various amylases was seriously affected by bone ash in the artificial digestion experiments, the two pro- teases that were studied im vitro were not unfavorably influenced when contained in solutions of normal reaction. Tryptic action seemed to be quickened by the bone ash. Lipase likewise was ren- dered more active im vitro, 1f it was at all affected by the ash. The results of the artificial digestion experiments warrant the hypothesis that large quantities of bone ash do not unfavorably influence the intestinal enzymotic processes, and that such doses of bone ash have little or no effect on gastric digestion. Such a hypothesis regarding the normal gastro-intestinal digestive proc- esses is further emphasized by the observation that none of the many dogs subjected to bone ash dosage in this laboratory ever indicated any symptoms of indigestion. The observations recorded by Steel and Gies '° in some of these connections have been confirmed in every instance. The use of bone ash in the diet of dogs seems to offer no me- tabolic disadvantages whatever. The advantages in its use have been referred to in the introduction to this paper and were men- tioned in some detail by Steel and Gies. In conclusion I desire to acknowledge my gratitude to Dr. Wil- liam J. Gies for his counsel and guidance throughout this work, and to whose interest and advice I owe its completion. 19 STEEL and Gies: Loc. cit. AN IMPROVED METHOD OF DESICCATION, WITH SOME APPLICATIONS TO BIOLOGICAL PROBLEMS By L. F. SHACKELL. [From the Physiological Laboratory of the St. Louis University School of Medicine.]' HE present communication embodies the results of an attempt made more than a year ago to shorten the time of vacuum desiccation needed to obtain the percentage of water in various materials. The resultant method has proved much more widely applicable than was anticipated in the beginning. The numerical data and applications to analytical chemistry presented in the first part of this paper are therefore incidental to the use of the latest modification — the essential of which is the desiccation of materials in the frozen condition —in the fields of therapeutics and pure biology subsequently mentioned. No attempt has been made to detail the work done along the latter lines, since a series of ex- tended researches involving the present method is in progress. Vacuum DESICCATION. The use of a vacuum has been recommended for many years in the drying of substances which would be changed in composi- tion by heating to the boiling-point of water. In general, no rule has been prescribed as to the degree of attenuation required. In a recent article on the determination of water in foods, etc., Bene- dict and Manning? have shown that the extent and rate of desic- cation of such materials is largely dependent on the rarefaction of the atmosphere within the containing vessel. These authors have proposed a “ chemical”? method for obtaining a high vacuum, which consists in vaporizing a few cubic centimetres of ethyl ether 1 The writer collected the numerical data given, and developed the present method, while Assistant Chemist to the Missouri Agricultural Experiment Station. 2 BENEDICT AND MANNING: This journal, 1905, xiii, p. 309. 325 320 L. F. Shackelil. in a tubulated desiccator by evacuating the latter with an ordinary water suction pump. The ether vapor displaces the air and sub- sequently dissolves in sulphuric acid contained in the desiccator. By this means, if proper attention is paid to the lubrication of the vessel, an air pressure of less than 5 mm. of mercury can easily be secured and maintained for practically an indefinite period. Using the method as just outlined, Benedict and Manning have determined the water content in various air-dried materials, or in substances containing less than 20 per cent water; “ the agreement of duplicates was striking in all cases. With animal materials (col- lagen, ossein, gelatin) in general the drying operation is complete at the end of two weeks.” With but slight modification the method was used for nearly a year in the chemical laboratories of the Missouri Agricultural Ex- periment Station with uniformly satisfactory results. Instead of air-dried substances, however, the material dried — fresh animal tissues — contained usually 60 per cent to 80 per cent of water. The writer early noticed that drying was much facilitated by rotating the desiccator, since otherwise the surface of the sulphuric acid soon became saturated with water, and desiccation from that point was coincident with the diffusion of the water throughout the concentrated acid below. This fact suggested a series of ex- periments to determine the minimal time required to completely dry various materials. At this time the attention of the writer was brought to the “ Geryk”’ vacuum pump,’ with its possible applica- tion to the problem at hand. This pump will lower the air pressure in an ordinary desiccator (2-3 litres) to a small fraction of a milli- metre of mercury within two minutes. Where water pressure is variable, this pump is a much more reliable source of high vacua than the ether-water-pump method, though it is believed that the analytical results which were obtained with the aid of such a pump, and are presented in the following tables, could have been dupli- cated under proper conditions with the method of Benedict and Manning. There are, however, several factors which tend to lessen the value of the latter method in water determinations. Where it is necessary to run hundreds of such determinations at the same * IT am indebted to Dr. Hermann Schlundt of the University of Missouri for sug- gesting the use of this pump, which does much to simplify the method to be described. A very efficient form suitable for hand or power can be obtained from the Cleveland Lamp’ Company, Cleveland, Ohio. An Improved Method of Desiccation. 327 time and many desiccators must be evacuated, much valuable time is lost in waiting for the complete volatilization of the ether. Car- bonization of the latter which is absorbed by the sulphuric acid gradually takes place. By reduction of the sulphuric acid sulphur dioxid is produced, and this is absorbed by many substances, espe- cially such as contain much fat. For the same reason the use of commercial sulphuric acid is questionable because of the sulphur dioxid dissolved therein. Without going into detail, the following modifications of ordinary vacuum desiccation were used to obtain the analytical data published herewith, namely: (1) The rapid pro- duction and maintenance, by means of the pump mentioned, of a very high vacuum, —less than 1 mm. of mercury column; (2) thorough mixing of the absorbing sulphuric acid by semi-occasional rotation, especially during the early period of drying. DISCUSSION OF ANALYTICAL DATA: In Table I are given the results obtained on a sample each of soil, corn chop, air-dried feces, milk, and honey. All the sub- stances were weighed out into crucibles previously ignited and tared, using a tared cover to prevent absorption of moisture by the dry sample during the period of weighing. The honey used was of the consistency of putty, and an attempt was made to dry it without dilution; this failed from the fact that whenever a high vacuum was used the honey foamed exces- sively. Water determinations were, however, run upon the fresh sample of honey by careful manipulation to compare with the per- centages obtained from the diluted material. A weighed amount (14.1102 gm.) of the fresh honey was dissolved in water, and the volume made up to 50 c.c.; 10 c.c. portions — equivalent to 2.8220 gm. of the original sample — were taken in duplicate for the de-_ termination of water. This diluted honey, and the samples of milk, of which the same volume was taken, were absorbed in cotton pre- viously tared with the respective crucibles. In each table, in the first line opposite the description of the sample are given the percentages of water obtained; below these in the case of each sample are given the dry weights. These indi- cate simply the weight variations during drying. In every instance the results of duplicate determinations are given. In this series the first weights were taken after a desiccation of tw2nty-four hours, 328 L. F. Shackell. and every twelve hours thereafter until the samples became con- stant. It will be noticed in Table I that ninety-six hours sufficed to completely dry the materials enumerated. The practically abso- lute desiccation in this length of time of an air-dried feces (steer), as well as of a corn chop, indicates the value of the modifications previously mentioned in the rapid drying of materials containing much cellulose. Benedict and Manning admit that desiccation of such substances by their method requires generally a drying period of more than two weeks. Because of the time required to completely dry the liquids in this series it was thought that the cotton used as an absorbent might not be thoroughly dry, and it was found to be moreover very hy- groscopic. A new series (II) was therefore started, in which mate- rials were used that would perhaps be most affected by drying at the temperature of boiling water. These included samples of levu- lose, butter, cheese, milk, vinegar, and soap. 10.4267 gm. of levu- lose (crystalline) were dissolved in water and diluted to 50 c.c.; 5 c.c. portions of this solution — 1.0427 gm. of fresh sample — and the same volumes of the milk and vinegar were used for the determinations of water. In this series the same containers (cru- cibles) and method of drying were used, except that clean, ignited sea sand was tried instead of cotton as an absorbent for the liquids. From the results, sand has much to recommend itself in this con- nection; it does not spatter during the removal of the air from the desiccator; like cotton, it spreads the liquid, thus facilitating desiccation; it does not absorb a weighable amount of water, and hygroscopicity is limited to the amount of dry substance in the original sample. As a check on the vacuum method, control determinations of water were made on the samples of Series II in strict accordance with the methods prescribed for the several materials by the Asso- ciation of Official Agricultural Chemists.* The results of Series II by the vacuum method are given in Table II, and the results by official methods are given in Table III. Each weighing in Table I] was made after a drying period of twelve hours. Inspection of Tables I and II shows that the mate- rials investigated, which it is believed are quite representative, could be dried within five days, and in the case of some — soil, butter, * See Bulletin 107 (Revised) of the United States Department of Agriculture. = E€ES'6Z 00+S'6Z $945 6Z 6£S9'6Z sy8ay . ; +0°ST 08'bT LS'bT LLOL 0278°z mast TY 4auoy €£98°67 689862 6bL8'6Z Z0L6'62 » stysia\ = 66'+1 6L+1 8S'bT Oz'TI 072787 uaa tog (PIMP) [#ouop] SOI9LZ | OLTI'LZ LSZ9'LZ 6649 LZ 016942 SECL LZ 9OCLL LZ SIYSIO A 1 & 6l'TT 10'IT LL‘01 87'6 86'8 80'8 €L°9 io quaa seg II 4eu0H OTLLL2 | BSLbL2 | 9E8LLZ | OLT8ZZ | sszezz | opsssz | 9o88'zz aeons SU3PM oF gouoy gT II 10'TT yLOT | 956 S76 S78 OTL yu99 Jog S8bT'6Z LOST'6Z COST 6Z 8PST'6Z 06ST'6Z O£9T 62 LS8T 62 Dee S9M oy any 8E'b8 9£'r8 9&8 | — ZEb8 82'b8 has Z0'r8 yuao 19g J ZOIb'9Z 661h'9Z QEth'9Z | BLZH'9Z p67h'9Z 9seh'9z CSSb'9Z one SBE oni 9r'b8 Sh'b8 yrs | «| BEB 9£' 8 Of +8 IT'+8 yua9 Jog ZZ10'8Z €Z10'8Z 6£10°8Z Lb10'82 SS10'8z I1Z0'8Z oom |e eee es 269 769 - = 98'9 €8°9 08°9 09°9 - = uaa Jog é 13°9 $8°9 ; ; : S19 SZSO'E yua9 19g (Aip are) *] sa00,q OfT8'6Z | EFI8'6Z S608'6Z | -SSER'EZ Ebb8'6Z 67862 | 622862 : SWBPM 6Y'ET 8t'eT LS€T 80°ET 16°21 SL2L aarat L662'S W999 TI doy) a0) 910262 £661 62 010Z'6z 8022'62 O1E7'6Z Z8ET 67 ILS7'62 SIUM .. g 2S'€1 LS'€T PST aaa 06°21 slat ard SEL'y | yuao sag 1 POH MOO eace ° $3 / aay ore €S8'0T on ees 9709 bE bSO09 bE S}YSIOM : 68°72 862 9S8¢'6 I [0s quo. 19g An Improved Method of Desiccation. no : ‘a]dures > JO 1YSI9\\ Bet Se A ‘a]dures yo uonduodsacy ‘sInoy] — ‘Sutkip Jo powag ‘NOLLOD NO Gaauosdy saindIyT T1y — ‘NOILvOoIsaq WaAnOVA — ‘saraag Isa I GDIGgVL LF Shacketl: ‘Q[QIONAD BY} WOIZ PUPS a]}II] B JO SSO] OF onp ‘YRIY YNsoyYy , SOST 62 OF TZ $820 TE 95°02 y6ST 62 £17 LSEO TE ch 0% TTLT 62 10 12 LCvO TE 62 02 C89E CE T6 CE S602 0€ 06 GE Sev6 SS 1 60'S ySO6'€S | c6'T 96 F007 62 9b 0c L190 TE v6 61 COLE CE 98°F 9ETT OF S8°CE Oho Ss VI 0906 €S 98'T 8 Eve 62 99°61 Tv80 TE €S'61 SOLE CE T8°CE O8T? OF 6L°CE 8rr6 SS LOT SL06°€S CLT CL O87 62 S6 8T CLOT TE OT 61 6276 SS LOW. 90T6'€S cy Tt POPE 67 €L LT OSPT TE Or 8T cOb6 SS ST LC16 €S co T 09 8r 80bb 62 96 ST 8E07 TE TEL ELOL ES c9 98 ers +S 89°98 O86¢ ZE 99° CE L8ET OF SSE 9056 SS Iv 1 9VT6 €S vIT O8SS 62 LLeT 9E87 TE €8°ST 90E7 CS 6S 66 9162 SS 19 66 VLOL ES 79°98 OCbS "PS 89°98 O8Tr cE CE CE 6687 OF 62 CE CvoL TE ce OT S9€9 0E ve OL 0SS6 SS 66° 9026'€S or 4689 62 Oe TI ey8e Te L6 ET ELEC ZS 85°66 yC67 SS 6°66 T69L°€S 65°98 VES $S 99°98 Te6h CE cy Te POSE OF SV TE Ch6L TE cf OT €9€9 0F ve OL T6S6 SS 6S" O26 ES co 88eeS 986E°S 00S0'S 0050'S OST S OSPT'S 9S97'8 C8SE8 C8So'L 809 LOOT LCvO'T “md 96 ‘sInoyT — ‘sutdip jo potiog ‘ojduies jo YSIO A efi tn gs STUDIO AR sit” eerie USO RIOR II deog vo esR pO Sua a GU Seer irae creDmalye ne deog s]ySI0 es ae (poypusip) “Tp reBour, S}USIO/N teie (payusip) *[ aesour,, Ms pee Co Ley a oe Sess {USSF Ia IT SUN > toss 5. SIUSIOA\ a TRS ATION IO, I AH te SUF 3 MEE II 9994) epee UO Eee tee Senses J 9es994;) et eee EONS co oS ae). ae yi IJ 19}Ng 7 7 7 Tot SayBIeAA . jue Car Gi AATIOON LOST IT tong SIYSIO MA ae qUd9 19g (UOT}NIOS) *T] aso]NAVy] S1YBIOM uounNyos) ‘7 asoynaar Judd 19g (uoRNfos) *] esopnaay ‘ajdures yo uondtosaq “adNVS NO Gaaduosady Sarnory TIV — ‘NOILVOOISHq( WANOVA — ‘Salads CQNOOUS ’ lel HTAVL esiccation. An Improved Method of D €698°ST | Z2Z8°ST | OSZ8°ST | OO8S'ST.| EZ98°sST 69°98 99°98 79°98 | 65°98 cS 98 CESO'ST | T8SO'ST | 6£90°ST | 8Ezo'st | LZ80'ST 19°98 95°98 T$°98 | 1v98 cE 98 OZrS6r | S612 VIPT TS CL IZ | 9017 81 < ae 8b OT a pity 69 8T fs tea 19°91 ie $619'9Z | 8819'9Z ag ZL 9 189 a vS89'SZ | €989°Sz oy 0s°9 ey9 Orbs 6r 16 '1Z VIF TS CL IZ c0S6'FT +966 9ZS9'bT v9 66 TZ06'ST cE 98 SE0T ST cL 98 £065 TE SG EE OIPZ TE vy ee CLIZ 81 Or OT C8EE 8T LS 9T 1229 97 0S'9 1989°SZ vy 9 9189p STSE'E OOT OT OOT OT 0672 OT 062 0T 8IIZ's PSOL'S 6LIb'€ O9bb'€ Lov0'T LCr0'T “ud SIYSI9\A JUDD 19 SIYSIOA, JU9S 19g ‘TI dog ‘[ dros S1YS319,\\ }U9) 19g Q SIYBIOM JUdD Jog (PaTstp) “Ty reBour, (PaTINSIP) “T aeBour, STYSII\\ oe : ° queoasq TNA _ 7 oo F siysiem . JU 19g I ISN SIYSIOM . quad 19g II 2sa0y49 SIYSIOA\ ‘ é Jud. JOG I asa0y9 SIYSION . wud. Og IT 49330 SIYSION . JU9D 1ag Poe S1Y S19 Judd 19g SIYS19\\ JUdD Jog (uoNN]os) “Ty asopnaary (uoNNOS) *] asopnaary « SGOHLAW IVIOMIQ,, Ad ‘ajduies Jo ‘ajdwies jo uondiosacq WYSION NOILVOOISaC, WONOVA NO MOTH — “SATUAS ANOOUS Tl WIaVvL 332 L. FF. Shackell. milk, vinegar, and the like — within a comparatively few hours. In the case of the cheese and soap —the latter giving the only discordant results — dried in the vacuum, the difficulty in abstrac- tion of the last traces of water is explained by the fact that these substances dried to hard masses, and it is probable that traces of water were held quite firmly within the hard, outer crust. Later experiments have shown that if clean, ignited sand is mixed with fresh meat much of the shrinkage of the sample is prevented, and complete desiccation is proportionately facilitated. This is easy to do in the case of meats, cheese, and the like if the container is tared with sufficient sand and a suitable stirrer for thoroughly mixing the former with the sample. Where it is possible, it is suggested that weighed portions of soaps be dissolved in and made up with water to a definite volume, and aliquots absorbed on sand as described previously for honey, milk, ete. Though the data given in the tables are meagre, they still present several points for consideration. An inspection of Tables I and If in which vacuum desiccation was employed shows that in the large majority of cases given exceedingly accurate results, as shown by duplicates, have been obtained. It is well here to call attention to the possibility of obtaining good duplicates where the samples are not thoroughly dry. In this connection the reader is asked to compare the first and last percen- tages given for the corn chop in Table I and for the cheese in Table Il. In each case striking duplication was obtained, though there was still more than I per cent of water in each sample. The obvious conclusion is that drying should always be continued until constancy in weight is obtained; furthermore, to prevent the same conditions from prevailing during drying, duplicates should not be allowed to dry in the same desiccator. In comparing the results obtained by the vacuum with those by “official methods,” it will be seen that in a majority of instances there was a greater loss of volatile material by heating the samples to drive off the water than by volatilizing the latter in a high vacuum. This is simply corroborative of conclusions that many other workers have reached, and shows the necessity of using the vacuum method where anything like absolute values are desired in water determinations. Benedict and Manning have shown the value of a very high vacuum in obtaining completeness of desiccation and accuracy of An Improved Method of Desiccation. 3868 result. To this the writer desires to add his observation that, given a high vacuum, whether obtained “ chemically” or mechanically, the time of desiccation will still be unduly prolonged unless precau- tions are taken to keep the absorbent sulphuric acid at the same concentration throughout by rotation and thorough mixing in the desiccators. The frequency of the rotation will, of course, depend entirely upon the amount of water being absorbed, a little experi- ence enabling the operator to judge for himself. During the first few hours of drying, however, in which the largest part of the water is lost by the samples, the writer finds it necessary to rotate each desiccator for a few seconds every fifteen minutes to half an hour.® As stated in the prefatory remarks, the object of the work just recorded was a reduction in the time required for complete vacuum desiccation. The results show that this has been accomplished, the time element in several cases being shortened to but a small frac- tion of that required for the same or similar materials by the method of Benedict and Manning. This fact, it 1s hoped, empha- sizes the value of the simple expedient employed by the writer of keeping the sulphuric acid homogeneous during the drying period, and should make the vacuum method much more generally used, especially in work of research calibre; even in technical work, where time is an all-important factor, it seems quite possible that this method can be used to advantage for the determination of water in materials like soaps, low-grade sugars, molasses, etc., where the existent methods are entirely empirical and the results often grossly inaccurate. Vacuum DESICCATION OF FROZEN MATERIALS. Previous to the work just detailed, which was completed in July, 1908, the writer had undertaken at the University of Missouri, under the direction of Dr. P. F. Trowbridge, a study of the gly- cogen content in liver and muscle of steers under various condi- tions of feeding. Results were so discordant, however, due to the very rapid post-mortem hydrolysis of the glycogen, especially in the liver, that no definite conclusions could be reached. The rapid 5 Tt is obvious that various types of mechanical rotators can be devised, where the routine determination of water by this method is carried on to any extent. 334 LL. F. Shackell. drying of materials in the high vacuum now reverted the attention of the author to a problem related to that just mentioned, namely, the cause of post-mortem hydrolysis of glycogen. This question was energetically debated for several years by Noel Paton® and Pavy.‘ According to the former’s view, the post-mortem change is due in the main to exaggerated life action of the liver cells, though the formation of an amylolytic enzyme which causes subsequent slow hydrolysis is possible. Pavy maintains, on the other hand, that conversion of glycogen to dextrose under these circumstances is entirely the result of unorganized ferment action. The latter idea is generally regarded as correct; nevertheless it seemed that by the use of the vacuum method of drying a definite solution of the problem could be effected. The line of reasoning follows: 1. The extensive vasculation of the liver, and the fact that by crushing the freshly excised gland unusually rapid hydrolysis takes place, seems to argue that the latter may be due in part to a dias- tatic ferment in the blood, the presence of which has been demon- strated by Bial and others. Rapid irrigation of the liver with physiological saline should obviate this factor, if any such import- ance attaches to it. 2. By drying such bloodless liver in a vacuum, any diastatic enzyme in the gland tissue should be preserved; the practical absence of fat makes it possible, as experiments have shown, to reduce the dried material to an impalpable powder, thus insuring the destruction of any living cells. 3. The glycogen or sugar could be determined in weighed por- tions of the powdered liver after incubation in water for varying periods of time. Any decrease in the glycogen or increase in the sugar percentage would indicate the presence of a hydrolytic agent. There was still left unanswered the question of preventing hy- drolysis during the period of drying. This could seemingly be done only by keeping the substance frozen. Bearing in mind the fact that 1ce may be volatilized in a very high vacuum without passing through the liquid phase, an attempt was made to dry frozen mate- rials without allowing the ice in them at any moment to liquefy. Very gratifying results have been obtained; in fact all experiments * D. Nort Paton: Proceedings of the Royal Society, liv, p. 313; Journal of physiology, 1897, xxii, p. 121; Ibid., 1899, xxiv, p. 36. 7 F. W. Pavy: Journal of physiology, 1898, xxii, p. 391; Pavy and Srau, Ibid. 1902, XXVil, Pp. 452. An Improved Method of Desiccation. 335 to be cited as well as those at present under way are fundamentally dependent upon the rapid and complete desiccation of materials in the frozen condition. PRACTICAL APPLICATIONS OF THE FREEZING METHOD. Though the conception of the present method was due to the glycogenic problem just mentioned, the latter has given way to a study of questions of a more practical nature. For the present paper, a simple statement of results thus far obtained, more ex- tended investigations of which are in progress or projected, will indicate the significance and scope of this latest modification. 1. The mixing of sand with samples to prevent shrinkage and hardening and consequent occlusion of water need now be used only with those that contain too little water to be solidly frozen. Mate- rials containing sufficient water so that they can be frozen to solid masses — meats, fruits, vegetables —can be brought to complete dryness in less time than by any other modification of the vacuum method of desiccation. The reason for this is that the size and physical structure of the substance being dried are preserved prac- tically intact and shrinkage is prevented by its solidity while frozen. It is sufficient to mention in this connection that the writer has samples of beef round that have been dried for nearly four months. No attempt has been made to keep this material from contact with the air or ordinary atmospheric moisture, yet not the lightest de- terioration can be detected. The dried meat is perfectly porous throughout, a fact that accounts for the ease with which accurate moisture determinations can be made. This same porosity very greatly facilitates the extraction of the material with ether for the ordinary determination of fat or extraction by other organic solvents. It may not be amiss to suggest at this point the drying of large samples (100-500 gm.) —this has been done with as much ease as 5 gm. samples — in the extensive problems of animal nutrition that are being undertaken at the present time, and in which analyses of the fresh substance as a rule prevail; the fact that there is an extremely small chemical change, if any, during drying and none whatever afterward if the simplest precautions are taken, makes it always possible to refer to the equivalent of fresh material in case of question as to previous results. 336 L. F. Shackell. It is at times practically impossible to obtain thorough mixture of samples for analysis, as for example the white and gray matter of the central nervous system; with the use of this method such materials can be completely dried, triturated, and mixed in the powdered form to complete homogeneity. 2. The application of this method to immunity work early sug- gested itself. An investigation along this line was begun in Novem- ber last. Diphtheritic or tetanus antitoxin was not then available, but, through the kindness of Dr. J. W. Connaway of the Missouri Agricultural Experiment Station, the writer was supplied with the toxic blood as well as the immune serum of hog cholera. This was for the most part dried while frozen. Unavoidable delays and the difficulty of obtaining proper animals for the tests to be made render it impossible to report at present the results in this case. The fol- lowing experiments make it probable, however, that the present method can be used for the absolute preservation of immune sera as well as the concentration of low-potency toxins: (a) The serum of guinea pigs was desiccated while frozen. Though the dried material has been exposed to sunlight and the ordinary temperature for many weeks, its complement content remains unchanged, as shown by its successful use in the Wassermann sero-diagnostic re- action for syphilis. (b) Dog’s blood caught directly from the fem- oral artery into a receiver surrounded by a freezing mixture was congealed to a solid mass and completely dried without coagulation. On addition of water to a small portion of the dry residue, a typical fibrin clot invariably forms in a few.moments, showing that none of the factors involved in the formation of thrombin have been affected by this method of desiccation. (c) A rabbit’s brain con- taining the fixed virus of rabies, obtained through the kindness of Dr. D. L. Harris, City Bacteriologist of St. Louis, was dried in the frozen condition. An emulsion of a small amount of the dried brain injected subdurally into a rabbit causes typical symptoms of rabic paralysis, with subsequent death of the animal. A matter of considerable importance is the solubility of products of this method. A serious objection to the use of desiccated anti- toxins as now put on the market is that such products do not give clear solutions when diluted. The present method is not open to this objection, for every experiment so far has shown that, if per- tectly clear solutions are frozen and dried, the same clearness and perfect solubility will again obtain on dilution. The turbidity re- An Improved Method of Desiccation. a7 sulting on solution of serum residues when desiccated by ordinary vacuum methods may be attributed to concentration of the original solutions with probable accompaniment of chemical changes, even though comparatively low temperatures have been employed during + desiccation. A striking illustration of this point is the work on rabies just cited. As is well known, the Pasteur treatment of hy- drophobia depends on the attenuation of virulence in the spinal cords of inoculated rabbits when dried for several days over caustic potash. In this latter case concentration is undoubtedly destructive to the virus, as shown by the fact that a cord well dried by Pasteur’s method loses its virulence completely. On the other hand, the same material dried by the present method retains its virulence, because the disturbing factor, concentration, has been removed. The pres- ent method employs the only possible means by which concentration — that is, increase in percentage of solids in solution — can be pre- vented. From a physical standpoint it is interesting to record that drying proceeds from the surface toward the centre, and that when say 50 per cent of the water is removed from a certain material, one half is practically absolutely dry and the other half contains its original percentage of water, still in the frozen condition. 3. A very interesting phenomenon noted while drying blood was that the blood gases were retained quantitatively or nearly so in the dry residue; at least no change in the mercury levels of a differential manometer could be noted during the drying of 40 c.c. of brightly scarlet defibrinated blood. The residue was perfectly soluble in water or physiological saline, the corpuscles having been destroyed at some time during the process. Such a solution sub- jected to a vacuum without previous freezing gave off the gases in the same manner as fresh blood. It has not yet been determined whether elemental nitrogen is present in the dry residue, but the oxyhemoglobin is undoubtedly preserved. TECHNIC OF THE IMPROVED METHOD. To obtain results such as have been given, the following funda- mentals of the technic must be rigidly adhered to: (1) Well-made desiccators; (2) thorough solidification of the material, if freezing is used at all; (3) the production and maintenance of a very high vacuum; (4) the use of a proper lubricant for the stopcock and joints; (5) proper closure of desiccator stopcocks after exhaus- 338 iL NPs Soerell. tion; (6) thorough mixing at intervals of the sulphuric acid in the desiccators. 1. A desiccator whose inner diameter is more than 6 inches cannot safely be used with the very high vacuum now employed. Larger desiccators are almost invariably crushed by the external pressure. It is very important to have the parts of the desiccator fit well. All ground-glass surfaces, especially those of the stop- cock, should be as smooth as possible. 2. Freezing of the material before desiccation needs no discus- sion. The writer has always used an ordinary ice-and-salt mixture. It is obvious that unless thorough freezing is effected, that part of the material still semi-solid will shrink and harden in the same manner as unchilled samples. 3. In work on frozen substances, an efficient pump of the type previously mentioned is indispensable for the rapid production of a vacuum sufficiently high to prevent thawing. 4. In all this work the author has used as a lubricant a mixture of 5 parts by weight of commercial vaseline (petrolatum) and 3 parts of ordinary paraffin. The ingredients are melted together and heated to the boiling-point of the mixture for several minutes. Unless this precaution is taken to insure reciprocal solution of the components, small particles of paraffin are apt to be found through- out the lubricant, rendering it worthless for the purpose intended. The proportion of paraffin may be increased or decreased, de- pendent upon any considerable rise or fall in temperature of the laboratory. 5. Simple turning of the stopcock after exhaustion of the desic- cator has been found insufficient to entirely prevent entrance of air into the vessel; minute pathways in the lubricant are formed in the direction of revolution and allow small quantities of air to enter the desiccator despite additional closure by solid rubber stoppers, etc. To prevent this the writer proceeds as follows: the outer surface of the core and inner surface of the shell of the stopcock are lubricated, the first at its larger and the second at its smaller end. The core is then pushed into the shell until the two films of lubricant meet. This procedure tends to prevent stoppage of the exit tube with the lubricating mixture. After the stopcock is finally closed, the core is simply pressed tightly into the outer shell, thus destroying the pathways formed on closing. By this means it has been possible to maintain the highest obtainable vacuum indefinitely An Improved Method of Desiccation. 320 when necessary. A small manometer is always used within the desiccator to indicate any chance diminution in the height of the vacuum. 6. As detailed in the first part of this paper, thorough mixing of the sulphuric acid in the desiccators is absolutely necessary to prevent saturation of the surface of the acid with water, the latter causing increased tension of water vapor in the vessel sufficient to allow frozen material to thaw, and consequent failure of the experi- ment. Since the vapor tension of water is 4.6 mm. of mercury at o° C., it is obvious that the levels of a differential manometer must never show a difference of more than 3 mm., if it is desired to prevent melting of frozen material. Even after sulphuric acid has absorbed an equal volume of water it is very hygroscopic, and if kept well mixed will absorb water vapor as rapidly as the ice in any frozen material is volatilized. The poor thermal conduc- tivity of a high vacuum, as well as the fact that frozen substances * are cooled to several degrees below o° C. by the rapid volatilization of the ice in them, has made it possible to obtain eminently satis- factory results even in a very warm laboratory. SUMMARY. 1. The improvements in the present method over ordinary methods of vacuum desiccation are (a) the very rapid production of extreme vacua with the so-called “ Geryk”’ type pump; (0D) the freezing of the material prior to desiccation to obviate primarily any concentration of substances, and to a lesser degree shrinkage and hardening; (c) mixing of the sulphuric acid absorbent to pre- vent saturation of its exposed surface with its consequently greatly lessened efficiency. 2. The method affords a comparatively rapid and exceedingly accurate means for determining water in various materials, espe- cially such as can be frozen-solidly. Because of the friability and porosity of such dried materials, the extract obtained by organic solvents can be secured with greater ease than by the usual method of extracting oven-dried materials or those that have been shrivelled and hardened by ordinary methods of vacuum desiccation. Liquids need not be frozen for determinations of water, but can be absorbed on dry sand in proper containers. 340 LES Shacrell 3. Blood dried in the frozen condition retains the largest part if not all of the gases originally in it. To what extent these gases remain in chemical combination or physical occlusion has not been as yet determined. Fresh blood quickly frozen and dried retains its power of coagulation on later addition of water. 4. Experiments so far indicate that all materials, especially those unstable substances associated with immunity work, can be desic- cated as outlined and can be indefinitely preserved. It is generally recognized that no chemical changes take place in perfectly dry substances. The products of this method are for all purposes entirely moisture-free. Deterioration in them is there- fore absolutely precluded, providing the ordinary precautions of stoppered containers are taken to prevent contact with atmospheric moisture. If necessary, substances after drying can be hermetically sealed im vacuo. The special value of the method in medicine lies undoubtedly in the field of serum-therapy. One of the great practical problems has been the prevention of auto-degeneration in serums and toxins. It has been found possible, however, by the use of this method to obviate autolysis of such typically unstable substances as the com- plement of guinea-pig serum and the virus of hydrophobia. Though but a few of the most important results of the method have been mentioned in the present paper, the writer and his col- leagues appreciate the probability of its widely extended application to many current problems, several of which are being attacked at the present time. The writer wishes, in conclusion, to acknowledge his indebted- ness to Dr. E. P. Lyon, whose appreciation of the method and pro- vision of experimental facilities have furnished much of the in- centive for the more recent work here recorded, as well as for the series of researches at present in progress. fir ENELUENCE, OF THE TEMPERATURE OF THE PBART iON THE ACTIVITY OF THE VAGUS IN it TORTOISE. By G. N. STEWART. [From the Laboratory of Experimental Medicine, Western Reserve University.] M. LUDWIG and B. Luchsinger! long ago concluded that in . the tortoise, as in the frog, “not only does the vagus remain completely active at the highest temperatures which the heart can still endure, but it appears rather to have an increased activity at these lethal temperatures.” In an investigation of the influence of temperature and other factors on the heart and particularly on the action of the vagus and cardiac sympathetic nerves? in the frog, I incidentally made some observations on the inhibitory nerves of the ordinary European land tortoise, and concluded that “in general the inhibitory action of the right vagus of that animal is affected by the temperature of the heart in the same sense as that of the vagus in the frog; although the effect seems to be less marked than in the frog, and a much greater change of temperature is necessary to cause a sensible alter- ation in the inhibitory activity of the nerve. At very low tempera- tures it is unquestionably more difficult to obtain complete standstill of the heart than at the ordinary or at a higher temperature. But for a considerable range above and below the ordinary temperature it may be difficult to demonstrate any marked difference. It is by no means easy to show in the tortoise what is seen in the frog, that the minimum strength of stimulus needed to produce a given in- hibitory effect increases as the temperature falls and decreases as the temperature rises. It needs a considerable fall of temperature to appreciably increase the minimum stimulus.” 1 Lupwic and LucusInceR: Archiv fiir die gesammte Physiologie, 1881, xxv, Bp. 2rr. ? STEWART: Journal of physiology, 1892, xiii, p. 59. 341 242 G: N. Stewart. These conclusions have been called in question by E. G. Martin,? working with the terrapin, but confirmed by Bassin,* working in Kronecker’s laboratory with the European land tortoise. Bassin, who does not seem to know of any work published in English, states that the strength of stimulation of the vagus which causes inhibi- tion at the ordinary temperature causes as great an inhibition when the heart is heated to about 40°. Inspection of his curves and protocols shows indeed that complete inhibition may be obtained at the higher temperature with a strength of stimulation distinctly smaller than at the lower. I have not been able to spare time to repeat the work, but, par- ticularly since the appearance of Bassin’s paper, hold this to be unnecessary. With the single exception of Lépine and Tridon,? all observers who have worked with European tortoises are agreed that the inhibitory activity of the vagus is, at any rate, as great at temperatures around 40° as at ordinary air temperatures. Why, then, has Martin reached the opposite result? (1) Chiefly, I think, because he has adopted a different criterion of vagus activity from the other observers cited. I am merely noting this difference, not criticising it. Ludwig and Luchsinger, Bassin and myself, deter- mined a strength of stimulation sufficient to cause complete stop- page of the heart at the higher temperature, and then observed whether stoppage was caused by the same strength of stimulus at the lower temperature; or, vice versa, a strength of stimulation just insufficient to produce complete inhibition at the lower tem- perature was sought and its effect determined at the higher. The stimulation was kept up only long enough to produce a perfectly definite effect. Martin, on the other hand, in his first set of ob- servations, “ selected a strength of stimulus which would hold the heart at a practical standstill for several minutes at ordinary tem- perature, and then the effect of raising and lowering the tempera- tures was observed for this strength of stimulus.” In the second set “ advantage was taken of the fact that the heart of the terrapin can be maintained in standstill by stimulation of the vagus for a number of minutes without the effect of the stimulus becoming appreciably weaker. The heart was inhibited at ordinary tem- 3 Martin: This journal, 1904, xi, p. 387. * Bassin: Archiv fiir Anatomie und Physiologie, 1907, p. 420. ° LEPINE and Trrpon: Mémoires de la Société de Biologie, March 4, 1876. Influence of Temperature. 343 perature, and then, while the stimulus was still on, the temperature of the organ was raised ten degrees or more.” ° Now Martin found in general that at the higher temperature the heart during vagus stimulation gave a larger number of beats than at the lower temperature, or could not be so long prevented from beating. What was observed here was really the capacity of the heart to escape eventually from vagus control at the higher and lower temperatures, not the capacity of the vagus to cause complete, even if transient, inhibition, as in the experiments of | the other writers. The two tests could hardly be expected to give the same result. For if the vagus, acting upon an intracardial inhibitory mechanism whose excitability is increased by increase of temperature, is able to cause inhibition even more easily at high than at lower tempera- tures, it does not follow that the inhibition will be as long main- tained. One would rather expect that the automatic “ contractile energy,’ accumulating more rapidly at the high temperature in the automatic ganglia or the muscular fibres or in both, would sooner reach the threshold at which it overflows in spite of the continued stimulation of the inhibitory nerve. As I pointed out in my previ- ous paper,’ “although a high temperature is favorable to the ini- tiation of inhibition, it is not necessarily favorable to its continu- ance. So far is this from being the case that the standstill obtained by chemical stimulation of the medulla oblongata at the ordinary temperature can be removed by gradually raising the temperature of the heart. And with electrical stimulation of the mixed vagus nerve the standstill, which is more easily obtained at a high tem- perature than at a low, often passes off sooner.” $ Martin also made a few experiments in which he endeavored to see whether the minimal strength of stimulation needed to produce inhibition was altered by altering the temperature of the heart. The results, however, are such as are not easily inter- preted. For example, in one experiment (he cites only two) he got no inhibitory effect at all with the strongest stimulation which could safely be employed when the tem- perature of the heart was 31.5°, although complete stoppage was caused with a much weaker stimulus at 22° before the heart was heated and at 23° after it was cooled again. If this is a typical result, the terrapin’s heart must differ essentially in this respect from that of the tortoise, for I have never seen anything to suggest that a tem- perature of 31.5° will abolish the inhibitory power of the vagus, nor is there anything in the work of Ludwig and Luchsinger or of Bassin which even remotely indicates such a possibility. 7 STEWART: Of. cit., p. 99. 344 G. N. Stewart. Even to this test, however, the vagus action at the higher tem- perature in Martin’s experiments really was much more slightly impaired than would appear on a hasty comparison of the number of beats. at the two temperatures during the period of vagus stimu- lation. For to get the true measure of the extent of the inhibition we must consider the number (and strength) of the beats the heated heart would have executed at the given temperature had the nerve not been stimulated. When, for example, we read that at 29° the heart beat 9 times in three minutes during stimulation of the vagus, whereas at 20° it contracted only thrice in five minutes, we must remember that, had it been left to itself, it would have beat prob- ably not less than go or 100 times at the higher temperature and perhaps not half as frequently at the lower. The absolute amount of inhibitory effect, measured by the amount of work suppressed, is therefore very considerable at the higher temperature. (2) Martin did not use such high temperatures (only up to 31° or 32°) as the other observers (up to 40° or a little more). Now, as already mentioned,.I found that for a considerable range above and below the ordinary temperature it is not easy to demonstrate any marked difference in the minimal strength of stimulus which will cause inhibition. It is, of course, quite conceivable that the excitability of the intracardiac inhibitory mechanism or the capacity of the muscle fibres to respond to inhibitory stimuli may be a dif- ferent function of the temperature at different temperatures in the same heart, and that the curves may not be the same for the hearts of different animals. Further, the relation between the effect pro- duced on the contractile power of the heart and on the inhibitory mechanism by a given change of temperature need not be the same for all temperatures in one and the same or in different hearts. So that, if it is much easier, as I have shown, to demonstrate the progressive change in strength of the minimal stimulus required to produce inhibition for moderate increase or diminution of tem- perature from the ordinary temperature in the frog than in the land tortoise, there may also be a difference between the land tortoise of Europe and the terrapin in this regard. eae Pcie PARATUSSFOR STUDYING THE RESPIRATORY EXCHANGE. By FRANCIS G. BENEDICT. [From the Nutrition Laboratory of the Carnegie Institution of Washington, Boston, Mass.] INTRODUCTION. CAREFUL examination of the qualitative and quantitative changes in the air passing through the lungs furnishes most valuable data for interpreting the nature and extent of the oxida- tion processes in the body. The exhalation of carbon dioxide and the absorption of oxygen have been called the respiratory exchange. When the body is at rest and without food, the exchange is continu- ous, proceeds with considerable regularity, and, 1f rhythmical, the variations are usually fairly constant from day to day. The respira- tory exchange is, however, markedly influenced by the ingestion of food and muscular exercise and to a less extent by many other factors. Of the respiratory products, carbon dioxide has been longest and most accurately studied. The numerous delicate methods for de- termining this constituent in the air, the small amount present in normal air, the rapid excretion of carbon dioxide following mus- cular exercise, all contributed to the general feeling on the part of the earlier physiologists that a knowledge of the carbon metabo- lism was of utmost importance. When we consider that all three of the main organic materials in the body — the proteins, the fats, and the carbohydrates — yield carbon dioxide as the result of their partial or complete oxidation, it is seen that while the determina- tion of the total carbon dioxide output may indicate in a general way the amount of the total katabolism, it of itself cannot indicate in any way the nature of the material burned. To aid in apportioning the katabolism between the nitrogenous and the non-nitrogenous material of the body, it was found that the nitrogen excretion of the urine represented very closely the 345 346 Francis G. Benedict. amount of protein disintegrated. The proportion of nitrogen in the protein molecule is relatively constant, as is indeed the carbon content. From these factors the total carbon resulting from dis- integrated protein can be computed with considerable accuracy. In thus including the carbon of protein in the calculation it became necessary to determine the unoxidized carbon in the urine. This determination even at this date is far from exact, and only too frequently investigators rely upon the relatively constant ratio be- tween nitrogen and carbon in the urine found in healthy men to compute the carbon from the amount of nitrogen determined by means of the rapid and exact method of Kjeldahl. Deducting from the total carbon excretion that computed as be- longing to the disintegration of the protein, we have left the carbon derived from fat or carbohydrate. The apportionment of this remaining carbon between the katabolism of fat and carbohydrate has long been a source of much difficulty in metabolism experiments. For many years for want of direct determination of oxygen it was necessary to assume that during inanition the carbon other than carbon of protein was wholly derived from fat, and the calculations were based on this assumption. The recent researches in inani- tion,t however, have demonstrated that there may be a very con- siderable draft upon glycogen in the body, at least on the first day of fasting, and hence it is wholly erroneous to consider the carbon excretion other than that of protein as carbon of fat. In experi- ments where food is ingested, relying upon the well-known rapid absorption of carbohydrates in the diet, physiologists have usually assumed that the total carbohydrate in the diet was absorbed and oxidized inside of twenty-four hours, and hence the carbon un- assigned to protein has been considered as made up in part from the carbon resulting from the combustion of carbohydrates in the diet. Making due allowance for the amount of carbon that can be derived from the carbohydrates in the diet, the remaining carbon has been ascribed to the fat of the-diet, or, in case the diet was. inadequate, to the fat of the diet plus a certain amount of body fat. If to the determination of carbon dioxide exhaled is added a determination of the amount of oxygen absorbed, the interpreta- tion of the results is very much more satisfactory, for the ratio between the volume of carbon dioxide given off and the oxygen ' Benepict: Carnegie Institution of Washington, Publication No. 77, 1907. Apparatus for Studying Respiratory Exchange. 347 absorbed, the so-called respiratory quotient, gives a reasonably ac- curate indication of the nature of the substance burned. While with the combustion tube or the calorimetric bomb the oxidation of organic material proceeds with perfect regularity and completeness to carbon dioxide and water, inside the human body we have to deal with other conditions. The carbohydrates are, it is assumed, completely oxidized; the fat with normal man is like- wise assumed to be completely oxidized; but, on the other hand, the complex nitrogenous molecule of the protein is but partially disintegrated, and we have therefore of the original protein carbon part disintegrated in the form of carbon in the urine and part as carbon dioxide in the expired air. .In disease, where there may be abnormal metabolism, we have also the possibilities of a partial excretion of organic material in the urine other than that derived from protein. This is noticeably so in the case of B-oxybutyric acid and sugar found in the urine of diabetics. In calculating, therefore, the carbon dioxide result- ing from the disintegration of organic material in the body and particularly in calculating the oxygen required for this disintegra- tion, it is commonly assumed that in health no compounds resulting from the partial disintegration of either fat or carbohydrate are excreted in the urine. As a matter of fact, this is not, strictly speaking, true. The chemical composition of the various ingredients of the body as well as the food stuffs has been determined. There is not the uniform agreement that could be expected to be found when ana- lyzing definite, well-crystallized, organic materials of simple mo- lecular structure; but as the result of a large number of analyses, the percentage composition of body material has been assumed to be that shown in the following table: | | Mineral matters (includ- ing S.) Body material. per cent. per cent, | per cent. per cent. per cent, Reise axe te ek al TOL G7 52.80 | 7.00 22.00 UBL, Pat oe ef ee ee oo ee 76.19 11.80 12.10 Carbohydrate (Glycogen) - - -| - - 44.49 6.20 49.40 348 Francis G. Benedict. For materials in the food, the composition of starch, cane sugar, and glucose can be computed from the chemical formulas directly. For the composition of normal fat, average values given by Koenig? have been chosen, namely, C, 76.65 per cent, H, 11.92 per cent, and ‘Oj sni-4e3epercent! The calculation of the respiratory quotient for substances of simple molecular structure — namely, the starches, carbohydrates, and the fats —presents very little difficulty. Thus, for example, starch has the chemical formula (C,H, )O;),. For purposes of cal- culation the molecule represented by the formula C,H,)O; and not the multiple can be taken without affecting the mathematics in any way. The molecular weight, therefore, may be considered as 162. The chemical reaction may be expressed as follows: C,H,,O; .+- 6 O; = 6'CO7-- 5 EL: In order to oxidize, therefore, 6 atoms of carbon or rather the 72 gm. of carbon existing in, say, 162 gm. of starch, 12 atoms coresponding to 192 gm. of oxygen are necessary. As a result of this oxidation, 6 X 44 = 264 gm. of carbon dioxide are produced. The respiratory quotient deals with volumes rather than with weights, however, and on reducing these values to volumes, as- suming that 1 litre of O weighs 1.43 gm and 1 litre of COs, 1.966, we have, as a result of the combustion of this amount of starch, 34.26 litres of CO, produced and the same volume of oxygen ab- sorbed. Therefore: In order to calculate the respiratory quotient for fat, we can find, from the molecular composition given in the table above, that I gm. of human fat requires 2.844 gm. of oxygen in its combustion and 2.790 gm. of carbon dioxide are produced. There is, therefore, an absorption of 1990.8 c.c. of oxygen to form 1240.4 ¢.c. of carbon dioxide, and hence the respiratory quotient would be CO; . 1249.4 O, 1990.8 =0.713- The calculation for protein is somewhat more elaborate, owing to the fact that it is only incompletely burned, as has been pointed * Korenic: Chemie der menschlichen Nahrungs-und Genussmittel, third edition, i, p. 198. i, Apparatus for Studying Respiratory Exchange. 349 out above. The calculations have been made in a number of ways by different writers on this subject, each assuming a somewhat different molecular composition for the protein, and each ascribing in turn various values to the unoxidized portion of the protein excreted in feces as well as in the urine. Furthermore, there is considerable latitude among various observers as to what degree the sulphur of protein is oxidized, for there is unoxidized as well as completely oxidized sulphur in the urine. The following calcula- tion is, however, taken directly from Loewy,? in which it is as- sumed that 100 gm. of fat-free, dry substance of flesh contain Pea ot ©, 7.27 9m. of H, 22:68 gm: of O, 16:65 em of N, and 1.02 gm. of S. Of these it is assumed that there are found (Ce Ee ©: N. ‘Sy Imethefurine, =. -.. > . . -9.406 2.663 14.099 16.28 1.02 mathe feces. =. 2°: 2 = 1.471 0.212 0.889 0.37 0.0 TPasrereniinye Nem aoe ee oa al 41.50 4.40 7.690 0.00 0.0 By the combustion of 41.5 gm. of carbon and 4.4 gm. of lrydro- gen, there were used 145.87 gm. of oxygen. Deducting from this the 7.69 gm. originally in the protein and not excreted in the urine or feces, there were required from the air 138.18 gm. During the process of oxidation there were formed 152.17 gm. of CO,. Re- ducing these values to volumes, we then have the ratio: COs = 77:39 =—o.8or. OF 96.63 The oxygen required for combustion, the products of combus- tion, and the respiratory quotient for several typical materials have been calculated and placed in the accompanying table(see page 350). It will be seen that the values as used by Loewy differ slightly from those given in the table above, in that the amount of oxygen required to oxidize 100 gm. of protein was 138.18 gm, while in the above table the amount is 1.367 gm. per gram of protein. There is likewise a slight change in the carbon dioxide production. It should be stated, however, that apparently the substance upon which the greatest error falls in calculating the respiratory quotient con- tributes the least to the total katabolism, and hence the error is wholly negligible when the total katabolism is measured. 3 Loewy: OpPpENHEIMER’S Handbuch der Biochemie, 1908, iv, p. 156. Francis G. Benedict. 350 The carbohydrates have a respiratory quotient of 1.00; fats, in . general, of 0.71, and protein of 0.81. From these factors, there- ; fore, we can see that during inanition, when the subject is subsist- - ing for the greater part upon body fat and protein, the respiratory quotient would tend to approach 0.71, and, as a matter of fact, in a long series of observations it has been found that in the later days of a three- to six-day fast, the respiratory quotient is fairly constant at about 0.74.4 On the other hand, if a diet is taken con- sisting in large part of carbohydrates, the respiratory quotient tends to approach unity. RESPIRATORY QUOTIENTS FOR PROTEIN, FATS, AND CARBOHYDRATES. Products of the oxidation of 1 gm. : Oxygen required Respi- tooxidizelgm. | | ratory Materials. | Carbon dioxide. | quotient Water. | | Weight. Volume. | Weight. | Volume. | | | c.cm. gm. em. cm. em. Starch. . =| 21.185 829.3 | 1.629 332), 10.556 Cane sugar 1.122 | 785.5 : 5 | 0579 Glucose . . 1.066 756.2 | 2 | 0.600 Animal fat. 2.876 2013.2 ah O65 Human fat. 2.844 1990.8 4S 1055 Protein . . 1.367 956.9 ; | 0.340 Since the metabolism of the protein remains relatively constant from day to day and from hour to hour and is but a small propor- tion of the whole, the errors involved in its calculation are not of sufficient magnitude to influence seriously any deductions drawn from the results in which these calculations occur. Usually the disintegration of the protein is about 15 per cent of the total katab- olism, and Magnus-Levy® has calculated that if the remaining 85 per cent is wholly from carbohydrates, the respiratory quotient would be 0.971, and if, on the other hand, the remainder of the * BENEDICT: Loc. cit., p. 451. ° Macnus-Levy: von Noorpen’s Handbuch der Pathologie des Stoffwechsels, Pp: 257. Apparatus for Studying Respiratory Exchange. 351 energy is derived solely from fat, the respiratory quotient would be 0.772. Under ordinary conditions the respiratory quotient would lie between these two figures, and values above or below these points might reasonably be considered as due to faulty technique or to distinctly abnormal metabolism or to possible formation of fat from the carbohydrate or carbohydrate from fat. The con- sideration of these points, however, would lead us too far afield in this discussion.® A carefully determined respiratory quotient is therefore of direct value in indicating the nature of the material oxidized. The absolute values for the amounts of carbon dioxide exhaled and oxygen absorbed in a given amount of time, usually one minute or one hour or twenty-four hours, are also of great importance in indicating the quantitative relations of the total katabolism. From the twenty-four hours’ amount especially is it possible to strike a daily balance, and by determining or computing the carbon of the diet the adequacy of the ration for maintenance may be proved. Of still more importance has been the use made by a number of physiologists of the respiratory exchange to compute the total calorimetry by the so-called method of indirect calorimetry. From the determination of nitrogen in the urine (protein katab- olism) and the respiratory exchange, it is possible to apportion the total katabolism between protein, fat, and carbohydrate. It is generally assumed that all of the food materials are first transformed into similar substances found in the body, and the calculations may then with propriety be based upon the values for glycogen, body protein, and human fat. As each of these materials gives rise, when completely burned, to definite amounts of energy, it is customary to multiply the num- ber of grams of katabolized protein by the factor 4.1, fat by 9.54, and carbohydrates by 4.19, and then obtain the total energy result- ing from the oxidation. That this method gives a reasonably accurate indication of the calorimetric transformations in the body is commonly assumed. It is absolutely proved that the method of indirect calorimetry, when applied to experiments of not less than twenty-four hours’ dura- tion, does give accurate results for the total heat production. This has been shown not only with fasting men but likewise with those ® For an exhaustive treatment of this subject, see MAGNuS-LEvy: Loc. cit., p. 218. 352 Francis G. Benedict. consuming food. As a result, then, of the information given by the study of the respiratory exchange, it is clear that a method for this most important study should be perfected which in so far as possible shall be free from errors, be as little complicated as pos- sible, and permit of reasonably accurate and satisfactory results in the hands of others than skilled chemists especially trained to use the apparatus. Recognizing at a very early date the noticeable variations in com- position of inspired and expired air, and foreseeing with remarkable clearness the great advantage of the knowledge of the quantitative relations existing in these transformations, chemists and physiolo- gists began to study methods for separating the inspired and ex- pired air, for determining the changes in composition, and finally for noting the actual amount of carbon dioxide produced and oxygen absorbed in a definite period of time. The methods of studying the respiratory exchange are based upon two distinct principles: First,’ the animal or subject is placed in a chamber, large enough to have him remain with comfort, and the air is there allowed to become vitiated by products of oxida; tion, and the changes 1n the composition of the air from period to period studied; or there is a small movement of air throughout the chamber during the whole time, thus checking to a certain extent the rapid rise of the carbon dioxide content of the air in- side the chamber; or the air is rapidly circulated through the chamber so as to maintain the atmospheric conditions approxi- mately normal. The other method depends upon the separation of the expired and inspired air by means of suitable valves, the respiration being maintained through tubes in the nose, a mouthpiece held between the teeth or lips, or a mask, or, in the case of animals, a cannula in a tracheal fistula. The expired air is measured either in a spi- rometer or gas meter. During the past two decades Zuntz* and his co-workers have made admirable use of an apparatus employing a mouthpiece, valves ot 7 Tn connection with a description of an apparatus of this type a review of the earlier respiration chambers was given in Publication No. 42 of the Carnegie Insti- tution of Washington, 1906. ® For a complete discussion of the development of this remarkable apparatus, see LoEWY: OPPENHEIMER’S Handbuch der Biochemie, 1908, iv, p. 134. f Apparatus for Studying Respiratory Exchange. 353 similar to those of Speck? or of special fish bladders,’ and an Elster wet gas meter.!! Chauveau and Tissot !2 make use of rigid glass nosepieces, me- tallic valves, and collect the expired air in an ingeniously devised spirometer.'* This French apparatus has not, however, been ac- corded the general use given the apparatus of Zuntz. DESCRIPTION OF NEw APPARATUS. This apparatus is the logical outcome of an attempt to apply the principle used in the large respiration chambers in use in this labo- ratory to a movable type of apparatus. The precursor of these chambers has been described in great detail.'* In the large respira- tion chamber a man sits in an armchair or lies on a bed in a cham- ber of sufficient size to allow him to move about with comfort. A current of air is maintained by a rotary blower. As the air leaves the chamber, it contains, in addition to the nitrogen normally pres- ent, carbon dioxide, water vapor, and is somewhat deficient in oxygen, inasmuch as the oxygen has been used for the support of combustion in the body. The outgoing air is caused to pass through a purifying system consisting, first, of a sulphuric acid vessel which removes the water, second, a soda lime vessel which removes the carbon dioxide, and, third, a sulphuric acid vessel for absorbing the water yielded to the current by the moist soda lime; and the air is then returned to the chamber, the deficiency in the oxygen being made up by admitting oxygen from a steel cylinder of highly compressed gas. By noting the increment in weights of the different purifying systems and the loss in weight of the oxygen cylinder, a rough approximation of the total carbon dioxide production and the oxygen consumption is obtained. By making ® Speck: Schriften der Gesellschaft zur Beférderung der gesammte Naturwissen- schaften zu Marburg, x, 1871. See also Physiologie des menschlichen Athmens, Leipzig, 1892, p. 9. 70 Duric: Biochemische Zeitschrift, 1907, iv, p. 68. 1% Loewy: Loc. cit., p. 136. For details of construction, see description by ZUNTz: Landwirthschaftliche Jahrbiicher, 18809, xviii, p. 1; also FLtccr: Hygienische Unter- suchungsmethoden, p. 531. ™ CHAUVEAU and Tissor: Comptes rendus, cxxxii, p. 1532. 8 Tissot: Journal de physiologie et de pathologie générale, 1904, p. 692. 14 ATWATER and BENEDICT: Carnegie Institution of Washington, Publication No. 42, 1905. 354 Francis G. Benedict. due allowance for variations in pressure and temperature of the air in the chamber, variations in the carbon dioxide and water and oxygen contents of the air in the chamber, exact information re- garding the carbon dioxide production and oxygen absorption of the subject during periods as short as one hour can be obtained. This method obviously involves the use of an elaborate respiration chamber, which, as a matter of fact, is fitted with calorimetric appliances and hence cannot be used as a portable or, oncieee, as a semi-portable apparatus. The large respiration calorimeter, although giving exact data regarding the respiratory exchange and the heat production, is un- fortunately not adapted, either by reason of the expense of instal- lation or the technique of manipulation, for general use in most hospitals, clinics, or laboratories. Hence an apparatus which, while not giving necessarily the calorimetric data, will give with great exactness the gaseous exchange and thus permit an approximation of the energy transformations, can be of great service to many laboratories and hospitals. It was planned, therefore, to attempt to so adjust this apparatus that, by taking the expired air from the mouth or nose, the carbon dioxide exhaled and the oxygen actually absorbed could be determined with great exactness. With this end in view, a small apparatus was constructed, using a rotary blower, sulphuric acid and soda lime purifying vessels, and supply- ing the oxygen either from a cylinder of the compressed gas or from the decomposition of sodium peroxide. The whole apparatus is placed upon one small portable table, with an electric motor d1- rectly belted to the blower. Thus the whole apparatus can be moved at will around the laboratory. In a hospital it can easily be taken to different wards and, indeed, to the bedside of a patient. This method, as outlined above, involves no gas analysis, and simply requires that the vessels absorbing the carbon dioxide and the oxygen cylinder should be weighed accurately. Such an apparatus permits the use of almost any form of nasal tube, mouthpiece, mask, or, indeed, a hood covering the entire head. In practice, it has been found advisable in the large major- ity of cases to use special forms of nosepieces described beyond. Occasionally the mouthpiece and nose clip of Zuntz have been used. The subject lying or sitting is fitted with the proper nose tubes, which are in turn connected to a three-way valve attached to the ventilating air pipe of the respiration apparatus. xs al ee eS oper ties oh a , erry fea oe Apparatus for Studying Respiratory Exchange. 355 The course of the ventilating air current is shown diagrammati- cally in Fig. 1. As the air leaves the lungs and passes into the con- stantly moving current of air, it is carried along by a rotary blower and forced through a suitable vessel containing strong sulphuric acid for removing the water vapor imparted to the air by the lungs. The air then leaves this absorber freed from water vapor but containing all Gees the carbon dioxide. For ee EQUALIZER the removal of this gas nt = the air is caused to pass /, —— perSbuce through. finely. divided (= aie soda lime, which absorbs appro BLOWER the carbon dioxide with great rapidity. As has been found from a large ween Jenn oonoe Se number of experiments, I eect ae mae es it is desirable to have Ficure 1.— Diagrammatic scheme of air circuit, and the soda lime somewhat PUT Ee arabe: moist,!® and hence the water vapor taken up by the dry air as it passes through the soda lime is removed from the air by passing the air through another vessel containing sulphuric acid. ‘The air is now dry and free from carbon dioxide, and if passed directly to the lungs would dry out the nasal and throat passages so quickly as to make respiration uncomfortable for the subject; hence it has been found advisable to moisten the air and thus render it respirable. Attached to the ventilating pipe near the point of entrance of the air into the lungs is a pan with a rubber diaphragm called a tension equalizer. Inasmuch as the whole circuit is closed, obvi- ously the inspiration of air into the lungs would cause a decreased tension in the system which would seriously interfere with the respiration were not some proper provision made for keeping the tension always that of the atmosphere. By means of the rubber dia- phragm, there is practically no difference between the external and internal tension; as the air is drawn into the lungs the rubber diaphragm sinks, and as the air is expelled from the lungs the diaphragm rises. The carbon dioxide formed by the process of oxidation in the body is absorbed by the soda lime, and conse- 45 Benepicr and Tower: Journal of the American Chemical Society, 1899, Xxi, p. 396. 350 Francis G. Benedict. quently, on the assumption that the same quantity of air is in- spired as is expired, the total volume of air becomes diminished by the amount of oxygen used in the process of the combustion of both organic hydrogen and carbon. It is necessary, therefore, to admit oxygen to maintain the percentage of oxygen in the air approximately normal, and likewise to avoid a diminished tension on the system in case the oxygen used out of the system was greater in amount than the volume of the tension equalizer. With this apparatus, therefore, it is seen from the diagram that the man simply breathes into and out of a pipe through which a current of air, freed from carbonic acid but moistened_to a com- fortable degree of humid- ity, is constantly passing. Valves are not used. The whole apparatus is mounted on a table. A somewhat diagrammatic FIGURE 2. — Diagrammatic arrangement of LEE representation is given in fon! apparatus, showing HORS tubes for EO, Bice cllye rotary blower tension equalizer, air-purifying apparatus, and S oxysenleymades driven by an electric motor is on the lower shelf of the table, and the air, leaving the lungs by means of the two nosepieces shown in the diagram, is sucked down by the blower and carried in a pipe beneath the lower shelf and introduced into two Wolff bottles containing pumice stone and sulphuric acid. The first of these absorbs a very large proportion of the moisture and the sec- ond is used for precaution. The dry air is then passed through the top of the table by a pipe and conducted to -the can containing soda lime. It then enters into a glass vessel (as a matter of fact, the lower part of an ordinary Kipp generator is used) in the bottom of which is placed sulphuric acid and in the upper section some large pieces of pumice stone to prevent spattered acid from leaving Apparatus for Studying Respiratory Exchange. 357 the vessel. The dry air freed from carbon dioxide is then con- ducted through a pipe into a second Kipp generator, which is, how- ever, supplied with water and with pumice stone moistened with water. A small funnel attached to the side opening in the bottom of the generator permits the addition of water as fast as it evapo- rates. The moistened air then ascends through a tube to the point where it is respired by the subject. Two pet cocks in this pipe permit, the one the introduction of oxygen from the cylinder of compressed gas and the other the connection with a very delicate petroleum manometer. Tension equalizer. — [he tension equalizer consists of a rubber diaphragm fitted to a copper can 16 cm. in diameter and g cm. high. On opposite sides near the bottom of the can are soldered two ordinary hose couplings such as are used on garden hose with an internal diameter of 16 mm. The rubber diaphragm used in these experiments is of thin sheet rubber; we have found a ladies’ pure rubber bathing cap to be especially serviceable and inexpensive for this purpose. The large volume of air in this tension equalizer and the fact that the entrance and exit are opposite each other would make it possible for a direct passage of carbon-dioxide-laden air directly across the can with a minimum amount of circulation, and hence the after-ventilation to sweep out the carbon dioxide might be incom- plete. Accordingly a semi-cylindrical piece of sheet copper is sol- dered to the can near the entrance coupling in such a way that the air striking this copper is deflected upward against the rubber dia- phragm. This insures a thorough circulatory movement of the air inside the tension equalizer. The ordinary size of ladies’ bath- ing cap permits considerable fluctuation of the volume of respira- tion, and consequently oxygen may be admitted at rather irregular periods. It is only necessary that the operator so adjust the supply of oxygen as to keep the bag from becoming either too much dis- tended or too much flattened. To aid in minimizing the resistance of the movement of this rubber diaphragm, it is so placed that the open end of the tension equalizer is vertical and the diaphragm tends to hang down on the side. Thus the weight of the dia- phragm is in large part taken up by the edge of the metal can. Furthermore, the tension equalizer is always pointed away from the subject, so that he cannot himself note the rise and fall of the diaphragm with each respiration. 358 Francis G. Benedict. Piping, hose, and couplings. —I[n the apparatus as here used, standard %-inch pipe is used. This has an actual internal diameter of 15 mm. The rubber hose has an internal diameter of 19 mm., and the total length necessary for an apparatus of this type is ap- proximately 2 meters. Since all of the work of maintaining the ventilating current of the air falls not upon the lungs of the sub- ject but upon the rotary blower, it is obvious that if the distance between the three-way valve and the tension equalizer is not too great, the remaining piping may -be of almost any length. Thus it is not out of the range of possibilities that the apparatus could be placed in one room of a hospital or clinic and pipes carried to another room. Since, however, the whole apparatus is very easily transported on the table, such a procedure would not be necessary except in rare instances. Rotary blower. — The blower used in this apparatus is of the so- called positive type, that is, there 1s a movable piston on an eccen- tric shaft which forces the air around a circular chamber and out through an opening in the bottom. It has been described in detail elsewhere.1® Pressures up to some 40 to 50 cm. of mercury are readily obtained with this blower, pressures far in excess of that demanded in work of this kind. A large wheel on the shaft of the blower is belted directly to the small electric motor, and by varying the resistances in the circuit the speed of the motor can be adjusted at will. The speed of the motor and the revolutions of the blower are so adjusted that the total amount of air passing through the ventilating pipe is not far from 35 litres per minute. To insure absence of leaks, the blower is immersed in oil; if a leak occurs at any. time, it is made manifest instantly by the bub- ~ bling of air through the oil. Drying apparatus. — The air current brings with it the water absorbed from the air moistener and a certain amount of water vapor from the lungs. Since, in the ordinary experiments on the respiratory exchange, no particular use is made of the determina- tion of water, and as the carbon dioxide is determined by weight rather than by volume, it is necessary to dry the air before it enters the carbon-dioxide absorbers. The air leaving the blower passes through the two ordinary Wolff bottles. For the sake of conven- lence in cleaning, 3-neck Wolff bottles are commonly used. They © ATWATER and BEeNEpIctT: Carnegie Institution of Washington, Publication No. 42, 1905, p. 18. Apparatus for Studying Respiratory Exchange. 359 are filled with acid to a certain level marked on the bottle, and as the water is absorbed and the acid becomes diluted, experience has shown that a certain increase in volume indicates the time for the renewal of acid. The acid in the second Wolff bottle rarely has to be renewed. Two Wolff bottles, fitted with glass tubing 16 mm. internal diameter, permit the passage of 35 litres of air per minute, completely depriving it of moisture without difficulty. Carbon dioxide absorber. — These absorbers are of very much the same type as those described for the large respiration chamber and are made of brass, silver-plated to resist the action of the alkali. They are 26 cm. long, 12 cm. in diameter, and at each end is a hose coupling of standard size, so that there is perfect interchange- ability of all parts. Such a can, filled with soda lime, will absorb about 60 gm. of carbon dioxide without allowing any to pass. Water absorber. — The dry air entering the soda lime can takes up in its passage through the can some of the moisture from the reagent, and since the amount of carbon dioxide absorbed is de- termined by weight, the amount of water removed from the can must be known accurately. In order to know how much water has been removed, the air passes through a Kipp generator con- taining strong sulphuric acid; the drying is accomplished by the bubbling of the gas once through the acid, but the pumice stone in the chamber above. which is also drenched with sulphuric acid, aids materially in the removal of the last traces. It has been found that one of these Kipp generators will dry the air to the same de- gree of moisture content that it has when it leaves the Wolff bottles below. This apparatus is likewise provided with ordinary hose couplings. Air moistener. — The dry air leaving the Kipp generator is moist- ened before being breathed again by passing it through a second Kipp generator with water. To this water 1s added a small amount of sodium bicarbonate to neutralize any possible acid fumes that may leave the Kipp generator. The passage of 35 litres of air through this water results in the saturation of the air to about 65 per cent, a degree of humidity that makes respiration very comfortable. Manometer. — At the beginning of the experiment with the motor not in motion, the whole system is filled with air under a definite pressure. At the end of the experiment stifficient oxygen should be admitted to bring the pressure to what it was at first. see | 360 Francis G. Benedict. The most delicate form of manometer for this purpose that we have found is that used so successfully by Pettersson !7 on his gas FIGURE 3. — General view of respiration apparatus in use. Rotary blower, electric motor, and two Wolff bottles on lower shelves. Air moistener in lower foreground. Car- bon dioxide absorbing apparatus, oxygen cylinder, petroleum manometer, on top shelf. Nosepieces in position on subject. Tension equalizer, and three-way valve supported on cross rod attached to upright standards. Balance for weighing absorb- ers in rear. Specimen nosepiece and bulb for dilating on top of table. analysis apparatus. It consists of a glass tube bent in the arc of | a large circle and containing a short column of petroleum. The movement of this petroleum column along the are of the circle for | a few degrees is a very delicate measure of pressure and leaves | little to be desired for an apparatus to be used as a manometer. . 7 PETTERSSON: Zeitschrift fiir analytische Chemie, 1886, xxv, p. 467. Apparatus for Studying Respiratory Exchange. 361 Supply of oxygen. — Although steel cylinders of compressed OXy- gen weigh more than is desirable, the gas can be obtained readily and of a high degree of purity, and by use of the balances described beyond, it is possible to weigh the cylinders to a centigram. The gas as it leaves the cylinder contains a small amount of carbon dioxide, water vapor, and nitrogen. The apparatus has been de- scribed in detail elsewhere.'S The general appearance of the cylin- der can be seen in Fig. 3. The carbon dioxide is removed by a soda lime tube and the gas dried by sulphuric acid in a special form of glass tube. Both purifiers are attached by rubber bands to the oxygen cylinder and weighed with it. In order to prevent any sudden escape of gas through the tubes, a rubber bag is attached to the valve. This rubber bag should be deflated completely each time the cylinder is weighed, and care should be taken in admitting the gas to deflate the bag before the valve is finally closed. More recently we have obtained small cylinders weighing but 3 kilos and containing 143 litres of 97 per cent oxygen from the Linde Air Products Company of Buffalo. The gas is practically free from carbon dioxide and water, and purifying attachments are unneces- sary. By means of a somewhat expensive reduction valve, the rubber bag may likewise be rejected. The oxygen ordinarily con- tains about 3 to 5 per cent of nitrogen, and a small correction on the weight of the gas is made, as is pointed out later. While the increasing use of cylinders of compressed oxygen in medical practice makes this gas available in most places, it was thought advisable to attempt to substitute some form of portable oxygen generator that could be used with equal success. A great many experiments were made with a generator '? supplying oxygen from sodium peroxide. The oxygen made by the interaction of sodium peroxide and water is remarkably pure, and the simplicity of the operation is such as to commend it for use. The form of generator modified for use in these experiments is shown in Fig. 4. A metal bell, 4, is submerged in a metal vessel containing water. A tin can containing fused sodium peroxide, sold under the trade name of oxone, is held in the bottom of the bell by two springs. Holes are punched through the top and 18 ATWATER and BENEDICT: Loc. cit., p. 32. 19 The generator is furnished by the Roessler & Hasslacher Chemical Co. of New York. 362 Francis G. Benedict. bottom of the can to allow the addition of water. On open- ing the valve C, which connects with a pipe screwed into the top of the bell, water rises inside the bell, comes in contact with the sodium peroxide, and generates oxygen. If the valve is closed, the oxygen generated forces the water down until it is beyond the reach of the sodium peroxide, and the gen- = G AA eration of gas ceases. The gas thus formed is remarkably pure, containing only moisture. The whole apparatus can be weighed on the balances used for weighing the soda lime ab- sorbers, and hence the amount of oxygen generated can best be noted by differences in weight. In order to dry the issuing gas, a U-tube filled with pumice stone drenched with sul- phuric acid is attached to the side of the FIGURE 4.—Generatorsup- cylinder by rubber bands. That the major plying oxygen from so- : : - ciumiecrriden Wi anne pOneion of the drying may be accomplished gas chamber; B, cartridge and at the same time the rate of flow of of sodium peroxide; C, gas indicated, the glass tube in one arm of needle valve; D, drying the U-tube is caused to dip into concen- tube with pumice stone . ere ; aide cuinnincescd. ue trated sulphuric acid in a short piece of test tube with sulphuric acid tube. The gas bubbles through the acid, for drying; G, valve com- passes through the pumice stone, and escapes municating to interior of through a T-tube at the top. Occasionally, generator; F, water level : ee: he when the rate of flow is very rapid, the oxygen will be generated so fast as to force the water away from B so rapidly that some gas may even escape under the edge of the bell, consequently the valve G in the top of the apparatus is opened and connected directly with the T-tube. The top of the apparatus is, as a matter of fact, se- curely fastened by two thumb screws to a rubber gasket indicated in heavy black on the diagram. To assist in regulating the flow of gas, the level of the water in the metal cylinder is made visible by means of a small, arbitrarily graduated water gauge, F, attached to the outside of the cylinder. By means of this gauge one can tell exactly the level of the water at any time, and after a few preliminary trials the point to which water may be allowed to rise without danger of loss of oxygen is readily found. GES ALSEA lil! A, = 1 Apparatus for Studying Respiratory Exchange. 363 In practice the U-tube is attached by rubber bands to the upper part of the cylinder. One objection to this apparatus is the fact that during the action there is intense heat and the cylinder and liquid become very much heated. This interferes considerably with accurate weighing. It has been found practical to place the generator in a pan of cold _ water, allowing water to rise until it is just below the U-tube. Under these conditions the excessive heat is easily controlled, and by wiping off the cylinder carefully, the weight is accurately and quickly obtained. In the generation of the hot gas steam is apt to condense in the pipe at the top of the bell. This will occasion- ally collect in the fine needle valve, C, and make the flow of gas intermittent, and difficulties have been experienced in the use of the apparatus at just this point. Wherever cylinders of compressed oxygen of suitable size and weight cannot be obtained, the above modified form of generator can be used with success. The cylinders of compressed gas are invariably to be recommended. Balances. — The oxygen cylinders with the purifying attachments weigh some 8 to 10 kilos. Fortunately balances can easily be ob- tained in the market capable of weighing these cylinders to within I centigram. The balance has been described elsewhere.?° They are furnished by most of the large supply houses. The balance here in use (shown in the background in Fig. 3) is of the size sold as having a carrying capacity of 10 kilograms in each pan. Usually, in placing the oxygen cylinders on the balance, it is neces- sary to have some form of hook to hold the cylinder in place. Since, however, differences rather than absolute weights are of value, this hook system can be left on the balance permanently. APPLIANCES FOR BREATHING. The successful use of this apparatus necessitates that the air should be breathed into and out of the ventilating air pipe with- out any escape of air into the room or without any entrance of room air to the lungs or the apparatus. This implies, then, that there should be a very close seal between the mouth or nose and 20 ATWATER and BeNEDIcT: Carnegie Institution of Washington, Publication No. 42, 1905, p. 57; BENepict and Miner: U. S. Dept. of Agriculture, Office of Ex- periment Stations, Bulletin 175, 1907, p. 20. 364 Francis G. Benedict. the ventilating air pipe. A very large number of experiments has been made in connection with this apparatus in which the Zuntz mouthpiece, the Tissot glass nosepieces, and several forms of masks have been used, and, as a result of all this experimenting, a form of nosepiece has been developed which gives by far the best - results. Nosepiece. — The difficulties experienced with the Tissot nose- piece have been serious, namely, the rigid glass tubes entering the nose must be forced into the nose with such great pressure that their use becomes very painful. Instead of relying upon the flexi- bility of the nose to adapt itself to the rigid glass surface, the attempt is made here to insert a nosepiece with an inflatable rubber cover in such a manner that the rubber could ee i ene fit into the inequalities in the nasal orifice and connecting with respira- thus produce tight closure. This result was tion apparatus; B, rubber obtained in a very simple way by preparing a finger cot; C, rubberstop- nosepiece as shown in Fig. 5. A glass tube, per; D, small rubber : : tube for idilatine mubber, 4 internal diameter, serves to conduct the cat air into and out of the nose. A thin rubber finger cot, B, has a hole cut in the end, and the end is then slipped over the glass tube, A. This is well tied to the glass with white silk and shellacked. The rubber finger cot is then turned inside out and turned back over the glass tube, A, and tied about a rubber stopper, C. This rubber stopper has a small glass tube passing through it leading to the space between the rubber finger cot and the large glass tube. A rubber tube D, pro- vided with a pinchcock, permits the introduction of air in the an- nular space between the glass and the rubber, thus allowing B to be inflated. After the inflation has proceeded to the proper point, the pinchcock is closed and the apparatus holds indefinitely. In use, the deflated apparatus is first inserted in the nose in the proper posi- tion; then, by using a syringe bulb, air is forced into the annular space until the desired tension is secured. As the rubber expands, it fills in perfectly the inequalities in the surface between the glass tube, A, and the nose. The second nasal tube is then inserted and inflated, and finally, if desired, the edge of the nose can be smeared with soapsuds and the subject can put pressure upon the tubes and test for any leak. When properly inserted, the tubes Figure 5.— Nose tube for Apparatus for Studying Respiratory Exchange. 365 cannot leak, and they may be worn an hour or more with perfect comfort. It has frequently happened with one of our subjects that he has slept through three or four experiments entirely unconscious of the presence of these tubes, thus showing that they cannot be very annoying or uncomfortable. The glass tube, 4, is in turn connected by a short piece of rubber tubing with a metal tube attached to the three-way valve on the ventilating air pipe. In only those cases where nose breathing is impossible or difficult is the mouth appa- ratus recommended. Occasionally it has been found advantageous to use the Zuntz mouthpiece and the noseclip to insure closure of the nose. Under these conditions it is necessary to secure tight closure around the mouth. The subject must not be allowed to fall asleep, as other- wise the muscles around the mouth will relax and a leak occur. The nose must be thoroughly tested for tightness. When the nose tubes are used, it is necessary that the mouth be kept absolutely closed. It frequently happens that the noise of the blower produces a sense of drowsiness in the subject and he falls asleep. The lips may open slightly and leakage occur. It is occasionally necessary to insure that the mouth is kept closed by attaching a small piece of surgeon’s plaster to the upper and lower lip, first having the subject draw the lips tightly together. With these precautions the subject can go to sleep and one can be sure that both mouth and nosepieces are perfectly tight. METHOD OF USE. The subject, covered, if necessary, with a blanket, lies with the head in a comfortable position on a couch, and he should remain in this position for some minutes before the experiment begins. Too great stress cannot be laid upon the fact that the subject should be resting quietly, with the respirations normal and the pulse normal. After a slight physical exertion or any psychic excitement abnormal results are almost invariably found in the first of a series of ex- periments, due usually to the increased carbon dioxide exhaled unaccompanied by a corresponding intake of oxygen. During this preliminary period the nosepieces are properly adjusted and inflated. The pulse can be taken as usual, but in most of the experiments it has been found of advantage to use a Bowles stethoscope with long air transmission through a rubber tube, so that the pulse can 360 Francis G. Benedict. be counted without the knowledge of the subject. While in most of these experiments a pneumograph placed about the trunk mid- way between the nipples and the umbilicus has been used, the respiration can be counted perfectly satisfactorily from the rise and fall of the chest wall. Prior to the experiment proper, the subject lies quietly on the couch and breathes through the nose or mouth pieces into the three-way valve, one opening of which connects with the ventilat- ing air pipe, the other open to the air. During the preliminary period he breathes through the side outlet and consequently is respiring ordinary air. Meanwhile the motor is started, and the air in the whole system is thoroughly mixed. The motor is then stopped and oxygen admitted from the cylinder until the rubber diaphragm is distended to such an extent that there is a slight positive, definitely known pressure on the system, as indicated by the petroleum manometer. At that point the oxygen supply is shut off, the manometer shut off, and the motor again started. The subject is carefully watched, and at the end of a normal expira- tion the three-way valve is suddenly turned by the operator, and the next inspiration consists of air removed from the ventilating air pipe. In some of the earlier experiments attempts were made to have the subject personally throw the valve at the end of a normal ex- piration, but it was soon found that this could not be carried out satisfactorily, and hence the valve was so arranged as to be turned by the observer unknown to the subject. On the top of the valve is attached a socket wrench connected by brass rods with two uni- versal joints to a hand wheel attached to the table (see Fig. 3). By turning this hand wheel the valve is thrown. It has been found that even subjects used to the apparatus are inclined to anticipate the moment of throwing the valve, and occasionally the respira- tion becomes abnormal. In order to diminish the resistance as much as possible, the ten- sion equalizer is placed as near the three-way valve as is convenient, 1.¢e., about Io cm. from it. Under these conditions there is no noticeable variation in tension, whether the three-way valve is open to the air or open to the ventilating pipe. The rate of ventilation is so adjusted that at no time is there any danger-of the subject’s breathing into the lungs air that has just been expired. For this purpose a number of experiments have Apparatus for Studying Respiratory Exchange. 367 shown that a rate of 35 litres per minute is amply sufficient to take care of this point. In the first place, at the end of each expiration there is an instant’s pause, during which time the last expired air is rapidly pushed along the tube, its place being taken by fresh, pure air. When the inspiration begins, this fresh air is carried to the lungs, and unless the rate of inhalation is more rapid than the flow of 35 litres per minute, obviously none of the vitiated air can be drawn back from the tension equalizer into the lungs. During an experiment, as the oxygen is consumed out of the air and the rubber bag or the tension equalizer sinks more and more into the can, oxygen is supplied from the steel cylinder which has previously been weighed. This cylinder is not the one that has been used for filling the system to a constant tension before the experiment begins, as the preliminary introduction of oxygen need not be quantitatively known. Toward the end of the experiment (ten to twenty minutes), it is necessary to see that the rubber bag is not too much distended to permit the last expiration to take place without producing any tension on the rubber. The patient is then carefully watched, and at the end of a normal exhalation the valve again thrown, this time so that the opening to the air pipe is closed and the subject is breathing through the side outlet into the open air. There is, then, in the system between the valve piece and the carbon dioxide absorber air which contains a large percentage of carbon dioxide. The current of ventilating air is maintained for some three or four minutes, during which time the system is thor- oughly swept out, and at the end of which time there is no appre- ciable amount of carbon dioxide remaining. At the end of three minutes the motor is stopped and oxygen again admitted, this time _ from the weighed cylinder used during the experiment, until the petroleum manometer indicates the same tension on the system that was there at the beginning of the experiment. In this apparatus, since for ordinary experiments the water ex- pired from the lungs is of no particular value, it is unnecessary to weigh the two Wolff bottles, and only the soda-lime can and the Kipp generator containing sulphuric acid are weighed. The loss in weight of the oxygen cylinder is also carefully recorded. In an experiment of but ten minutes’ duration the carbon dioxide excretion is rarely below 3 gm; consequently, if the carbon dioxide absorbing vessel is weighed to within 0.03 gm., the error is but I per cent. Similarly the amount of oxygen absorbed is rarely 308 Francis G. Benedict. less than 3 gm., and an error of weighing of 0.03 gm. again in-- volves an error of but 1 per cent. As a matter of fact, with bal- ances which can be easily obtained and at small cost, it is possible to weigh these absorbing vessels rapidly to 0.01 gm., and hence the errors in weighing may practically be neglected. Calculation of results. — While the amount of carbon dioxide ab- sorbed by the soda lime is usually greater than the amount of water given up to the dry air as it passes through this can, it may happen that an actual loss in the weight of the soda lime can oceurs sim- ultaneously with a large gain in the acid vessel. Obviously the algebraic sum of these variations in weight represents directly the amount of carbon dioxide exhaled during the experiment. The volume is found by multiplying the weight in grams by the factor 0.509. If it were possible to admit absolutely pure oxygen, the calcula- tions for oxygen would be equally simple. Where the sodium per- oxide generator is used, this corresponds to the case, although marked variations in the level of the water inside of the generator should be taken into consideration. With cylinders of compressed oxygen, however, the oxygen is by no means pure. It may at times contain but go per cent of oxygen, and hence apparently a large correction should be made. A close examination of the figures, however, shows that the cor- rection is by no means as great as one would at first sight think. What is measured in this apparatus is the amount of oxygen re- quired to replace an equal volume absorbed by the subject. The amount of oxygen admitted does not represent pure oxygen, but oxygen plus some nitrogen. For each litre of nitrogen admitted there would be a loss in weight of the cylinder amounting to but 1.26 gm. instead of 1.43 had the gas been pure oxygen. As a matter of fact, it has been found by testing that the correction to be applied with a cylinder containing, for example, 97 per cent of oxygen is about 0.4 per cent. Thus the loss in weight of the cylin- der should be increased by 0.4 per cent to give the true loss in weight had the cylinder contained pure oxygen. In many experi- ments this slight correction may be neglected. To find the volume of oxygen absorbed by the man during an experiment, the weight is multiplied by the factor 0.7. The res- piratory quotient is the volume of carbon dioxide divided by the volume of oxygen. 6 Se ae EP Se Se eee ——- Apparatus for Studying Respiratory Exchange. 369 CRITICISM OF NEw METHOD. In manipulating the whole system in the new method, it is as- sumed that the temperature of the system is absolutely the same at the beginning and the end of each experiment. Several possible sources of error may creep in here. As carbon dioxide is absorbed by the soda lime, and as water is absorbed by the sulphuric acid, there is considerable liberation of heat due to the chemical reac- tion. This results in the warming of the air in the spaces between the soda lime or the air above the acid, and consequently tends to increase somewhat the total volume inside of the system. This would decrease the oxygen introduced. The experimenter in mov- ing about the apparatus, the subject in breathing into the tubes, would, especially during the first experiment of a series, tend to warm the system slightly. On the other hand, as the dry air leav- ing the Kipp generator passes through the moistening chamber, there is an evaporation of water, cooling the water in the generator and consequently the air above it. Thus the rise in temperature developed by the heat of reaction of the carbon dioxide and soda- lime and the water in the sulphuric acid is partly compensated by the contraction of the air due to the cooling of the air above the water. This is particularly the case in the first experiment of the series, where the water in the moistener may be at room tempera- ture and is subsequently considerably lower. Another possible source of error is the assumption that there is a constant degree of humidity in the air after it leaves the mois- tener. At the beginning of the experiment, before the valve is thrown, air is circulated through the system, and all the air leaving the moistener has a water content which by experiment has been found to be not far from 65 per cent. The amount of water evap- _orated into the air current as it passes through the Kipp generator will depend upon two factors, — first, the speed of the ventilating circuit, and, second, the temperature of the water in the generator. Consequently, if there are fluctuations in either of these, there may be variations in the amount of moisture in the air current. As a matter of fact, the fluctuations in the rapidity of ventilation are for the most part, rather small, and, in the second place, after the first experiment of the series, the temperature of the water in the moistener remains relatively constant. These errors are in large 370 ; Francis G. Benedict. part avoided by running the apparatus for several minutes before an actual experiment begins. Another source of error lies in the fact that it is assumed that the barometric conditions remain constant throughout the experi- ment. As most of the experiments do not last over twenty minutes, it is to be doubted, however, whether there will be any material variation in barometric pressure during this time. The total volume of air, including that in the distended tension equalizer, is not far from 8000 c.c. A change in the barometer of I mm. would be equivalent to 1/760 X 8000, or about Io c.c., corresponding .to about 0.014 gm. of oxygen. This is, therefore, the possible maxi- mum error on this apparatus. Obviously the error would be plus or minus, depending upon whether the barometer rose or fell. For the strictest accuracy, therefore, one should note the barometer at the beginning and end of each period. Influence of variations in the residual air. — This method involves the assumption that the same quantity of air remains in the lungs at the end of each experiment that was present at the beginning. A large number of tests have seemed to indicate that with normal subjects lying quietly on a sofa with quiet respiration, this is the case. Obviously with this method of studying the respiratory exchange, any errors involved in the change of residual air would affect no- ticeably the oxygen determination, and it is here that one finds the weakest point of the whole system. While a few preliminary ex- periments have been made on the use of the forced expiration as the moment to throw the valves, as a rule it has been found that the normal quiet respiration with the subjects lying on a bed can be so readily judged by the assistant that the throwing of the valves is quite a simple matter, and the inequalities in the volume of the residual air in the lungs are apparently so small as not to affect the determination of the respiratory quotient. Influence of a leak on total metabolism. — In any of the methods now in use, including that here suggested, if there is a leak around the mouth or nose, the results are affected. Thus, in Zuntz’s ap- paratus, while a leak would not influence perceptibly the composi- tion of the expired air and consequently the respiratory quotient, obviously the total volume of expired air might be considerably affected. With the apparatus here described, the slightest leak influences enormously the determination of oxygen. On the other Apparatus for Studying Respiratory Exchange. 371 hand, a small leak is without appreciable effect in the determina- tion of the carbon dioxide production. DISADVANTAGES OF THE NEw APPARATUS. The type of apparatus described here has certain disadvantages that must obviously be taken into consideration. In the first place, it is impossible to make duplicate analyses. With the Zuntz method and the spirometer method, analyses can be made almost ad libitum. The new method, therefore, like the method of Hanriot and Richet,?? must be classed with those that do not permit duplicate analyses. On the contrary, the rapidity with which these deter- minations can be made leaves very little necessity for duplicate analyses. The duplicate analyses have value only for insuring the accuracy of the composition of the air, and there is no method of securing duplicate determinations of the total volume of air ex- pired in any given experiments. Hence this method fundamentally has no greater disadvantage in studying the total metabolism than has any other method involving a measure of the total products of respiration. Second, the method likewise does not permit any measurement of the total ventilation of the lungs per minute or of volume of each expiration. As yet these values have not been of great physi- ological significance, especially when determined in connection with respiratory exchange. They are of value in the Zuntz method, and would be of value in this method in indicating the normal respiration at the beginning of a series of experiments where the subject had not previously been used. It is seriously to be ques- tioned, however, whether similar data of equal value cannot be obtained by means of the pneumograph and tambour. This method has been used in this laboratory with considerable success to in- dicate normal respirations at the beginning of the experimenting. Third, it is obvious that_this apparatus cannot be adapted for a portable type of apparatus, thus making it impossible for study- ing many problems which have been investigated by means of the Zuntz apparatus. 21 Hanriot and RicHET: Comptes rendus, i881, civ, p. 435. 372 Francis G. Benedict. CoNTROL EXPERIMENTS WITH BURNING ETHER. Experience with the large respiration chambers has shown that it is absolutely necessary in studying the respiratory exchange to control the apparatus from time to time by some delicate chemical tests proving the correctness of the deter- minations of carbon dioxide and oxygen. It seemed desirable, in connection with the de- velopment of this apparatus, likewise to have a method of checking it exactly. In order to test the apparatus it is only necessary to insert’ a combustion chamber of special construction in the ventilating air pipe at about the point where the three-way valve is ordinarily at- tached. After many experiments with various forms of combustion chambers and various types of alcohol burners, it was found im- possible to secure satisfactory results by burn- ing alcohol. The final type of combustion FIGURE 6.—Combustion chamber for control experi- ments with burning ether. A, combustion chamber; B, ingoing ventilating air current; C, outgoing air current; D, burner; &, glass-window; F, F’, high tension sparking current lead wires; G, container for ether; H, supply of air under pressure; J, water cooler. chamber was devised to permit the burning of ether vapor. The apparatus is shown in Fig. 6 herewith. It consists of a large metal tee, A (stand- ard 2-inch). Into this is screwed an upright piece of pipe which is surrounded by a tin water jacket, J. On the top an elbow is at- tached, into which a pipe, C, is screwed. In the bottom of the tee, A, is screwed a short piece of pipe having a rubber stopper in it. through which are passed, first, a rubber tube, B, through which the ventilating current of air passes, then a small brass pipe to which is attached a burner, and finally two electric wires, F and F’. A glass plate, E, permits a careful inspection of the flame. Ether is supplied from a glass vessel, G, which is, as a matter of fact, an ordinary so-called sul- phur dioxide condensing tube. A current of air, entering the ether tube at H, passes over the ether, becomes saturated with ether vapor, enters the combustion chamber, and issues from the jet on the burner, D. By passing a high tension current through the wires Apparatus for Studying Respiratory Exchange. 373 F and F’, a spark is caused to jump across the gap just above the burner D, thus igniting the ether vapor. The excessive heat developed by the burning of the ether is absorbed readily by the water in /, and the gas issues at C, practically at room temperature. In order to maintain a constant flame, a steady pressure of air must be obtained, and this is secured by inserting a T-tube between the blower and the first sulphuric acid Wolff bottle. A small supply of air taken from this point suffices to carry the ether vapor into the combustion chamber. In the test the vessel G is weighed be- fore and after the experiment, and the amount of ether vaporized accurately known. At the end of the experiment the supply of ether vapor is shut off and the ventilating air current allowed to run for several minutes to sweep out the carbon dioxide already formed and to allow the whole system to acquire room tempera- ture. At the end of that time the oxygen supply is added until the manometer indicates the same tension as at the start. The ap- paratus is then disconnected, and the soda lime vessel, the sulphuric acid water absorber, and the oxygen cylinder weighed. From these data it is possible to compute, first, the amount of carbon dioxide produced in the combustion of a given amount of ether; second, the amount of oxygen required to oxidize a given weight of ether; and, third, the ratio of carbon dioxide formed to oxygen absorbed in burning the ether. In certain experiments it was somewhat difficult to regulate the flame so as to secure perfect combustion, and occasionally unburned ether passed through the system. Under such circumstances, therefore, the amount of carbon dioxide actu- ally absorbed was not so great as would be expected from the loss in weight of the ether vessel, but since the ether vapor thus un- burned was immediately absorbed by the sulphuric acid in the ab- sorbing vessels, it did not influence in any way the ratio of the carbon dioxide formed, and the oxygen absorbed of such ether as was burned, and consequently the respiratory quotient, so to speak, of ether was almost always found to correspond with the theoretical, irrespective of the absolute amount of ether burned. It was pos- sible in most experiments to adjust the flame of the burner so as to have perfect combustion. A typical test is given herewith. The experiment lasted fifteen minutes. 374 Francis G. Benedict. ETHER EVAPORATED. Found. Required. Carbon’ dioxides) 7) = + 11.62 gm. 11.71 gm. Onycen Eo est. pn a eae 12.78 gm. 12.78 gm. Respiratory quotient an . 0.662 0.666 Physiological contfol. — While there may be reasonable question- ing as to the true value of the respiratory quotient determined under conditions necessary with this or any other form of apparatus involving artificial breathing, it was possible to control the experi- ments with this particular form of apparatus by determining the respiratory quotient in a respiration chamber in this laboratory. Consequently it was possible to compare directly the respiratory quotient as determined by the new apparatus and that determined in a respiratory chamber where the subject was lying quietly and breathing normally. Several such comparisons have been made, and but one need be here given. Thus, three consecutive ten-minute experiments on the new ap- paratus two hours ‘after breakfast indicated a respiratory quotient of 0.87, and immediately following this, two one-hour periods in the large chamber gave values 0.89 and 0.90. Several hours later the values found with the new apparatus were 0.78, 0.82, and 0.79, and immediately following this, two consecutive one-hour periods in the respiration chamber, 0.77 and 0.76. While variations were found in the total metabolism during ‘these periods, since in one case the subject was breathing through nosepieces and was in the laboratory, surrounded by a number of observers, and in the other case he was lying quietly in the chamber, the results show that there is nothing abnormal in the measurements as obtained with this type of apparatus. In the development of this apparatus I have been much indebted to Mr. F. P. Fletcher, who was especially concerned with the ap- paratus for control with burning ether; Mr. J. A. Riche, who has conducted the greater number of experiments with men, and to Mr. T. M. Carpenter, who has personally supervised many of the details of construction and testing. CAN FUNCTIONAL UNION BE RE-ESTABLISHED BE- TWEEN THE MAMMALIAN AURICLES -AND VEN- FRICLES AFTER DESTRUCTION OF A SEGMENT OF THE AURICULO-VENTRICULAR BUNDLE? By JOSEPH ERLANGER [From the Physiological Laboratory of the University of Wisconsin.] (WitTH a Histotocicat Stupy sy W. S. MILLER.) HE sole functional connection between the auricles and ven- tricles of the mammalian heart, the auriculo-ventricular bundle, is composed of a tissue which in many ways resembles, although it is not identical with, heart tissue in general. When the continuity of this structure is completely interrupted in any way whatever, complete auriculo-ventricular heart block is the re- sult, and this is permanent.‘ In other words, regeneration of the severed auriculo-ventricular bundle to the extent that it may become capable of again functioning does not occur. May we then be justi- fied in concluding from this result that heart tissue in general does not regenerate? * This question is of the greatest practical signifi- 1 Some new evidence bearing upon these subjects, as well as references to the literature, will be found in a paper by the author soon to appear in the Journal of Experimental Medicine. So far as the author is aware, the question of regeneration of heart muscle has been studied by histological methods only. The prevailing opinion seems to be (THoREL, in Lubarsch and Ostertag’s Ergebnisse, 1903, ix, p. 861), that wounds of the heart muscle heal by the formation of scar tissue. The muscle cells show little if any tendency to regenerate. No effort seemingly has been made to determine if there is any restoration of conductivity across a healed wound. BERNSTEIN’s experiment (functional isolation of the tip of the frog’s ventricle by means of a crush) throws little or no light upon this question because the period of survival probably does not suffice for the completion of regenerative processes. According to TIGERSTEDT ‘(Physiologie des Kreislaufes, 1893, p. 157), the longest survival is recorded by AUBERT, who has succeeded in keeping a frog alive for six weeks after performing BERNSTEIN’S operation. Another serious objection to the use of BERNSTEIN’S experiment for the elucidation of this question is mentioned on p. 378. 375 370 Joseph Erlanger. cance, since, if it can be shown that functional union of heart tissue never occurs, then must we admit the impossibility of re-establishing by operative interference, in cases of heart block due to destruction of the auriculo-ventricular bundle, a connection between the auricles and ventricles which would restore the normal sequence of heart beat. The experiments which demonstrate that the auriculo-ventricular bundle, or a segment thereof, once destroyed does not regenerate, cannot be considered as conclusively proving the impossibility of establishing experimentally a functional union between the auri- cles and ventricles at some other place. The auriculo-ventricular bundle, it should be borne in mind in this connection, is a long and slender structure encased in connective tissue which in some places is particularly dense. Might not these relations of the auriculo- ventricular bundle interfere in some way with regenerative proc- esses which otherwise might possibly proceed to complete restora- tion of function? Might it not be possible, it may be further argued, to unite auricles to ventricles by an operative procedure in such a way that regenerative processes can occur? THE UNION oF AURICLES TO VENTRICLES. Stimulated by this line of thought, the making of a functional connection was attempted on several of the animals in which at the same time the effort was made to produce auriculo-ventricular heart block by damaging the auriculo-ventricular bundle. Auriculo-ven- tricular heart block was produced at the same time because it was thought that functional union of the coapted surfaces of the auricles and ventricles might be facilitated by the absence of any functioning connection between these chambers. Methods. — The method of procedure was about as follows: After having exposed the heart and after having apparently successfully damaged the auriculo-ventricular bundle, a part of the contiguous surfaces of the right auricle and right ventricle were denuded of epicardium and fat, and, after all hemorrhage had ceased, the de- nuded areas were carefully approximated with fine silk sutures. The operation is a difficult one, particularly the denudation of the auricle. This structure is so thin that there is great danger of opening it while stripping it of its epicardium. So difficult was the Can Functional Union be Re-established? 377 operation that it was rarely successful, and in the most favorable of our cases could only a few square millimetres of the auricle be thus laid bare. Dogs alone were used. They were anesthetized with morphine and ether. Results. — The experiment was attempted four times. Two of the animals, Nos. 1 and 2,° died a few hours after the operation. Another, No. 4, survived the operation, but recovered of the block in the course of a day or two. Dog No. 3, however, survived the operation with a relatively complete heart block, from which it recovered completely in the course of twenty-six days. Evidently recovery in this case could have been due either to the successful operative union of the auricle to the ventricle or to a restoration of conductivity in the auriculo-ventricular bundle. For the purpose of deciding this question the heart was exposed on the sixty-first day, and, while recording the movements of the auricles and ventricles, the sutured area was crushed in a mass ligature. The heart beat remained normal. Complete heart block was then re-established by crushing the auriculo-ventricular bundle. The recovery was therefore due to a complete restoration of conductivity in the auriculo-ventricular bundle, which, as is discussed in another place, had escaped destruction at the time of the first operation. THE REGENERATIVE CAPABILITIES OF AURICULAR TISSUE. Owing to the difficulties in the way of a successful experiment of the kind above described, we thought it might be well to test the question of functional regeneration of cardiac tissue under even more favorable conditions. This was done as follows: Methods. — The auricular appendage was drawn through a clamp so made that with it the appendage could be crushed along a line parallel to, and near its base and extending from one edge of the appendage, either anterior or posterior, to points more than half- way across the ventral and dorsal surfaces. Immediately thereafter ligatures were laid to mark the apex of the appendage and the line of crush. For the latter purpose three ligatures were tied, namely, one at the edge of the auricle crushed by the clamp, and one each on the ventral and dorsal surfaces to mark the other ends of the line of crush. The wounds were then closed and the animals allowed 3 See paper soon to appear in the Journal of Experimental Medicine. 378 Joseph Erlanger. to live. The general technique of the operation was similar to that employed in producing chronic auriculo-ventricular heart block. By crushing only part way across the base of the appendage the appendicular muscle was left in a position to respond to cardiac impulses. In this way any danger of atrophy from disuse was ob- viated, a danger which otherwise would have defeated the object of the experiment. The failure to take this precaution renders valueless in this connection Bernstein’s experiment, which otherwise might be considered as conclusively proving the impossibility of functional regeneration in the case of the frog’s ventricle. Results. — Two kittens thus operated upon died within a few hours. Three dogs experimented upon in the same way survived the operation. One of them, however, died some months later of causes not related to the subject in hand. Two of the animals, Nos. 3 and 5, lived for the final test, which was made on the one hun- dred and thirty-second and two hundred and sixty-eighth days, respectively. At the final test the heart was exposed and, after having located the ligatures in the auricle, arrangements were made for stimulat- ing the apex by the unipolar method and for recording the contrac- tions of the body of the auricle by the method of air transmission. It was found that stimuli applied at any point on the surface of the auricular appendage resulted in irregularities in the beat of the body of the auricle. Then the auricular tissue was crushed in a clamp along a line perpendicular to, and intersecting the base of, the appendage at about the middle points of the ventral and dorsal surfaces. This line met the edge of the auricles a little to one side or the other of the apex of the appendage and presumably inter- sected the old line of crush on the surfaces of the appendage, the ligatures laid at the time of the first operation to mark the latter line serving as guides in this connection. By this operation the auricular appendage was divided for our purposes into two areas, namely, one presumably completely isolated from the rest of the heart by the old and new lines of crush, which area, for the sake of con- venience, we shall designate area Y; the other, termed area X, lying to the opposite side of the new line of crush and presumably still connected functionally with the rest of the heart at the base of the auricular appendage where the tissue had not been crushed at the first operation. Then each area in turn was stimulated tetani- cally, the stimulating electrode being applied to one or the other at different points but as far as possible from the old line of crush. _ f Can Functional Union be Re-established? 379 The results of the tests were as follows: In the case of Experi- ment III the auricle became irregular wherever in either area the stimulus was applied. Therefore neither area was completely iso- lated. For the purpose of demonstrating that the new line of crush was thorough and that the impulse was probably passing through tige line as marked by the ligatures, the auricular tissue was crushed along a line parallel and immediately distal to the latter. Now the stimulus applied to area Y no longer elicited irregularities of the parts of the heart lying without it. In the case of Experiment V a stimulus of considerably more than minimal density when applied to area X invariably caused the auricle to become irregular, whereas no irregularities whatever were recorded when the stimulus was applied to area Y. Area Y, evi- dently, was completely isolated. Discussion. — The results of these two experiments are therefore apparently contradictory. It would seem at first sight that in Ex- periment III functional regeneration had occurred over the old line of crush, whereas in Experiment V functional union had failed to take place. A decision as to the results of which experiment are to be considered as conclusive can be made only by taking into con- sideration the circumstances attendant upon each of the experiments. It should first be made clear that the difficulties in the way of a successful operation were very great. In the first place, we could not be sure as to the exact limits of the original line of crush. Although the marking ligatures were placed as carefully as possible at the ends of the line of crush where seen immediately after re- moving the clamp, it is possible that the auricular tissue may not have been thoroughly crushed quite up to that point; indeed, that some of the tissue included in the grasp of the clamp may have escaped thorough disintegration. Any one who has attempted to destroy the continuity of the auriculo-ventricular bundle by com- pression is in a position to appreciate the difficulties in the way of thoroughly and permanently destroying the functional continuity of two regions of the heart. We therefore cannot be certain that area Y was completely bounded by tissue that had been thoroughly destroyed. An experiment which shows that there is no functional * The histological study by Dr. W. S. Mixxer of the material obtained from Ex- periment V is appended to this paper. It is to be regretted that the material from Experiment III could not be utilized for this study. The series of sections obtained from Experiment V, however, leaves nothing to be desired. 380 Joseph Erlanger. union over a line of crush is consequently of far greater significance than one that shows the existence of functional union. In the latter case functional union may never have been interrupted, although the possibility must be admitted that it may have been formed anew; whereas in the case of the former there is no such alternative. When, in connection with these considerations, it is recalled that Experiment V was the last of our series and was performed under the very best of conditions, and that at the time we were perfectly certain of our procedures, whereas this was not so in the case of Experiment III, which was the first made upon the dog as subject, we have not the slightest hesitancy in accepting the results of Ex- periment V as conclusive. Furthermore the histological study of Experiment V substantiates this conclusion in a most satisfactory manner. It shows that all muscular connection of area Y with the surrounding parts of the heart had been completely severed, whereas area X was still in con- nection with the body of the auricle through a band of uninjured muscular tissue which in the natural state was probably more than 0.7 mm. wide. It is of the greatest interest that, although there was no muscular connection between area Y and the body of the auricle, these parts were none the less connected by several nerve trunks which either had escaped destruction at the time of the first operation or had grown across the old scar. Despite this nervous connection, impulses started in area Y were not conducted to the parts of the heart below. Evidently the cardiac excitation wave passes through myocardium and not through nerve trunks. The myocardium may for our pur- poses be considered as consisting of muscle and of the nerve plexus intimately associated therewith. There consequently remained open to the cardiac excitation wave, as in the normal heart, two possible paths of exit from area X. Assuming that the excitation wave normally takes but one of these paths, it evidently still remains to ~ be determined which is the one. Our experiments throw but little light upon the question. The heart tissue was so prepared for histological examination that the course of the intrinsic nerve plexus could not be traced. If we may however assume that the nerve plexus, just as nerve trunks, can grow across a scar, then must area Y have been in connection with the heart through the plexus. Ad- mitting this, we are led to the conclusion that the cardiac excitation wave is conducted through muscle fibres only. Can Functional Union be Re-established? 381 We conclude therefore: 1. That functional union cannot be re-established between two parts of the heart whose functional continuity has been completely severed by thorough destruction of the connecting heart tissue. There is not the slightest hope of relieving heart block thus caused either by surgical interference or by the operation of natural re- generative processes.° 2. That the cardiac excitation wave is not conducted through the auricle in nerve trunks located in its tissue. 3. That the cardiac impulse is conducted through muscle fibres rather than through nerve fibres. Since heart tissue, as we under- stand it, is composed of muscle and nerve plexus, and since nerve fibres ordinarily regenerate and thus re-establish their former con- nections, these experiments would seem to support the myogenic theory of the heart beat. This conclusion is stated with some re- serve, however, since our knowledge of regenerative processes in nerve plexuses is at present in a very unsatisfactory state. HISTOLOGICAL STUDY. By W. S. MILLER. Method. — Fixed in Tellyesnicky’s fluid; imbedded in celloidin; cut serially by a modification of Obregia’s method in 478 sections p 40 thick; stained with haematoxylin and eosin. The entire auricle was not sectioned; only a small portion, however, was left out of the series. That this was a negligible quantity may be seen by reference to Fig. 1, in which the line of the first crush, that of the second crush, and the plane of section are indicated. First crush. — The first evidence of destruction of heart muscle is seen in section 102 and is situated at the outer rounded edge of the section. The destruction of the muscle is complete (Fig. 2). In section 124 the line of crush has moved around to the ventral 5 We do not by this mean to imply that a bundle made functionally insufficient by some process that does not completely destroy its anatomical continuity, such, for example, as compression, infiltration, or recoverable degenerations, cannot again resume its function with the return of conditions to the normal. Indeed there are already on record instances of recovery from complete auriculo-ventricular heart - block due probably to the relief from compression exerted upon the bundle from without. 382 Joseph Erlanger. side of the auricle while the opposite side is intact (Fig. 3).. In section 138 a double line of crush is first noticed; that is, scar tissue A / ace: . ges FicurE 1.— Outline of portion of auricle sectioned. A, line of first crush; B, line of second crush; S, plane of section; the sections were cut as far as the crest of the convex base line. The break in line A indicates approximately the band of intact heart muscle. X, Y, areas mentioned in text. is found on both sides of the section with intact muscle between (Fig. 4). From this point the two lines of scar tissue move FicurE 2.—JIn this and the following fig- FIGURE 3. ures the crushed heart muscle is indicated in solid black. Figures 2, 3, 4, and 5 are - camera lucida tracings of the sections named in the text. farther apart, and through the remainder of the series each side of the section shows scar tissue 0.9 mm. in width extending through ' the entire wall of the auricle. No heart muscle crosses these two lines of crush which come to lie nearly opposite each other, that of Can Functional Union be Re-established? 383 the ventral side being AI WayR ait ated slightly in advance of the dorsal side (Fig. 5, section 259). _ Second crush. — It is somex what difficult to locate the ex- act line of the second crush owing to the amount of the extravasated blood, but if the | line of greatest injury be taken it can be stated that the line extends obliquely across the auricle from section 333 FicuRE 4. to section 276, where the line of the second crush crosses the old scar of the first crush. There is therefore an irregularly triangular area (area X) which is completely isolated from the rest of the auricle except for a nar- row band, 0.56 mm. in width, situated on the dorsal side of the auricle. FicureE 5. In some places nerves can be traced across the line of the old crush; they are however found in area Y, that is, outside of the triangular area described above. In some sections the heart muscle appears to be regenerating; but whether this appearance is due to an actual new growth or recovery of some individual muscle ele- ments which were injured at the time of the operation I am unable to determine. Wherever they are found they always lie along the old scar, but never penetrate into it. PSEUDO-FATIGUE OF THE SPINAE COkw) By FREDERIC S. LEE anp SUMNER EVERINGHAM. [From the Department of Physiology of Columbia University, at the College of Physicians and Surgeons, New York.] T is to be regretted that comparatively little is known of the phenomenon of fatigue in the central nervous system. Ver- worn’s! demonstration of the significance of oxygen and carbon dioxide in the activity of the spinal cord is valuable, although the employment of strychnin therein might not be universally approved. The belief seems to be still prevalent that the brain and spinal cord are more susceptible to fatigue than are the parts of the organism lying outside of them. Mosso, Lombard, and Waller have contributed experimental evidence in advocacy of this idea. Against it we find results obtained by Kraepelin, G. E. Muller, Henri, R. Muller, Hough, Woodworth, Storey, and Joteyko. In his “ Introduction to Human Physiology ’’ Waller ? makes the fol- lowing statement: “If a series of induction shocks is applied to a frog’s brain and bulb until the gastrocnemius has ceased to respond, a second series of contractions may be elicited by switching the cur- rent to a pair of electrodes applied to the sciatic nerve; if, when the muscle has ceased to respond to this excitation, the current is switched to electrodes applied to the muscle, a third series of con- tractions is obtained; from which we learn that maximum action of the superior organ does not elicit maximum action of the sub- ordinate organ — in other words, that central fatigue is limitative of peripheral fatigue — and we may formulate, as a probable con- clusion, that the incidence of normal voluntary fatigue is in dimin- ishing gravity from centre to periphery, — relatively greatest at the former, relatively least at the latter.’’ The present paper presents certain results of an attempt to study the fatigue of the central nervous system by direct stimulation of it. 1 Verworn: Archiv fiir Physiologie, Supplementband, 1900. 2 Watter: An introduction to human physiology, London, 1891, p. 551. > 354 Pseudo-Fatigue of the Spinal Cord. 385 We have experimented with both frogs and turtles during the months of October to March inclusive. Although we have worked with great care, we have found direct stimulation of parts of the brain, such as the crura cerebri or the medulla, to result in only meagre contractions of the muscles of the fore legs, and no contrac- tions of those of the hind legs. This has been the case, whatever the strength of stimulus and even when the circulation of the blood posterior to the cerebrum has not in any way been interfered with. With the spinal cord it is different, and most of our work has been on this structure. In both frogs and turtles it is easy to stimulate directly the spinal cord and obtain reflex contractions of a single muscle, the nerves to all other muscles being cut. We have used for a stimulus in some cases a rapidly interrupted induction current, but in most experiments a series of single induction shocks. Some- times we have used platinum electrodes, which were placed in close contact with and across the ventral surface of the cord, so as to enable the current to reach the motor tracts readily. At other times non-polarizable brush electrodes have been employed, the brushes being sufficiently long to enable them to be tied around the cord. All contractions have been recorded upon a slowly moving drum by means of an isotonic lever. With the frog our custom has been, after destroying the brain, to prepare both gastrocnemius muscles for the graphic record of their contractions, to cut one sciatic nerve near the spinal cord, and care- fully to expose the cord. The same stimuli were then applied simul- taneously to the dorsal cord and to the cut sciatic nerve, exciting one gastrocnemius through descending tracts, its motor centre and nerve, and the other through its nerve only. Simultaneous records of the contractions of the two muscles were made. Sometimes we found that the muscle stimulated through the cord ceased to respond before its opposite ceased, but sometimes the perplexing reverse occurred. Contractions resulting from cord stimulation were usu- ally not so great as those resulting from sciatic stimulation, which also was perplexing. Sometimes our object was defeated by an escape of the stimulating current from the cord into the nerve. Attempts to discover how far polarization occurred were frustrated by the short length of the cord. It was difficult to maintain, as fully as we wished, the circulation of the blood through the cord. We found it impossible also to-obtain a systematic and continued series of reflex contractions from stimulation of the skin, which we very 286 Frederic S. Lee and Sumner Everingham. | S much desired for comparison. In brief, we found the frog’s spinal cord an unsatisfactory object for our purposes. We therefore turned to the turtle. It was when we undertook experiments on this species that strong suspicions were aroused of the impossibility of fatiguing the spinal 1 Hl ANN F | FicurE 1.— Record of contractions of the ischio-caudali-tibialis (semimembranosus) of the turtle. To avoid shock the brain was destroyed twenty-four hours previously. A and B are reflex contractions of the muscle resulting from a mechanical stimulation, by pinching, of the skin in the region of the ischium. Between A and B is a series of reflex contractions resulting from electrical stimulation of the dorsal spinal cord. Break shocks, 16 per minute; secondary coil at 11.5 cm. with one Grove cell. FIGURE 2.— Record of contractions of the ischio-caudali-tibialis of the turtle. To avoid shock the brain was destroyed two hours previously. The blood supply of the spinal cord was left intact. The record begins with a series of reflex contractions resulting from electrical stimulation of the dorsal spinal cord. Break shocks, 12 per minute; appears to be well ’ secondary coil at 12 cm. with one Grove cell. After ‘‘fatigue’ marked, at B, the skin was pinched near the ischium, and the resulting reflex con- tractions of the muscle were obtained. cord by direct stimulation. It is possible in this species, as in the frog, to stimulate the cord above the sciatic centre and obtain con- tractions of an isolated muscle supplied by that nerve. The ampli- tude of the contractions is not great, however, and they cease com- paratively early. Hence the total amount of work performed by the muscle is little. This is the more striking when the record of such an.experiment is compared with that of the same muscle when acting Pseudo-Fatigue of the Spinal Cord. 387 under the reflex influence of a stimulation of the skin. Records of typical experiments of this nature are shown in Figs. 1 and 2. In both of these records there is observed an enormous dispropor- tion between the reflex contractions arising from direct stimulation of the cord and those resulting from stimulation of the end organs of the afferent nerve, yet the two centripetal impulses undoubtedly make use of the same centrifugal nerve mechanism involving the same motor neurone. The disproportion is not a matter of strength of stimulus, since, by even the most intense stimuli applied directly to the cord, only moderate contractions are elicited, while a slight pinch of the skin suffices for a most intense contraction. It is not a matter of shock caused by destruction of the brain. We have tested this in a variety of ways, and have found neither the early nor the late destruction of the brain materially to affect the result. It is not a matter of anemia of the cord; we have taken great care to avoid anemia. It is not primarily a matter of polarization of the cord by the electric current. Moving the electrodes along the cord posteriorly and thus bringing new regions into activity has either no pronounced effect or produces only a slight increase in intensity. Moreover, the disproportion is present with the very first contrac- tion. The disproportion is always present, no matter what portion of the cord is stimulated, whether the cervical, dorsal, or lumbar. The most reasonable explanation of the disproportion which has occurred to us is that stimulation of the cord excites both motor and inhibitory impulses, and that the subsequent contraction is the alge- braic resultant of the effects of the two. Stimulation of the cord as a whole is a stimulation of a variety of tracts. Presumably there is sent to the motor centres that give origin to the nerves of the muscle in question, other than simple motor impulses, and it is not improbable that they may interfere with the purely motor result that is desired. That, however, such a motor result is capable of being elicited from the motor neurones in question is shown by the experiment of pinching the skin and obtaining the enormous con- ‘traction of the muscle, which represents the natural protective re- flex from such a stimulation, namely, the flexion of the leg at the knee joint, and the pulling of the leg under cover of the carapace. In view of the possibility of this enormous reflex, it is evident that direct stimulation of the cord is very inefficient as a means of bring- ing into action the latent powers of the motor neurones in question, and the picture of fatigue as represented by such experiments as 388 Frederic S. Lee and Sumner Everingham. those of Figs. 1 and 2, is no true picture of fatigue of the spinal cord. The two modes of stimulation above employed bring into play the same set of motor neurones, but the latter are reached by differ- ent paths through different synapses. Sherrington® has studied the ‘ fatigue’ of reflex motor mechanisms, and especially that of a mechanism in which a common efferent path may be reached by two or more afferent paths. When stimulation of one of the affer- ent paths has ceased to be effective, contractions have again occurred from stimulation of the other. The experiment shows that neither the common efferent path consisting of motor neurone (cell body and nerve fibre), nor the muscle, is the seat of the “fatigue.” Sher- rington believes it to be localized at the synapse between the afferent and efferent paths. One cannot help thinking that the “ fatigue ” of such an experiment, which follows a few contractions, is not genuine or complete fatigue at all, comparable to that resulting from the action of toxic fatigue substances or the loss of substance essential to activity. It is a temporary condition, from which re- covery is easy and rapid. Perhaps it is due to a mild temporary asphyxia. That a particular part of the spinal cord is capable of performing an enormous amount of work has been abundantly demonstrated by the newer ergographic methods. Beside the ergographic records we can place certain graphic records that we have obtained from the turtle, where the pinching of the skin in the ischial region occurred approximately every third second and the reflex contractions were confined, by severing efferent nerves, to a single ischio-caudali- tibialis. This is a crude method of experimentation, but we have continued it during more than one hour without observing any evi- dence of fatigue in the preparation. Before these and the ergo- graphic records those obtained by Sherrington in his scratch reflexes and by ourselves through direct cord stimulation pale into insignifi- cance and urge irresistibly the conviction that neither Sherrington’s nor our ‘‘ fatigue’ is genuine or complete fatigue. We have chosen to call this condition pseudo-fatigue. Whether Sherrington’s pseudo- fatigue and ours are identical in nature is not entirely certain, but the one fact of which we feel very sure is that, by no method yet 3 SHERRINGTON: The integrative action of the nervous system, New York, 1906; also, Address to the Physiological Section, Proceedings of the British Association for the Advancement of Science, 1904. Pseudo-Fatigue of the Spinal Cord. 389 discovered of stimulating the spinal cord directly, has complete fatigue of the nerve substance of the cord been obtained. Hence neither our results nor Waller’s can be considered as demonstrating in any way the early fatigability of the central nervous system. We find indeed, in both frog and turtle, that after the muscle has ceased to respond to direct cord stimulation, stimulation of the nerve or the muscle directly will elicit further contractions. If however the cord is then not really fatigued, such a result is of no significance. Therefore Waller’s “ probable conclusion, that the incidence of nor- mal voluntary fatigue is in diminishing gravity from centre to peri- phery, — relatively greatest at the former, relatively least at the latter,” cannot be said to be demonstrated. Whether our pseudo-fatigue, as demonstrated in Figs. 1 and 2, is a condition localized at the synapses, is at present uncertain. It is clear, however, that it represents no fatigue whatever of the bodies of the motor neurones. ‘The reflex contractions resulting from the stimulation of afferent nerves after long-continued cord stimulation and the production of pronounced pseudo-fatigue are usually more intense than before the stimulation of the cord has occurred. This fact is highly significant. The problem whether the central nervous system is more or less resistant to fatigue in comparison with peripheral organs is still un- solved. In considering it the following few facts seem to be perti- nent. Sensations of fatigue have their primary sources in the fatigue of tissues or organs which are situated outside the central nervous system. In starvation the brain and the spinal cord are the last part of the body to lose in weight; their substance and working power are maintained to the last at the expense of other tissues. In certain diseases, notably in syphilis, the brain and spinal cord are attacked only after other organs have become involved. Thus, under various untoward conditions the integrity of the central ner- yous system appears to be long preserved. It would seem to be entirely in harmony with this fact and with the hierarchical position of this system in the living body that it should be resistant to fatigue. : CONCLUSIONS. 1. Direct electrical stimulation of the tracts of the spinal cord does not bring into activity the total working power of the central nervous motor mechanism of an individual muscle. 390 Frederic S. Lee and Sumner Everingham. 2. The slight “ fatigue’ that may thus be demonstrated does not represent complete fatigue of the central nervous motor mechanism of the muscle, and is no measure of the working capacity of the mechanism. 3. The result of such an experiment does not justify the in- ference that the central nervous system succumbs readily to fatigue. ee INNERVATION OF THE CORONARY VESSELS. By CARL J. WIGGERS. [From the Physiological Laboratory of the University of Michigan.| CONTENTS. I. The value and limitations of the perfusion method . .......2.2.2.. 391 Errors arising from insertion of the perfusion cannula ........ . 392 Errors arising in measuring the venous outflow .........-... 394 Enrorssntroduced by the beating heart -1- 25 : - - 2. 22 = Sr gal tae iif The reaction of the coronary vessels to adrenalin . ...-...2..-.-. = 57 3090 PRTERIOUSEC VIG ENCE Oho Tere en dans Sey, eyoen Se Sus et ee ee 396 Mieiniodsremployeduintmismesearcht =) 5) = =e en ls, 2 See nee ee 397 Saswilis, Gina! (doves treo he U Sto) 1A ee oe aS eS Se Sos SS 398 Ill. The effect of nerve stimulation on the flow through the intact heart . ... 399 PSTMIGLISBODSCLVATONSS coed co nr i ye Sk euler heed oe ae 400 MONO Saal noe ee ae a eee REE, lc eeliys. © 400 RUSTON BES Be SS PEN ot 5. Popa. Siate Se nc, ah nits |e) es pecan pena, eee 404 Il. THE VALUE AND LIMITATIONS OF THE PERFUSION METHOD. HE criterion by which coronary vasomotion has generally been judged has been the reaction given by the coronary vessels ‘when the heart was perfused and the flow through them was measured. Porter! was the first to employ this method in study- ing coronary innervation. He inserted a cannula into the aorta, perfused the heart with defibrinated sheep’s blood, and measured the variations in flow from a cannula in the superior vena cava as a criterion of vasomotor changes. Maas,” in 1899, made similar experiments, using the outflow from the inferior vena cava as an index of coronary flow. In 1904 Schafer * pointed out what he deemed an imperfection and source of error in this method, con- - tending that the aortic valves were not effectively closed and that considerable leakage of fluid into the ventricle was liable to occur. * Porter: Boston medical and surgical journal, Tan. 8, 1896. * Maas: Archiv fiir die gesammte Physiologie, 1899, Ixxiv, p. 28r. 3 ScHarerR: Archiv des sciences biologiques, 1899, xi, suppl. vol. p. 281. 391 392 Carl J. Wiggers. For this reason he modified the perfusion method by pushing the cannula beyond the valves. This allowed the fluid to distend the left ventricle and pass by the side of the cannula into the coronaries, its back escape through the pulmonary veins being prevented by a ligature. As a measure of the coronary flow, Schafer used the venous blood escaping from the caval openings into a funnel below. Errors arising from insertion of the perfusion cannula. — The most “convenient method of perfusing the heart is that utilized by Porter and by Maas. A cannula is inserted into the aorta or one of its branches, and the coronary arteries are perfused, since the pres- sure tended to close the aortic valves. As doubt has arisen concern- ing the perfect closure of these valves, a series of experiments was undertaken to test this point, for outflow changes obviously are in no wise trustworthy unless all the fluid passes through the coronaries. The hearts of dogs, cats, and rabbits were perfused by this method, and the amount of fluid passing through the coronaries was measured by collecting the overflow from the filled right auricle and ventricle in the pan of a drip-recording apparatus, recently de- scribed. The fluid entering the left ventricle was drained by a cannula pushed directly through its musculature, and was led by a thin rubber tube to a closed receptacle, which in turn was connected with a bellows recorder; in short, the method employed by Brodie and Dixon ® to continually record outflow was used. The bellows recorder differed, however, from those ordinarily used, in that it possessed a device for altering the magnification of the bellows moveinent and wrote a vertical line instead of an arc on the drum. A series of 20 experiments showed that the leakage through the aortic valves varied markedly. In some preparations the leak could by adjustment be reduced to a very small quantity, but in other prep- arations it persisted after the most painstaking adjustment. I have records in which the leak equalled from 36 to 48 per cent of the flow through the coronaries. That these figures represent a leak through the valves rather than a flow through the Thebesian vessels into the left ventricle is evidenced by the fact that the flow decreased to from 3 to 8 per cent when the cannula was inserted into the coronaries. The fallacy in the method consists, however, not so much in the degree of leak as in its tendency to vary. Thus the leak ‘ WiccErs: This journal, 1908, xxiii, p. 23. ° BropiE and Drxon: Journal of physiology, 1904, xxx, p. 478. The Innervation of the Coronary Vessels. 393 has been observed to increase (1) as cooled hearts were gradually warmed; (2) when the pressure rose; (3) after the introduction of adrenalin, probably because it reduced the tonus of the aortic ring; and (4) when no cause was apparent. Fig. 1 illustrates how such a change of leakage may exert an inverse effect on the flow through the coronaries, without materially changing the perfu- sion pressure. These results seem to war- rant the following conclusions: 1. The method cannot be util- ized to study the effect of adrenalin or any other drug which affects the tonus of the aortic ring. 2. The method may be used to test the effect of nerve stimulation on the coronaries, but to render re- sults free from criticism, they should be accompanied by rec- Ficure 1.— Three sevenths the original size. ords showing that the amount P.P., perfusion pressure; 7, time in ten sec- : ; or = onds; V. O., flow through coronaries recorded of fluid entering the left ven- : : : I yy author’s apparatus; A. L., leak through tricle has not changed during aortic valves recorded by Brodie’s method. stimulation.® Several times | V..S., vagus stimulation. Published to show have obtained what seemed per- at X’ sudden increase in leak causing change Sean : in flow through coronaries at X. Taken on ceptible changes in the outflow : _ : non-beating heart. record during nerve stimula- tion which might be attributed to vasomotor changes, only to find that the increased leak through the valves counterbalanced the de- crease, making the total effect nil. Schafer’s modification is, however, no less open to criticism. In this procedure the pressure under which the coronaries were per- fused, as well as the rhythm of the pressure, was determined by the contraction of the left ventricle. As long as the heart beat regu- § This cannot be accomplished by recording the constancy of the perfusion pressure alone. Fig. 1 shows that this need not be an index of competent-valves. Either the outflow from the left ventricle must be recorded, as was done by Magrath and Kennedy (Journal of Experimental Medicine, 1897, ii, p. 13) or by recording the changes in intraventricular pressure with a sensitive membrane manometer, as was done by Miss Hyde (This journal, 1808, i, p. 215). 3904 Carl J. Wiggers. larly the pulsating pressure remained constant, but, when nerves were stimulated or drugs introduced which affected the rate, rhythm, or tonus of the heart, the vessels were no longer supplied by the same pressure. The curves published in Schafer’s article, as well as that produced in Fig. 2, give evidence of this. It seems that this pro- cedure violates a prin- ciple in the perfusion of organs, namely, that the supplying pressure must remain constant or vary within fixed limits if the outflow changes are to Aven cite be used as a criterion ft of vasomotor changes. The objections to both of these methods may be removed by insert- ing the cannula directly - into the coronary ves- | | FP FB AB BN fA Pt sels, but this technic ‘IGURE 2.— original size. Schafer’s 2 c FIGURE 2. One half the original size. Schafer’s method has so far not been em- of perfusion. HA, heart beats recorded by straw lever; P.P., perfusion pressure; V.O., venous outflow; 7, ployed, to my knowl- time in ten seconds. At X pointer caught. Effect of edge, in order to test adrenalin, 0.15 mg. Heart rate increased, tonus de- the effect of drugs or creased, systole augmented, perfusion pressure elevated, nerve stimulation on the venous outflow thus mechanically increased. . coronary vessels. Errors arising in measuring the venous outflow. — The fluid per- fused through the heart returns by the coronary veins into the right auricle and by the Thebesian vessels into the right ventricle, only a small quantity passing by similar vessels into the left ventricle. As there is no single vessel into which a cannula may be inserted and the total outflow thus received, the right auricle and ventricle have been used as intermediary reservoirs between the place of venous discharge and the recording device. Thus Porter, as well as Maas, believed that the outflow from the right auricle and ven- tricle through a vena cava could be used as an index of the flow from the coronary veins. Maas, however, made no attempt to measure the flow from the Thebesian vessels, while Porter did so in one experiment only. The return flow’through these Thebesian vessels is so considerable that it should always be recorded in esti- The Innervation of the Coronary Vessels. 395 mating changes in the flow of blood through the heart. Schafer allowed fluid to fill the right chambers, and then recorded the over- flow from the auricle and ventricle by allowing fluid to leak out of the ven cave and pulmonary artery. This method adequately measures the blood returned to the right heart as long as its beat or tonus remains unaltered, for as much blood will drip away as is added by the heart vessels. When the beat of the heart is changed through the stimulation of nerves or the administration of drugs, the overflow no longer gives a true indication of coronary flow. I have observed, for example, that the first few augmented beats in- duced by adrenalin caused a greater amount of fluid to be expelled from the cavity of the right heart, which, when recorded, produced the erroneous impression that the coronary flow had been suddenly increased. It is evident, then, that the flow from the right chambers gives reliable information concerning the coronary flow only when the fluid passing through the Thebesian vessels is recorded, as well as that from the coronary veins, and when such outflow records are accompanied by graphic records showing or warranting the assump- tion that the size of these chambers was not altered by changes in contraction or tonus. Errors introduced by the beating heart. — Since fluid is pushed through the intramural vessels by the compressing action of systole, and the refilling of these vessels is determined to a great extent by the degree of relaxation during diastole, it follows that changes in outflow cannot be interpreted as due to vasomotor action unless it can also be shown that neither the tonus nor contractions have altered.’ I have a number of records showing that nerve stimula- tion may cause a change in tonus without change in heart rate, and this may account for the changes in flow observed. Maas empha- sizes the statement that in order to bring out vasomotor changes a strong current is necessary. I have been able to show that changes in outflow may indeed be obtained with strong currents when weak ones fail, but inspection of the cardiac tracings, 1f delicately re- corded, reveal beautiful tonus changes, probably due to a spread of current to the heart. The curves published by Maas himself, as Figs. 10 and 15, show, when a line is drawn perpendicular to the lower portion of the heart contractions, that they rise (tonus 7 PortER: This journal, 1900, xxiv, p. XXiv. 396 Carl J. Wiggers. change’). Since Porter took no graphic records to rule out these changes, and as his observations on non-beating hearts were limited apparently to a single experiment, his results offer, as he himself words it, “ probable rather than quite certain evidence’ of a vaso- motor influence.S Schafer accompanied his experiments with trac- ings, but assigned all changes in flow to synchronous changes in heart beat. I have been able to corroborate Schafer’s results showing that adrenalin increases the flow through the beating heart. Such a reaction, however, need not necessarily be interpreted, as it was by Schafer, as indicating the absence of an adrenalin action, for, by increasing the amplitude and rate of cardiac contractions, the ves- sels would be more vigorously massaged, and so an increased flow might occur in spite of a constrictor action that the adrenalin might have. The effect of cardiac changes on the flow through the heart may be eliminated by inducing standstill. After trying various pro- cedures to attain this end I have found that perfusion with a non- oxygenated sodium chloride solution accomplishes it. If the per- fusion occurs immediately after death, the influence of nerves on the heart vessels may be tested, while, 1f an interval of several hours is allowed to elapse before perfusion, adrenalin and other drugs will not, as a rule, cause a revival of contractions, but will still affect the blood vessels. This preliminary investigation has shown that any decrease or in- crease in outflow from the right heart cannot be interpreted as due to vasomotor changes unless it can be shown (1) that the amount of fluid supplied to the coronary vessels in a given time remains constant, (2) that the size of the right chambers, which exist as intermediary reservoirs between the heart veins and the registering apparatus remains constant, and (3) that the massaging effect on the intramural vessels has not altered. Il. Tuer REACTION OF THE CORONARIES TO ADRENALIN. Previous evidence. — Adrenalin has come to be regarded as a con- venient agent for obtaining presumptive evidence of the innerva- tion of blood vessels, for, even if a constrictor reaction is not unt- versally accepted as a proof of nerve control, the lack of such a % PorTER: This journal, 1900, iii, p. XXiv. $ f : The Innervation of the Coronary Vessels. 397 reaction is enough to arouse a question as to the existence of such an innervation. So far a constrictor reaction of the coronary ves- sels has not been demonstrated. In fact, Schafer * not only failed to obtain any evidence of constriction, but showed that the outflow from the coronaries was often increased. In 1907 Langendorff ® pointed out the difficulty of determining accurately by the perfusion of hearts whether or not adrenalin exerted any action. He accord- ingly abandoned the method, and recorded instead the movements of a strip of artery from an ox heart, which was suspended in warm Locke’s solution. Adrenalin, tested by this method, caused only a relaxation, while an induction shock caused a contraction. Methods employed in this research. — The coronary arteries were perfused by a stream of normal salt solution regularly interrupted by a stopcock which was actuated through a cam and motor. The perfusion cannulas were inserted into the mouths of the two cor- onaries, or simply into one while the other was clamped or ligated. This procedure was found necessary, for sometimes the terminal nature of the coronaries did not prove true when the: isolated and suspended heart was perfused at various intervals after death with a solution having less viscosity than the blood. Perfusion with normal salt solution instead of the customary Locke’s solution or defibrinated blood brought the heart to complete standstill, so that the simultaneous effect of the drug on the heart was entirely elim- inated. The venous outflow was generally determined by recording the overflow from the right auricle, right ventricle, and left ventricle after all of these chambers had previously been filled with saline solution. In this way the total flow from the Thebesian vessels as well as from the.coronary veins was obtained. This overflow could be taken as evidence of the amount flowing through the coronaries, since the size of the chambers did not vary owing to the cessation of the heart. In a number of experiments the flow from the right ventricle was drained by a cannula passed through its musculature, and in still others the coronary sinus was incised and the flow allowed to drip away directly. : The changes in the diastolic portion of the perfusion-pressure oscillations were simultaneously utilized to corroborate the changes in calibre of the vessels. As the vessels were supplied by a stream of fluid rhythmically interrupted, it followed that the pressure in a laterally connected manometer rose whenever the stopcock was open, 9 LaNGENDOREF: Centralblatt fiir Physiologie, 1907, xxi. 298 Carl J. Wiggers. and fell whenever it was closed. The extent of the fall depended entirely on the amount of fluid passing through the heart vessels during the interval in which the stopcock remained closed. The effect of a constriction would, by diminishing the amount of fluid passing through an organ, cause less of a fall, and so give the ap- pearance of a rise in the diastolic pressure in a series of such oscil- lations long before the systolic pressure could be influenced. In a FIGURE 3. — Effect of 3 various-sized doses of adrenalin on dog’s coronaries as indicated by change in diastolic portion of perfusion pressure. Recorded by membrane ma- nometer, March 15, 1908. number of experiments a membrane.manometer was substituted for the mercury manometer, and the oscillatory changes thus magnified. In this way the slightest vascular changes could be determined. The adrenalin used was the crystalline product supplied by Parke, Davis, & Co., or tablet titurates made of the same substance. This was dissolved in salt solution in the desired quantities just before use, so that deterioration of the solution and the presence of pre- servatives were avoided. Results and their discussion. — By these pré cedures it was deter- mined that adrenalin was capable of constricting the coronary ves- sels in the dog, cat, and rabbit, while a dilatation was never obtained. In dogs doses ranging from 0.1 to 3 mg. caused a decreased flow and a rise in the diastolic pressure. The reaction became weaker as the dose was diminished, but a dilation never occurred (Fig. 3). The Innervation of the Coronary Vessels. 399 Adrenalin failed to react more than eighteen hours after death, although digitalein was still found to respond. In the cat and rabbit a similar constriction was observed. Though the change in outflow and oscillations was never very great, owing to the small size of the coronary vessels, it was al- ways clearly evident when the records were properly magnified (Fig. 4). A word of explanation re- garding the results of other workers seems necessary. In Schafer’s results the slight vasomotor change present in hearts as small as those of rabbits and cats were prob- ably obscured by the greater Beeeet ot the drug Smt FIGURE 4 Ine half the original size Effect = of adrenalin on cat’s heart when not beating. heart (consult previous sec- pp. perfusion pressure; V. O., venous outflow tion). In regard to the re- recorded by author’s apparatus; X, X’, relative sult of Langendorff, it may Position of pointers. be said that either the vessels of the ox heart behave differently to adrenalin, or he used an old solution which had lost the constrict- ing power of the adrenalin, but retained the dilator influence of its preservative, the chloretone. IJ. THe Errect or NERVE STIMULATION ON THE FLOW THROUGH THE INTACT HEART. By using the perfusion method I have been unable to obtain a change of flow during nerve stimulation which could not be equally well explained by some factor other than a vasomotor influence. If the perfusion cannula is tied into the aorta or one of its branches, as it has been by previous investigators, the nerve supply may be kept intact, but the results are complicated by the loss of fluid through the aortic valves. _If the technic employed in the previous part of this research to study the action of adrenalin is used, the nerves supplying the vessels are probably included within the liga- tures that tie the cannulas into the coronary vessels. Beset with these difficulties, I abandoned the perfusion method temporarily and 400 Carl J. Wiggers. made an attempt to study the effect of nerve stimulation on the vessels of the intact heart. (See Addendum.) Previous work. — Observations in regard to the effect of nerve stimulation on the vessels of the intact heart have been at variance with each other. Inspection of the vessels seems to have been the favorite procedure. Panum is reported in Schmidt's Jahrbuch of 1858 to have observed a constriction of the coronary vessels after vagus stimulation. In 1869, however, Meyer ?? thought he saw a_ dilation when the vagus was stimu- lated. Similar observations were reported in 1891 by Martin”? To ‘what ‘extent concomitant changes in the heart and _ general pressure were ruled out is not clearly FicurE 5.— Diagram showing relation of indicated. (Quite recently Do- heart vessels. C. A., coronary artery; giel and Archangelsky * Le C. V., coronary vein. Vein woundedat A duced the heart to standstill and ligature to prevent back flow applied py yaous stimulation, and then at B. oo 6 observed a constriction of the arteries on stimulating the annulus of vieussen and the inferior (middle) cervical ganglion. Their article is illustrated by photo- graphic plates. 3 Method employed in this research. — [n this research the amount of blood flowing from one of the cut veins of the heart during nerve stimulation was compared with a similar flow when no stimulation occurred. A dog was anesthetized with morphine and chloretone, and the carotid artery together with the nerves to be stimulated was isolated. After instituting artificial respiration the thorax was opened and the heart exposed. The animal was then inverted and suspended in a hammock arrangement above an outflow-recording apparatus, recently described.4 When the carotid pressure or the 0 MEYER, quoted by Doctet and ARCHANGELSKY: Archiv fiir Physiologie, 1907, p. 482. Martin: Transactions of the Medical and Chirurgical Faculty of Maryland, * Doctet and ARCHANGELSKY: Archiv fiir Physiologie, 1907, p. 482. The Innervation of the Coronary Vessels. 401 contractions of auricles and ventricles were being satisfactorily recorded, one of the veins accompanying the descending branch of the left coronary artery (Fig. 5) was wounded with the point of a pin or scalpel, and the blood leaving the wound was caught in the pan of the registering apparatus below. As coagulation over the vessel was mechanically delayed Muee sbeatine of, the heart, the ——————_. 7.1 quantity of blood flowing from ie \ the vein during equal-time inter- vals remained constant for a con- siderable length of time, with the result that it was recorded as a straight oblique line on the mov- ing drum. That the method may be used to indicate changes in the amount of blood flowing through the heart veins was shown by the fact that when the VagOSyMDPa- FrgurE 6.— One half the original size thetic of either side was stimu- V.S., stimulating the vagus with a mod- lated the well-known diminution erate current; B. P., effect on blood pres- ee dow was obtained. SOnICh gave sure and heart rate. These factors cause : change in venous outflow (V. O.). place to an increased flow when stimulation ceased. Fig. 6 illustrates the nature of these results. This diminution was due in part to the fact that the blood pressure lowered, and in part to the removal or diminution of the normal squeezing action of the cardiac systole. These were not the only factors concerned, however, for a decreased flow still occurred when the effect of the vagus on the heart was prevented by the previous introduction of atropin or by the use of a weak current which had no effect on cardiac contractions (Fig. 7). These results having been obtained from two to five times con- secutively in 7 experiments, were presented before the American Physiological Society at Baltimore as probable evidence of a vaso- constrictor influence of the vagus on the heart vessels. The reason- ing by which the conclusion was reached was somewhat as follows: The results certainly indicate that the quantity of blood contained in the large veins of the heart is decreased during such stimulation. Into these veins blood flows (1) from the arteries by the capillaries, 402 Carl J. Wiggers. (2) from the ventricles (especially the right) by the Thebesian vessels, and (3) from the auricles by a back flow through the incom- petent venous valves. When the outflow from an opening in one of the veins decreases after nerve stimulation, in spite of the fact that changes in the blood pressure and in the contraction of auricles and ventricles have not occurred, there is no reason to suppose that a a I aT a a ea a Ly pe Rt. psa - nd ca Venous outflow. Fi J FIGURE 7.— One half the original size. Effect of stimulating vagus (weak current) on venous outflow when no change in contractions of auricle and ventricle were present. the back flow has diminished unless the large veins have themselves been constricted. It cannot be so certainly said that the flow through the Thebesian vessels was not diminished, but, making this assump- tion,-one can only conclude that these must have been constricted by nerve stimulation. Since, then, the decrease is most probably not due to these factors, it only remains to assume that the arterioles were constricted. Such evidence, however, remains open to objection as long as the back flow from the auricles is not actually controlled. Theoretically this back flow should be easily prevented by ligating one of these veins and wounding it peripherally. Practically obstacles to the employment of this technic immediately arose, — a first in including the vasomotor nerves within the ligature, a second in producing an undue venous stasis in the heart vessels. After some experimenting, however, it was found that a place could be located where a ligature might be applied by a needle and yet not cause either of these effects. This place was where the veins and artery parted company, the The Innervation of the Coronary Vessels. 403 zontal direction (Fig. 5). When a ligature was so placed, the end of the vein toward the auricle was seen to swell mark- edly, while the veins of the heart diminished in size. The outflow from a wounded vein which before had been a spurt- ing stream was now trans- formed into a dropping one. This introduced a further obstacle. The beating heart which ade- quately prevented clot formation as long as a spurting stream left the vein was no longer able to do so when the out- flow was reduced, thus clot formation interfered seriously with accurate records. It was conse- quently necessary to ren- der the blood non-coagu- lable. Accordingly 150 to 250 c.c. of blood were withdrawn at_ half-hour intervals, defibrinated, and then reinjected until little or no coagulation occurred. In the first experiments in which this procedure was adopted the results after vagus stimulation were negative. [n all of these cases, however, a great deal of hemorrhage had occurred on opening former assuming a_hori- | = | : | | vey! Blood of animal V. O., venous outflow; B. P., blood pressure; T, time in { aha Matinee Sere errr Effect of stimulating the vagus with a weak current twice consecutively. Ww igi A Avvul\yvoyyondVvethavdndthvlvadbdbvnwnnl rd Mvayendivaiyltnythyyd Avni fro Bc eee SEES SSP TES ET RCT ORTON ERROR eer TROUPE ET TET rendered non-coagulable, back flow from auricle prevented by ligature of vein. CurvE 8. — About two thirds the original size. ten seconds and in seconds. Ll Sryivarlyw wy {vyellywlyylaatnys V.S V.O 404 Carl J. Wiggers. the thorax, because of the non-coagulability of the blood, and often the heart had to be kept beating by the use of strychnine and digitalis. At other times a marked diffusion of hemoglobin had taken place throughout the entire heart. In later experiments the thorax was always opened before defibrination of the blood, thus utilizing its coagulating power to check hemorrhage. Great care was also taken not to whip the blood too vigorously nor to heat it above 38° C. be- fore injection. By attention to these details the animal with its blood defibrinated was in as good a condition as before. In such cases stimulation of the vagus caused the flow from the peripheral end of the vein to decrease. This decrease could be obtained many times in succession. In Fig. 8 is presented such a decreased outflow, obtained six consecutive times without any change in blood pressure or heart action. CONCLUSIONS. The conclusion that the coronary vessels are not influenced merely in a passive manner by the blood pressure and the massaging action of the heart, but that they possess a nerve control, is supported by the following evidence: 1. The coronary vessels react to adrenalin by constricting, when the simultaneous effect on the heart is eliminated. 2. Stimulation of the vagus nerve causes in the dog a decreased outflow from a wounded heart vein. This happens in spite of the fact that changes in the blood pressure, contractions of the auricle and ventricles, and back flow do not occur. It is with pleasure that I acknowledge the suggestions made by Professor Schafer during his visit to the laboratory last summer. ADDENDUM. PRELIMINARY Note: — Recently the writer and Mr. H. Cum- mings have devised a method of perfusion which promises to deter- mine whether nerve stimulation can cause changes in flow through the perfused heart which may be attributed only to vasomotor changes. The mouth of the left coronary artery is closed by ap- proximating the intima around it with a stitch. The right coronary is ligated and a cannula is inserted into the central end of one of the } The Innervation of the Coronary Vessels. 405 descending rami of the left coronary artery. By this backward per- fusion through the arteries, the nerve supply of the vessels is not in- cluded within the ligature tying the cannula into the coronaries. Experimentation by this method will be resumed next autumn and a report on the results made later. THE COAGULATION OE BEOOD: BY LL. J; REDTGER: [From the Physiological Laboratory of the Johns Hopkins University.] ECENT investigators of the phenomenon of blood coagulation have come to rather widely divergent conclusions. Mora- witz,' attempting to harmonize the older views of Schmidt with the later work of Pekelharing, Hammarsten, and others, gives four factors entering into the process. Three of these pre-exist in the circulating blood, — the fibrinogen, the thrombogen or proferment, and the salts of calcium. The fourth factor arises at the moment of coagulation, and is produced by the disintegration of the formed elements of the blood,—the platelets and probably also the leu- cocytes. This substance is the thrombokinase, which activates the thrombogen in the presence of calcium salts to form thrombin. The thrombin then attacks the fibrinogen molecule and in the manner of a ferment or catalyzer changes it to fibrin. The thrombokinase may also be derived from the various tissues of the body, and when so extracted serves to activate the thrombin and initiate coagulation. The suggestion of a “ kinase,” as made by Pawlow in the study of the entero-kinase in the intestines, is here applied to explain the action of the “ zymoplastic ’’ substances of earlier workers. On the other hand, Nolf,? impressed by the fact that a clot is soon re-dissolved by an act of autolysis, especially in the case of clots of fibrinogen produced by thrombin, sees in the phenomenon of coagulation, as the paramount question, not what makes blood clot, but what makes the clot persist. According to his view coagu- lation is a normal process in nutrition and assimilation, whereby the fibrinogen of the blood is rendered available as food for the tissues. ' Morawitz: Beitrige zur chemischen Physiologie und Pathologie, 1904, iv, p. 381. Morawitz: Ergebnisse der Physiologie, 1905, iv, p. 307; contains a full bibliography. * Notr: Archives internationelles de physiologie, 1906, iv, p. 165. Ibid., 1908, Vi, pp. I, 115, 306. Jbid., 1909, vii, pp. 281, 380. 406 The Coagulation of Blood. 407 Coagulation is normally and continually going on in the blood. It consists of a quantitative and mutual precipitation of two colloids, the thrombin and the fibrinogen. The thrombin is formed by the union of the “ hepato-thrombin,” which has its source in the liver and corresponds in a way to the thrombogen of Morawitz, and “leuco-thrombin,”’ which is derived from the leucocytes and plate- lets of the blood. In the circulating blood small amounts of throm- bin are formed around the leucocytes in the manner described. This precipitates a small amount of fibrinogen as fibrin, which covers the leucocyte with a thin ultra-microscopic layer of fibrin. This layer of fibrin renders the leucocyte neutral to some extent, in the matter of further production of leuco-thrombin. ‘This precipitated cuticle of fibrin is now attacked by the real enzyme of the thrombin, the “thrombozyme,”’ and dissolved. This autolytic action is a true ' fermentative process, and serves primarily purposes of tissue nutri- tion. In abnormal conditions, such as produce ordinary clots, the leucocytes and platelets are destroyed in great numbers, and the ex- cessive amounts of leucothrombin so formed unite with hepato- thrombin to form the active thrombin that produces the clot. The real question here is, what makes the clot persist? This anti-fibrin- olytic substance he believes is the hepato-thrombin, to the inhibiting action of which on the fibrinolysis the clot owes its persistence. Coagulation ‘is therefore, according to Nolf, a late phylogenetic adaptation. It is a secondary modification, for preventing hem- orrhage, of a primitive nutritive and assimilative process. Loeb,? approaching the problem from a.study of blood coagula- tion in the invertebrates, assigns the primary role in the process to the coagulins. These coagulins may be extracted from the tissues, and when activated with calcium salts have an effect on the fibrin- . ogen similar to ordinary thrombin. These coagulins are not merely zymoplastic substances, nor are they kinases. They do not activate any proferment, but independently attack the fibrinogen molecule. Neither are the coagulins identical with ordinary thrombin. They have a much greater specificity than thrombin, and are more stable. The fact that tissue extracts properly irrigated will not clot care- fully prepared fibrinogen solutions, even when calcium chloride is 3 Lors, Leo: Archiv fiir pathologische Anatomie, 1906, clxxxv, p. 160; LOEB: Beitraige zur chemischen Physiologie und Pathologie, 1906, viii, p. 67; Ibid., 1907, ix, p. 185. LorB: Biochemisches Centralblatt, 1907, v1, pp. 829, 889, excellent bibliography given. Lors: Beitrige zur chemischen Physiologie und Pathologie, 1909, XVi, p. 157. 408 Leds Reetieer- added, is explained by assuming that the repeated salting out has changed the character of the original fibrinogen. Mellanby,* working with the blood of birds, assigns the main role in the coagulation of blood to the kinases. These kinases, extracted from the tissues, activate the proferment and form fibrin. The proferment seems bound in a way to the fibrinogen, so that the amount of proferment varies directly with the amount of fibrinogen present. As the source of his kinase he used the macerated testes. The fibrinogen solutions were made by diluting the clear bird’s plasma with twenty times its volume of water, acidulating, centri- fuging, and re-dissolving the precipitate in the original volume of water. The present investigation was carried on under the direction of Dr. W. H. Howell, to whose stimulating interest and many valuable suggestions most of any credit which this work may have is largely — due. The first requisite in the investigation of blood coagulation is a trustworthy testing solution. The ideal solution would be the cir- culating blood, if it could be kept in its normal condition. This is of course impossible, at least with mammalian blood, inasmuch as it begins to clot immediately after taking. Recourse must therefore be had to artificial plasmas. It is believed, for reasons to be stated further on, that the most satisfactory testing plasma is a solution of fibrinogen, prepared essentially after the manner of Hammarsten, with two modifications. The method of preparation of mie fibrinogen solutions used in this work was as follows: On account of the fact that the blood of the cat does not “lake” so readily as that of many other mammals, this animal was used in the majority of experiments. The blood was taken under careful conditions from the carotid artery, through a cannula and short tube either oiled with some neutral oil or washed out with a 1 per cent solution of sodium oxalate. A I per cent solution of sodium oxalate was added in the proportion of one part of oxalate solution to nine parts of blood. The blood was then immediately centrifuged, until a clear supernatant plasma was obtained. Sodium chloride was added to a little more than 50 per cent saturation, and the precipitate after centrifuging, and washing in a saturated solution of sodium * MeELLANBy: The journal of physiology, 1909, xxxviii, p. 28. The Coagulation of Blood. 409 chloride by decantation, was re-dissolved in a 2 per cent solution of sodium chloride of volume equal to the original plasma. A second and a third precipitation in the same manner produced a final precipitate, which was dissolved in 0.9 per cent solution of sodium chloride. If such a solution of fibrinogen still remained distinctly opalescent, 1 c.c. of a dilute solution of barium chloride and I c.c. of a dilute solution of sodium phosphate were added. The dilution of the two solutions was approximately such that 1 c.c. of the barium chloride and 1 c.c. of the sodium phosphate neutral- ized each other. The precipitate of barium phosphate so formed gradually fell to the bottom, and with it were swept out the particles giving the opalescence, leaving finally a clear solution of fibrinogen, not very different from the clear appearance of distilled water. The solution of fibrinogen was then placed in a collodion tube and dialyzed against a large volume of 0.9 per cent solution of sodium chloride for about eighteen hours. This insures a practically complete removal of any oxalate remaining, or of any excess of barium chloride or sodium phosphate. It also insures to the fibrin- ogen solution a known concentration of sodium chloride. Such test solutions of fibrinogen do not clot spontaneously, even when kept until putrefactive changes arise. They give no appear- ance of a clot on the addition of calcium salts, showing that no pro-stage remains in the fluid which can be activated by calcium. No clotting occurs either when Ringer’s solution is added, such as occurs frequently in fibrinogen solutions, which have been precipi- tated only once or twice and which remain strongly opalescent. Probably in such opalescent solutions some of the fibrinogen has already been affected by the thrombin present. It is believed there- fore that by this method a solution of fibrinogen was obtained which was entirely free from any substance which could give rise to the thrombin. Only with the aid of such solutions can reliable experiments be made upon the presence or absence of a coagulating agent. That solutions of fibrinogen prepared as here described are, how- ever, not materially changed from the condition found in the blood is shown by the fact that with serum or thrombin they form a typical clot, which resembles in every essential the normal blood clot. The promptness of the action and its similarity in every way to clots of other forms of test solutions prove clearly that the fibrinogen in question has retained its characteristic properties, and by its purity 410 ie. J Ren oer: and stability is peculiarly fitted to study the real coagulating effects of other factors entering into the process. With this fibrinogen solu- tion the main constituents of normal plasma were tested as to their coagulating properties. As already pointed out in describing the method of preparing fibrinogen, the inorganic salts of the blood in normal (Ringer) concentrations had no effect, either added singly or together or in other possible combinations. Similar negative re- sults were found in the case of solutions of lecithin and cholesterin. Carbon dioxide and oxygen gave negative results. The experiments - showed, as has long been known, that the initiating factor of coagu- lation is not among the simpler chemical substances which exist pre- formed in the circulating blood. THE THROMBIN. The observation that serum added to a solution containing fibrin- ogen will induce clotting dates back to the work of Buchanan. Later Alexander Schmidt succeeded in getting the active substance in somewhat purer form by treating the serum with alcohol and extracting the dried residue with water. This substance he desig- nated fibrin ferment, or thrombin. The thrombin used in these experiments was prepared after Schmidt’s method. Blood was drawn from an animal and allowed to clot. As soon as the serum was sufficiently pressed out of the clot, it was diluted with twenty times its volume of alcohol (95 per cent) and allowed to stand for a period: ranging from several days to several months. The duration of the action of alcohol seemed to have no very appreciable effect on the quantity or strength of the thrombin extracted. After the proteins had been precipitated by the alcohol they were filtered off, dried, and extracted with water. Into this aqueous extract other things as well as thrombin will enter. Evidently the free salts of the serum will be dissolved. Some of the proteins also go into solution, especially if the aqueous extraction is continued for any considerable length of time. Unfortunately therefore such thrombin solutions are not very pure, and offer serious. obstacles to a study of the chemical nature of this substance. If however the extraction is made of short duration, it is possible to secure a thrombin solution which is very active in producing coagulation and which nevertheless gives only the faintest traces of The Coagulation of Blood. 411 protein. The excess of salts may be removed by dialyzing the thrombin solution in a collodion tube against pure water. Experi- ments soon showed, however, that the dissolved proteins interfered in no way with the thrombin action, and some of the strongest thrombin solutions were secured by submitting the dried residue not only to prolonged aqueous extraction, but even to prolonged putre- faction, until the protein residue disappeared. Such solutions usu- ally gained in amount of thrombin, as the processes of putrefaction continued, the possible explanation for which will appear in another connection. It is evident that the thrombin extracted from the serum is of the greatest importance in the process of coagulation. It appears in the plasma at the time of coagulation and is the agent which effects the change from the liquid fibrinogen to the solid fibrin. It seemed desirable therefore to determine the conditions under which it acts, its origin, and as far as possible its chemical nature. Rapidity of coagulation with thrombin. — The rapidity of coagula- tion was found to be directly proportional to the amount of throm- bin added, as indicated in the following experiment : 20 c.c. fibrinogen + 40 drops thrombin extract, 14 min. 20 c.c. fibrinogen + 20 drops thrombin extract, 26 min. 20 c.c. fibrinogen + 10 drops thrombin extract, 50 min. 20 c.c. fibrinogen + 5 drops thrombin extract, 90 min. The time required for coagulation was determined by the moment when the glass could be inverted without spilling the contents. Such a rate, however, holds good only below certain limits. If the amounts of thrombin become excessively large, the rapidity of coagulation does not vary in similar proportion. There is soon reached a maximum amount beyond which the further increase of the amount of thrombin no longer increases the rapidity of coagula- tion. The evident reasons for this appear in a study of the quantt- tative relations of thrombin and fibrinogen. Quantitative relations of thrombin and fibrinogen. — The original statement of Alexander Schmidt that minimal amounts of his _“fibrin-ferment ’’ always produced complete coagulation was soon questioned. It was found that partial clots could arise, or that a second or even third successive clot might form in the same fibrino- gen solution or plasma. Careful quantitative experiments seem, however, not to have been made. For the experiment on this point 412 1, Teetveer 500 c.c. of clear plasma of four cats were taken, and from this plasma 500 c.c. of fibrinogen solution were prepared. This amount was divided into four equal parts. To part one 5 drops of a tested | thrombin solution were added; to part two, 10 drops; to part three, 20 drops, and to part four, 40 drops. The glasses of fibrinogen were then allowed to stand untouched for twenty-four hours at room temperature, to allow ample time for coagulation. The fibrin from each glass was then collected on a filter, washed, and placed in a weighing tube, which together with the filter had been previ- ously heated several times at a temperature of 110° C. to approx- imately constant weights. The weighing bottles with the several amounts of fibrin were then heated for a number of hours at 110° C., weighed, and reheated until fairly constant weights re- sulted. Final weighing gave the following results: 5 drops thrombin . . 0.2046 gm. fibrin. to drops thrombin . . 0.3573 gm fibrin. 20 drops thrombin . . 0.6089 gm. fibrin. 40 drops thrombin . . 1.5872 gm. fibrin. It will be seen, from these results, that while they fall short of exact mathematical proportionality, yet they reveal very clearly that there is in thé formation of the fibrin a definite and direct quantitative proportion. The bearing of such facts on the supposed fermenta- tive nature of thrombin is treated farther on. It need not be pointed out that such quantitative relationships will appear only when the amount of thrombin is below the maximum amount required to affect the entire fibrinogen. Adding thrombin beyond this maxi- mum amount can of course be of no moment, there being no further fibrinogen available. Effects of temperature on coagulation with thrombin.— Similar amounts of fibrinogen and thrombin submitted to different tem- peratures showed that temperatures between 17° C. and 41° C. had practically no varying influence on the time of coagulation. In one experiment on this point the following results were obtained: Temperature. Time of clotting. Tee an yw Nea. Sex} (OS, 2 hours yg Oa ee eae 28 min. Dip Oe RE te, tettee es TSE 28 min. CS aie Se el ae 27 min The Coagulation of Blood. AI3 These results agree very closely with the determinations made by Mellanby on this point. The temperature coefficient of coagulation between the temperatures of 17° C. and 41° C. is practically 1, Whether this is to be interpreted to mean that the process of coagu- lation is essentially a physical or physico-chemical one, or at least is overshadowed by physical moments, is perhaps an open question. It is well to recall, however, that in the experiments here noted the thrombin was ready formed. The time element in the coagulation, as determined by a coagulometer, on freshly drawn blood is a dif- ferent matter. In this case there is a definite relation between the temperature and the rapidity of the coagulation, as indicated in a few figures taken from Addis: ® Temperature. Coagulation time. 2c min. sec. MORNIN MS: oa sew Lyte s 2 OMA Sg he Ree a! I a) on EY ORO Bese GM hoe ak ss ick AP ta a? SD 16.5 Pree Eee hh) Sines ee OPE TO Ls aa a aS 2 oy 7934 19.5 Se stares tei erie (oe DONG 2th aoe its Ryyeem eh Ma 12 boo It is obvious that in these latter experiments additional factors are concerned, which give rise to the thrombin, — either processes of secretion, or disintegration of formed. elements, or chemical activa- tion, which drop out when ready-formed thrombin is used. PROPERTIES OF THROMBIN. 1. Its stability with reference to heat. In Alexander Schmidt’s reports of the fibrin-ferment discovered by him, as given in his book ‘‘ Zur Blutlehre,”’ he cites, as one proof of the fermentative nature of the substance, that it is completely destroyed by boiling. A footnote, however, shows that he had overstated this point, inas- much as he calls attention to the fact that “ ordinarily one finds that boiling does not completely destroy the ferment.” Howell also notes 5 Appis: Quarterly journal of experimental physiology, 1908, i, p. 305. ADDIS: Quarterly journal of medicine, 1909, ii, p. 149. AIA | iL evernger: the same fact. In the experiments made upon this point in the present investigation it was found that in aqueous solutions fairly free from proteins the thrombin is not wholly destroyed, and the freer the solution is from proteins, which coagulate on heating, the greater is the amount of thrombin which can bear boiling. Ordi- nary fresh thrombin solutions, after several minutes of boiling, still cause typical clotting. It is therefore not so thermolabile as often asserted, and its reaction to heat can hardly be cited as proof of its essential similarity to other enzymes. In dried form it withstands a temperature of at least 135° C. for half an hour. In regaining the thrombin from such dried extracts it is necessary to submit it for several days to the dissolving action of water. The necessity for this no doubt lies in the excessively hard baking to which the mass has been exposed, and the difficulty with which subsequent extraction takes place. The extraction may be facilitated by allowing putrefactive changes to arise, which probably, in a manner to be described later, liberate the thrombin enclosed. The thrombin is, however, completely destroyed when charred or ashed. Extracts after such treatment show no trace of its action. Moreover, thrombin is only slightly diffusible. Placed in a collo- dion tube, it may be dialyzed against pure water for six or seven hours without serious loss. The conclusion is warranted therefore that thrombin is not a simple inorganic compound. Recovery of inactivated thrombin. — While aqueous solutions of thrombin withstand, in large part, the boiling temperature, solutions of thrombin containing proteins are at first completely inactivated by that temperature. That the thrombin has, however, not been really destroyed is shown by the fact that its action may be recov- ered by simple processes. This recovery may be accomplished by submitting the solution of inactivated thrombin to the action of a small amount of sodium hydroxide. Not quite so effective is treat- ment with a dilute mineral acid for a few minutes. In this way amounts of thrombin may be regained depending upon the maxi- mum temperature used. Instead of the action of the sodium hy- droxide or acid, similar results are obtained by submitting the in- activated thrombin solution to putrefactive changes. Such a protein- containing solution of thrombin we have in the serum of the blood. Heating serum to 56° or 57° C. is usually stated to be sufficient to destroy the thrombin alee and to insure a thrombin-free serum. This was found not to be strictly true. Active serum may be heated The Coagulation of Blood. 415 above 57° C. and its thrombin action regained in part at least. If serum is heated to 65° C. for several minutes, it is found to be in- active when added to a fibrinogen solution. If however such an inactivated serum is treated with a small amount of sodium hy- droxide for a few minutes and then neutralized, it will coagulate fibrinogen, whereas a control of heated serum not so activated has no coagulating effect on solutions of fibrinogen. It is an old observation that serum soon loses its active thrombin on standing, and becomes ineffective in producing a clot. According to older observers the thrombin was supposed to disappear entirely. When later such inactive serum was activated with sodium hy- droxide or mineral acids, and a thrombin activity as much as thirty times the original strength of fresh serum was secured, it was ex- plained as being due to a proferment in the blood which had been activated and from which new thrombin had been formed. Such was the 8 proferment of Morawitz, in distinction from the a profer- ment which is activated by calcium alone. Later the disappearance of the thrombin from serum was explained as due to the formation of a metathrombin, which may be changed back to thrombin by the alkali or acid activation referred to. The parallelism between the gradual disappearance of thrombin from serum on standing and its recovery by activation, and the dis- appearance of thrombin from serum on cheating to 65° and its re- covery by a similar activation, strongly points to the conclusion that these two things are essentially the same phenomenon, and that the disappearance of the thrombin is probably due to its combining in some loose way with some substance, possibly the proteins of the serum, from which it may in turn be liberated by agencies which tend to destroy this combination. That the disappearance of the thrombin is due to such a loose combination with some substance, and that activation by alkalies is due to the breaking apart of this connection, seems probable from the further fact that serum, which has become completely inactive on standing, becomes later, if sub- mitted to putrefactive processes, quite active again. To procure such an active condition by putrefaction often requires a number of days, on account of the remarkable resistance shown by serum to putrefactive changes. Whether this union is of a chemical nature or not, is not easily told. It seems, however, clearly more than mere mechanical adsorption, for neither in the standing serum nor in the serum inactivated by heating to 56° C. is there any precipitation to 416 Te 2 eet pier make mere mechanical adsorption possible. The process is evidently here more nearly allied to a loose but truly chemical combination. Thrombin solutions containing proteins, when heated to points above the coagulation points of these proteins, may no doubt carry down the thrombin in a more mechanically adsorbed way. Experiments make it certain that thrombin does not unite with all proteins. If it is a protein with which it unites, then there is a marked specificity in this respect. If, for instance, serum be allowed to stand, not only will the thrombin originally in it disappear, but it will also inactivate large amounts of thrombin added to it. In some experiments serum was able to neutralize its own volume of a thrombin solution whose activity as tested on fibrinogen was greater than that of fresh serum. If however thrombin is added to a solu- tion of egg-albumen and the mixture is allowed to stand for several days, while its coagulating power is tested from time to time on fibrinogen solutions, it is found that practically none of the thrombin has disappeared, whereas blood serum in a similar experiment loses its coagulating power, unless it is re-activated or is submitted to slight putrefactive changes. If serum be heated to 80° C. before the thrombin is added, and the mixture is then allowed to stand, it is found, on testing with fibrinogen solutions, that the thrombin does not completely disappear. Clotting is evident, although not so immediate and complete, if the mixture has stood for several days, as with fresh thrombin solutions. The partial disappearance of the thrombin in this case may be due to the fact that the preliminary heating has only partly destroyed or altered the substance which combines with the thrombin. If this substance consists of the pro- teins of the blood, we can understand why the serum after being heated to 80° C. still possesses some power to inactivate the throm- bin, since it is well known that not all of the protein is coagulated at this temperature. The view that the union of the thrombin takes place with some blood protein is strengthened further by experiments which showed that thrombin is not neutralized by other known constituents of the blood. Attention was especially directed to cholesterin. Cholesterin solutions, both aqueous and soapy, had however no effect on throm- bin solutions after days of standing. It has been pointed out that the thrombin and fibrinogen unite in quantitatively proportional amounts in the formation of fibrin, and that coagulation in this respect is the mutual precipitation of these The Coagulation of Blood. AI7 two colloids as suggested by Nolf. But the union seems a relatively loose one. If less thrombin is added to an amount of fibrinogen than is required to precipitate the entire fibrinogen, a partial clot will form. If this clot be removed, no thrombin remains in the fibrinogen solution, and it will remain fluid unless new thrombin is added. But the fibrin removed yields relatively large amounts of thrombin, when “digested” according to the method of Gamgee, in a 5 per cent solution of sodium chloride at 40° C. for several hours. In this “solution of the fibrin” the thrombin is liberated anew. The same result may be reached by submitting the fibrin to" putrefactive processes. As the fibrin dissolves, tests with fibrinogen solutions indicate a larger and larger amount of thrombin liberated. The question naturally arises whether, if it is possible to regain the thrombin as thrombin from fibrin, it may not be possible by proper treatment to regain the fibrinogen as fibrinogen. ISOLATION OF THROMBIN. In approaching the problem of isolating the thrombin, experi- ments were made to determine the media which dissolve thrombin and those out of which it is precipitated. It is evidently not dis- solved by alcohol. In Schmidt’s method for obtaining thrombin it is the dried precipitate after the removal of the alcohol that yields the thrombin. That practically no thrombin is dissolved in the supernatant alcohol is shown by the fact that when the alcohol is slowly evaporated and the residue is extracted with water, no per- ceptible traces of thrombin are obtained. It is equally insoluble in ether, in chloroform, and in acetone. The dried alcohol precipitate of serum extracted with these liquids gives no trace of thrombin when the extract is evaporated and re-extracted with water. Aqueous extracts, however, of the dried alcohol pre- cipitate, after having been submitted to the ether, chloroform, or acetone for hours, yield the usual amount of thrombin, showing that none has been dissolved out. But thrombin is very soluble in aqueous solutions. The dried material which had been precipitated from serum by excess of alcohol, when shaken with twice its volume of distilled water for fifteen seconds only, showed a considerable thrombin activity. Attempts to extract the thrombin with water soon showed that 418 Te J etpeenr, not only the free salts of the dried material, but some of the protein as well, went into solution, especially so if the extraction was con- tinued a few hours. The salts are readily removed by dialysis against distilled water, and it was hoped the thrombin could be sep- arated from the excess of proteins by reprecipitating the proteins with alcohol and extracting anew with water. In this manner it was found that the proteins were thrown out in large part in three re- precipitations, but the thrombin activity was greatly reduced, in fact practically destroyed. Whether this is due to the fact that the “thrombin was coagulated along with the proteins is not certain. It is possible that the toxic action of alcohol is more marked in purer thrombin solutions, while in solutions containing considerable amounts of blood proteins it is protected in some unkown way. The method of reprecipitation gave in all cases finally negative results. To reduce the amount of proteins to a minimum, the dried alcohol precipitate of the serum was extracted with water for a period not exceeding one minute. In this short time, as stated, considerable amounts of thrombin were extracted, as shown in tests with fibrino- gen solutions. Such solutions of thrombin showed only the faintest traces of protein when treated with the ordinary color tests, and they remained clear on boiling. When evaporated only a small amorphous residue remained. That thrombin is, however, not a nucleo-protein, as believed by Pekelharing, is indicated by the fact that a thrombin solution may be acidulated with acetic acid, and the nucleo-proteins precipitated. If this precipitate is removed by centrifuging or filtering, and the filtrate at once neutralized, the thrombin activity is not materially reduced. The neutralization may not be postponed too long, as acids and alkalies, if allowed to act sufficiently long, are destructive in their effects. While this experiment shows that we are not dealing with an ordinary nucleoprotein, it is more difficult to arrive at conclusions concerning the protein nature of thrombin. There are some facts which argue against its protein structure. As pointed out by Schmidt and others, it is possible to secure thrombin solutions which are quite active and which give only the faintest traces of protein. As stated in an earlier part of this paper, it stands boiling, and so differs at least from the ordinary proteins associated with it in the blood. More significant, however, is the resistance of thrombin to putre- The Coagulation of Blood. 419 factive changes. Bottles of thrombin containing large amounts of dissolved proteins were allowed to stand for months, until the putre- factive changes had run their normal course. Such thrombin solu- tions lost none of their activity, in fact they gained materially in the initial stages of the process. It should be pointed out, however, that such solutions, even after several months of standing, still show very distinct protein reactions, so that it would be entirely wrong to hold that the cleavage due to putrefaction had been complete. On the other hand, some of the properties of thrombin are similar to those of the proteins. It is not readily diffusible. Dialyzed against pure water, it may, as stated before, be separated from . salts present without serious loss. It is, however, slightly diffusible through a collodion membrane, and when allowed to dialyze against a smaller amount of water, it may be shown that distinct traces of thrombin appear on the outside after several hours. Like proteins, it seems to be broken down by peptic and tryptic digestion. Solutions of thrombin submitted to ordinary peptic (acid) or tryptic (neutral) digestion always lost their coagulating activity. But other possibilities than those of regular peptic or tryptic digestion exist. In the case of the pepsin-hydrochloric ex- periment it may be that the prolonged action of the acid is toxic. Control experiments seemed to show that this was true. In the case of the tryptic digestion there was the disturbing action of the tryp- sin itself upon the fibrinogen of the test solution. Relatively small amounts of trypsin (neutral) in fibrinogen solutions induce active lytic effects, before coagulation is effected. In some experiments the mixture of thrombin solution and trypsin after having been left at room temperature for eighteen hours was first heated to 4o° C. for two hours, and subsequently in order to destroy the trypsin was heated to 65° C. Then by careful activation an attempt was made to restore the thrombin. But all such attempts led to negative re- sults, and it seems probable that thrombin is destroyed by the digestive action of both pepsin and trypsin. While it is impossible therefore to speak definitively as to whether thrombin is a protein or not, it is clear that it is not a simple inorganic compound. Some of the reactions of thrombin suggest its possible relationship with histones or protamines. Attempts have been made to test experi- mentally whether relatively pure thrombin solutions give the char- acteristic histone and protamine reactions, especially the precipitation out of aqueous solution by ammonia. The work on this point has not proceeded far enough to warrant any definite conclusions. 420 Ls JoReti ser: SoURCE OF THROMBIN. Thrombin exists in serum, but what its antecedents are in the cir- culating blood is a question that has been variously answered by different investigators. It has been assumed by many that there exists in the circulating blood a proferment or thrombogen or hepato-thrombin, different names by different authors for essentially the same thing. This pro-thrombin Nolf believes is formed in the liver ; according to Morawitz it is derived from the formed elements of the blood. Activated by calcium salts, and by a kinase derived from leucocytes or tissues, the pro-stage becomes the active throm- bin. By others the existence of a definite thrombogen or profer- ment in normal blood is denied. A series of experiments was under- taken to determine the existence or non-existence of such a body. It was necessary in these experiments to treat the blood in such a manner, at the very moment of taking, as to prevent the usual changes which occur at that instant and which usher in coagulation. The procedure consisted in laying bare the carotid artery of the cat and inserting a cannula. The cannula and short tube were thor- oughly oiled with a neutral sterile oil. On opening the artery the first few cubic centimetres of blood were rejected. About 50 c.c. of the blood were then led into a large jar of 95 per cent alcohol. Vigorous stirring of the alcohol brought the small stream of sub- merged blood into general and instantaneous contact with the greatly excessive amount of alcohol and so insured immediate precipitation and fixation. The remaining amount of blood drawn from the artery (50 c.c.) was led into a beaker and allowed to clot. The serum from this clot was precipitated with twenty times its volume of alcohol. It was thus possible to have two parts of blood from | the same animal, —one part precipitated by alcohol, before any coagulative changes occurred, the other part after the process of coagulation had run its course. The two precipitates were then dried, and aqueous extracts were made in the usual way for prepar- ing thrombin solutions. Such an extract from the serum showed, of course, the usual thrombin activity. Extracts from the blood, taken immediately under alcohol, showed no traces of thrombin. Tested with fibrinogen solutions, no evidence of clotting of any kind appeared. Many similar experiments by Schmidt and others had shown that, with greater and greater care in taking blood under The Coagulation of Blood. 421 alcohol, less and less evidence of ‘‘ ferment ” appeared. It is certain, therefore, that in circulating blood no appreciable amounts of read \- formed thrombin exist. Does a proferment or thrombogen exist? To determine this, if possible, extracts of the blood caught imme- diately under alcohol were activated with dilute sodium hydroxide and dilute mineral acids without giving any traces of thrombin. Activation with calcium salts in varying proportion proved equally negative. No evidence of any kind could be found that there ex- isted in the blood extract any substance which could be activated by calcium salts, alkalies, or acids to real thrombin activity. But, as stated above, those holding to the view that a thrombogen exists in circulating blood assert that the activation is by a kinase in the presence of calcium salts. Extracts of the alcohol blood were therefore made and calcium salts were added in proper proportion. Assuming that a prothrombin was present in the circulating blood, these extracts lacked only the kinase to produce active thrombin. This kinase, according to Morawitz, Nolf, Mellanby, and others, may come from the tissues. Extracts of tissues which had been carefully irrigated with a Ringer’s solution were added to the alcohol-blood extract, and after standing for varying lengths of time the coagulating power of the mixture was tested with fibrino- gen solutions. In no case was any evidence of thrombin action observed, in spite of the fact that all the supposed elements entering into the formation of active thrombin were present. If it be urged that the treatment with alcohol has destroyed the proferment, it need only be recalled that in the preparation of alcohol plasmas by the alcohol-precipitation method, authors assert the existence of the proferment unimpaired.* It seems probable from such results that thrombogen does not exist in circulating blood. In fact, in reading the literature on this point, one is impressed with the fact that the reasons for the existence of a thrombogen in circulating blood rest more upon speculative considerations than upon experimental evi- dence. It would seem that the normal existence of such a body, whose assumed presence in quantity certainly tends to complicate the interpretation of the process of coagulation, should be granted only when definite experimental evidence makes such an assumption necessary. No such necessity exists to-day, and thrombogen as a normal constituent of blood is a theoretical rather than a demon- strable thing. But thrombin does not exist as such in plasma. It must have material antecedents from which it is produced. The 422 1S J ARCT OCR. conclusion seems probable that it has its source under abnormal conditions in greatly accelerated processes which affect the leuco- cytes or platelets of the blood. Whether the substance formed is a result of secretory activity or a product of cell disintegration, it is - impossible to say. But a number of facts point unmistakably to the formed elements of the blood as the source of the pro-stage. Plasma in which the corpuscles have been removed immediately after taking has a greatly increased resistance to coagulation in avian and reptilian blood. In centrifuged blood experiments showed that parts of the oxalated plasma taken from the lower part of the tube near the layer of corpuscles yielded a much more active throm- bin solution than a similar amount taken from the top of the tube, where the plasma was much freer from corpuscles. Substances in actual solution in the plasma could scarcely have shown such an unequal distribution. The absence of thrombin or its pro-stage in the blood taken under alcohol as just described finds a ready expla- nation in the assumption that the corpuscles were in this case co- agulated by the action of the alcohol before the normal processes of disintegration were possible. Whether the substance so formed is of an ordinary protein or histone or protamine character, or 1s a still simpler organic substance, is undetermined, but no matter what its nature may be it is unable to produce the normal formation of fibrin unless activated by calcium salts. It is possible that other as yet undetermined conditions assist in this activation. This activa- tion may find its simplest explanation in assuming that the calcium enters chemically into-the thrombin molecule, as suggested by Mel- lanby, and by Loeb for the coagulins. If therefore the free calcium is thrown out of the blood by chemical agents, it remains fluid be- cause the prothrombin becomes an active coagulating agent only when combined with this minimal amount of calcium, for the role of which strontium alone may be substituted with any success. Sodium-oxalate plasma. — The fluidity of oxalated blood is, as has often been pointed out, due to the precipitation of the calcium. It is not due to a mere inhibition exercised by the oxalate. This was shown by taking oxalated plasma and dialyzing the same in a collo- dion tube against 0.9 per cent solution of sodium chloride for hours. In this way any excess of oxalate is removed, as may be shown by chemical tests. Centrifuging will remove any precipitate, and a fluid plasma is thus obtained which shows no tendency to coagulate when allowed to stand for days. Additions of very slight amounts The Coagulation of Blood. 423 of calcium chloride bring about a speedy clotting, showing that the prothrombin is lacking only in that salt. A point of interest is in the observed fact that such plasma solutions may stand for several days, and yet give immediate clotting upon addition of calcium chloride. The pro-stage of the thrombin does not disappear from the plasma readily. In serum solutions, however, where the acti- vated thrombin is present, this latter disappears very rapidly, as was pointed out in an earlier part of this paper, possibly by combin- ing with some specific proteins in the serum. The thrombin before activation with calcium salts seems not yet to have such combining properties. The fluoride plasma.— The action of fluorides in preventing the coagulation of the blood has been variously interpreted by different authors. Thus, according to Arthus, to whom we are indebted for the introduction of this method, the calcium salts of the blood are precipitated. In this respect it acts like the alkali oxalates. But, unlike the oxalates, Arthus holds that fluoride prevents the forma- tion of a proferment, because calcium salts in excess do not coagu- late fluoride blood. This inhibiting action was believed to be due to a certain fixation of the formed elements of the blood. Bordet and Gengou, noticing the precipitate which is produced when cal- cium salts are added to fluoride plasma, held that the prothrombin together with some fibrinogen was thrown out of the solution and thus caused the plasma to remain fluid. Mellanby attributes his fail- ure to have the fluoride blood clot with excess of calcium to the pre- cipitation of the fibrinogen when calcium salts are added to fluoride plasma. Nolf considers the fluoride blood a definite indicator, not only for thrombin, but for the presence or absence of his throm- bozyme. Fluoride blood does not clot, according to Nolf, because there is no thrombozyme present, its formation having been pre- vented by the action of the fluoride presumably on the leucocytes and platelets. In the following experiments it will be shown that none of the foregoing interpretations is entirely correct, but that the action of the fluoride is primarily identical with the action of the oxalate. If a solution of fibrinogen is made from fluoride plasma by pre- cipitating the fibrinogen several times with a 50 per cent saturated solution of sodium chloride, according to the method described earlier in this paper, and this fibrinogen solution is dialyzed in a col- lodion tube against an excess of 0.9 per cent sodium chloride solu- 424 LE Ie eenrcer: tion, the fibrinogen clots. The clot is firm and typical and the “serum” expressed is peculiarly rich in thrombin. This makes it improbable that the calcium salts are definitely precipitated out of the solution. It is further evident that the prothrombin is present in a fluoride plasma. The prothrombin remains inactivated because the calcium is bound in a peculiar way by the fluoride. This loose connection is broken in the dialysis, the calcium thus freed activates the prothrombin, and active thrombin results. As this occurs in a clear plasma, it is not due to secondary changes of the leucocytes, brought about by the dialysis. It is possible that the calcium and the fluoride are bound to some protein of the plasma in such a way that the calcium activation of the proferment is prevented, and yet loosely enough to be dissociated by the dialysis. If a solution of calcium chloride, of a concentration of I gm. anhydrous calcium chloride to 25 c.c. of water, be added to a fluoride plasma, a copious precipitate appears, which is largely composed of protein material thrown down together with the calcium fluoride. If the calcium chloride solution be added carefully, drop by drop, until the point is reached when the addition of a small drop produces no further precipitate, it will be found that the fluoride plasma so treated clots promptly and normally. Such results show that the statements usually made that calcium salts added in excess to fluo- ride plasma produce no coagulation are not strictly correct. If in the manner described the fluoride is precipitated by the calcium chloride solution and a slight excess of calcium salts remains, the clotting is similar in every way to that produced in oxalate solutions under similar treatment. Which of the proteins of the plasma are thrown down with the calcium fluoride precipitate has not yet been determined, but it is certainly not the fibrinogen which is primarily precipitated. This is evident from the fact that fluoride plasmas still clot, normally after the heavy precipitate which forms on addi- tion of calcium chloride is removed. To test more directly whether fibrinogen suffers any material loss, sodium fluoride was added to a fibrinogen solution to a proportion similar to that in fluoride plasma. Calcium chloride of the concentration described gave, when added, only a slight precipitate, even after standing several hours. Addition of thrombin to the fibrinogen solution so treated with sodium fluoride and calcium chloride clotted in a manner similar to the control of pure fibrinogen. It is probable, therefore, that the protein precipitate in the experiment is not to any considerable ex- The Coagulation of Blood. 425 tent fibrinogen. If however the calcium chloride is in excess beyond a certain point, the coagulation is prevented, and if the excess be considerable, thrombin added to the fluoride plasma remains in- effective. ‘he inhibition is therefore of the formed thrombin itself. That it is the great excess of the calcium chloride which inhibits the thrombin is shown in the further experiment that when sodium fluoride is added to remove the extra excess it brings about a prompt clotting. It would seem therefore clear that the action of the fluoride is essentially the same as that of the oxalate, with the difference that the fluoride is bound in some way with the calcium and protein and remains in solution in ordinary fluoride plasma. The inability of the calcium so bound to activate the proferment present as in ox- alated blood explains the fluidity of the plasma. That the fluoride cannot bind ready-formed thrombin is shown by the fact that such ready-formed thrombin added to fluoride blood produces a normal coagulation. Dialysis breaks the loose combination, liberates the calcium, activation of the proferment follows, and thrombin results. On the other hand, the addition of a solution of calcium chloride of the concentration given to a fluoride plasma precipitates this fluoride-calcium-protein compound, leaving the proferment and the fibrinogen practically unaltered. ia2, 2. see =A On, oy tehs es Ig Lato es ify ea ein The filtrates from the glutaminic acid hydrochloride were united and concentrated under diminished pressure to a thick syrup and esterified by the method of Emil Fischer. The free esters were then liberated, shaken out, and dried over anhydrous sodium sulphate. The aqueous layer was freed from inorganic salts and the esterifica- tion repeated. The total esters thus obtained yielded the following fractions when distilled in the usual way : Fraction. Temperature. Pressure. Weight. I 100° 18.00 mm. 2523 .om: II 70° 0.80 “ 48.88 “ III 100° O.33) 98.77 “ IV 110° o25. 48.76 “ V 190° O28 64.738" — Anatal Wises 2} oy ie! 2 ASS 292.42 gm. The undistilled residue weighed 79.00 gm. Hydrolysis of Ox Muscle. qa Fraction I.— The esters of this fraction were saponified by boil- ing with water for eight and one-half hours, the solution evaporated to dryness under diminished pressure, and proline extracted with boiling absolute alcohol. The part remaining undissolved when subjected to fractional crystallization yielded a fraction containing the more soluble amino-acids from which 0.91 gm. leucine and 2.30 gm. of a mixture of leucine and valine were obtained. The leucine had the following composition : Carbon and hydrogen, 0.1799 gm. subst., gave 0.3632 gm. CO, and 0.1561 gm. HO. Calculated for C,H,,NO, = C 54.92; H 9.99 per cent. Found ye oe = @ 55-06; H 9.71 ce “ In order to separate the mixture of leucine and valine, use was made of the method of Van Slyke and Levene’ by which the in- soluble lead salt of leucine is precipitated from an ammoniacal solu- tion under definite conditions. There were thus obtained, after re- generating the free acids, 1.04 gm. of leucine and 1.01 gm. of valine. The rest of this fraction was then esterified, and 1.8 gm. of glycocoll ester hydrochloride, equivalent to 0.98 gm. of free glyco- coll, isolated. The glycocoll ester hydrochloride crystallized from alcohol in prisms which melted sharply at 144°. Chlorine, 0.1114 gm. subst., gave 0.1138 gm. AgCl. Calculated for C,H,,O,NCI = Cl 25.41 per cent. OUNCE Seno: a ta as —_Ch2520, °F After removing the hydrochloric acid from the filtrates from the glycocoll ester hydrochloride, 6.65 gm. of alanine were obtained, which crystallized in the characteristic prisms and gave the follow- ing results on analysis: Carbon and hydrogen, 0.1351 gm. subst., gave 0.2016 gm. CO, and 0.0921 gm. H,O. Calculated for C,H,O,N = C 40.42; H 7.92 per cent. MOG. Saybia, =a 4ov7o; El 7.63) °° | Fraction II. — After saponifying the esters of this fraction with boiling water as usual and evaporating the solution to dryness under 7 Van SLYKE and LEVENE: Proceedings of the Society for Biology and Experi- mental Medicine, 1909, vi, p. 54. 442 Thomas B. Osborne and D. Breese Jones. diminished pressure, the proline was extracted with boiling absolute alcohol. From the residue which was insoluble in alcohol, there were isolated 14:30 gm. of leucine, 0.91 gm. of valine, and 10.38 gm. of alanine. The alanine crystallized from water in hard dense prisms. Carbon and hydrogen, 0.1571 gm. subst., gave 0.2344 gm. CO, and 0.1088 gm. HO: Calculated for C;H,0,N = C 40.42; H 7.92 per cent. Hound2 eee =C 40.69 ; H 7.75 “ ‘ A fraction was also obtained consisting of a mixture of leucine and valine from which 0.61 gm. of leucine and 1.8 gm. of valine were isolated by the lead method. Analysis showed the leucine to have the following composition: Carbon and hydrogen, 0.1696 gm. subst., gave 0.3404 gm. CO, and 0.1487 gm. H,0O. Calculated for C,H,,NO, = C 54.92; H 9.99 per cent. WOUNG as. see) a = Co54-745eo:01, a The valine gave the following results when analyzed: Carbon and hydrogen, 0.1230 gm. subst., gave 0.2311 gm. CO, and 0.1040 gm. H,0. Calculated for C,H,,O,.N = C 51.24; H 9.47 per cent. ’ Found =, 2). Sacae = VG (51-24; cE ole: The different fractions of valine were united and racemized by heating with baryta in an autoclave for twenty-seven hours at 175°- 180°. The baryta was removed quantitatively from the solution, and the valine converted into the phenyl-hydantoic acid derivative, which crystallized from water in hexagonal plates and decomposed at about 160°. Carbon and hydrogen, 0.1380 gm. subst., gave 0.3095 gm. CO, and 0.0846 gm. H,O. Calculated for C,,H,,O,N, = C 60.98; H 6.83 per cent. Pound: , {ae acontes 42 = {C6107 3 6:80: eee Fraction III. — The amino-acids of this fraction consisted almost entirely of leucine and proline. After saponifying the esters and evaporating the solution to dryness, the proline was extracted with Hydrolysis of Ox Muscle. 443 boiling absolute alcohol. There were isolated from the residue in- soluble in alcohol 36.41 gm. of leucine. Carbon and hydrogen, 0.1486 gm. subst., gave 0.2788 gm. CO, and 0.1193 gm. HO: Calculated for C,H,,NO, = C 54.92; H 9.99 per cent. Bomnd sates oo. 5 =Ci 54-6650 EO 5a. The filtrate from the leucine was boiled with copper hydroxide, and 2.57 gm. of copper aspartate were separated. Copper, 0.1069 gm. subst., air dried, gave 0.0311 gm. CuO. Calculated for C,H;0,NCu 44 H,O = Cu 23.07 per cent. PV lees ae gia y=) = = aie == Cio794 tis ts The united alcoholic extracts containing the proline from Frac- tions I, II, and III were evaporated to dryness under diminished pressure, and the residue redissolved in absolute alcohol. After standing for several days an insignificant amount of substance separated, which was filtered off and the filtrate evaporated to dry- ness and the proline converted into its copper salt. No appreciable amount of racemic-proline copper could be separated. The l-pro- line, after drying to constant weight at 110°, weighed 33.73 gm., equivalent to 26.61 gm. of free proline. For identification the copper salt was decomposed with hydrogen sulphide, and the free proline converted into the phenyl-hydantoine derivative. It crystallized from water in beautiful prisms which melted sharply at 144°. Carbon and hydrogen, 0.1840 gm. subst., gave 0.4473 gm. CO, and 0.0929 gm. HO. Calculated for C,,H,,.N,O, = C 66.64; H 5.59 per cent. BOUNCE Cots | a) (he xen = = GxOOssOe bag i05) «le Fraction IV. — After extracting the phenylalanine ester from this fraction by shaking out with ether and saponifying the ether soluble esters with strong hydrochloric acid, 3.41 gm. of phenylalanine hy- drochloride were obtained. The aqueous layer which remained after the extraction with ether was saponified by boiling with baryta, and 7.43 gm. of aspartic acid were isolated in the form of a barium salt. Analysis of the free aspartic acid obtained by decomposing the barium salt with sul- phuric acid showed it to have the following composition : 444 Thomas B. Osborne and D. Breese Jones. Carbon and hydrogen, 0.2327 gm. subst., gave 0.3107 gm. CO, and 0.1093 gm. EO: Calculated for C,H,O,N = C 36.07; H 5.30 per cent. Ideal 4 6 eh SA =C 36.41; H 5.26 66 ‘“ The filtrate from the barium aspartate, when freed from barium, was boiled with copper hydroxide, and 5.87 gm. of copper aspartate were obtained. Copper, 0.1363 gm. subst., air-dried, gave 0.0394 gm. CuO. Calculated for C,H;0,NCu 44 H,O = Cu 23.07 per cent. POUnG: ioe Se cee eee ay Oiler itey ey No serine or other definite product was obtained from the solu- tion which remained. Fraction V. — The phenylalanine ester was extracted from this fraction with ether and converted, in the usual way, into the hydro- chloride, of which 12.02 gm. were obtained. By treating the filtrate from the phenylalanine hydrochloride with ammonia 1.75 gm. of free phenylalanine were separated. Carbon and hydrogen, 0.1453 gm. subst., gave 0.3498 gm. CO, and 0.0907 gm. H,O. Calculated for C,H,,O.N = C 65.42; H 6.72 per cent. Pound) 2.5 ween = C<65:00;-1 6:05). yn From the aqueous layer, after saponifying with baryta, 5.07 gm. of aspartic acid, in the form of the barium salt, were isolated. The free acid reddened but did not decompose at about 300°. The barium was removed quantitatively from the filtrate from the barium aspartate, and the solution concentrated and saturated with hydrochloric acid gas. After standing for several days in an ice box 17.41 gm. of glutaminic acid hydrochloride separated, which decomposed at 197°. Chlorine, 0.1833 gm. subst., gave 0.1421 gm. AgCl. Calculated for C,H,,O,NCl = Cl 19.35 per cent. Found) \-> 2a nf 3.2 =O ST uae7) a ees When the hydrochloric acid was removed from the filtrate with silver sulphate and the amino-acids converted into their copper salts, 8.40 gm. of copper aspartate separated, which without being re- crystallized gave the following analysis: Hydrolysis of Ox Muscle. 445 Copper, 0.1555 gm. subst., gave 0.0448 gm. CuO. Calculated for C,H;0,NCu 44 H,O = Cu 23.06 per cent. JROCTLAG GC. ae = CU 24-020ee. ke After removing the copper from the filtrate from the copper aspartate and saturating the concentrated solution with hydrochloric acid gas, 1.21 gm. of glutaminic acid hydrochloride separated. THE RESIDUE AFTER DISTILLATION. The residue remaining after the distillation of the esters, which weighed 79 gm., was dissolved in boiling alcohol. On cooling, 1.15 gm. of crystalline substance separated. The filtrate from this was saponified by boiling with baryta, and after removing the baryta quantitatively with sulphuric acid, 7.13 gm. of glutaminic acid hy- drochloride were isolated by the usual process. The free glutaminic acid decomposed at 203° and had the fol- lowing composition : Carbon and hydrogen, 0.2476 gm. subst., gave 0.3713 gm. CO, and 0.1341 gm. ELO. Calculated for C;H,O,N = C 40.79; H 6.17 per cent. ONG) severest or, «Fe = qoloa: March & April . ... 9.9° = rate of .0357; 29.8° = rate of .1875; = 2.55 March & April . . . . 24.8° = rate of .0812; 29.8° = rate of .1875; = 4.53 FROG. November & December 8.8° = rate of .0167; 23.1° = rate of .1204; = 5.71} Gil 10.6° = rate of .0249; 24.7° = rate of .0927; = 1.94 March & April . . . . 10.6° = rate of .0249; 29.8° = rate of .1510; = 3.16 March & April . . . . 24.7° = rate of .0924; 29.8° = rate of .1510; = 3.20 ICL eo eae re 18.2° = rate of .1642; 27.7° = rate of .1807; = 1.19! NECTURUS. 0.93 } 1.62! November & December 8.7° = rate of .0358; 23.0° = rate of .0474; rate of .0197; 24.8° = rate of .0389; TURTLE. _rate of .0046; 24.8° = rate of .0525; = 7.81! = rate of .0583; 28.7° = rate of .1333; = 2.12 — &, | el NI Ne} ° | ee yon be Ce Ke Lee 1 See discussion in the text. 458 Oscar Riddle. of such reactions by 2 or 3, should of course apply to the digestive process. It is of considerable interest to know whether in living animals this result is attained, within what limits it is attained, and whether the cold-blooded animals vary in this respect. Obviously, in the living, digesting organism the temperature changes may affect not only the action of the ferment present on the hydrolysis of food materials, but such other factors conditioning the digestion rate as the amount and acidity of the secretion, the rate of absorp- tion, etc. The measurements and calculations recorded in Table V represent, I believe, the only attempt thus far made to answer the above-mentioned questions. The tabulated 1! results indicate (1) that within certain not very wide ranges of temperature the rule of van’t Hoff applies to the digestive processes in living cold-blooded vertebrates, the average of eight valid coefficients being 2.62. (2) The range of temperature within which the speed of digestion is doubled with a rise of: 10° is different for the different classes of vertebrates; this range being smallest or most restricted in the two types of Amphibia used. (3) The five coefficients which in the table have been designated with the reference figure I reveal influences of the temperature used, which are quite apart from their effect on the speed of chemical re- actions: those numbers which are greater than 3.00 indicating that the lower temperature of the two temperatures compared exercises a destructive or inhibitive action on the digestive secretions; whereas numbers smaller than 2.00 indicate that the higher temperature of the two temperatures compared likewise inhibits or destroys ferment action. Rate at Tn Rate at T X a Tn being the higher temperature, T the lower temperature, and x the difference between these two. 1 These coefficients are obtained as the quotients resulting from + o. x » q PitteekReLATION OF IONS TO CONTRACTILE PRO= SEsot>s.—1V. THE» INFLUENCE OF VARIOUS ELECTROLYTES IN RESTORING MUSCULAR CON-— Pexotieimry AFTER ITS LOSS IN SOLUTIONS OF SUGAR AND OF MAGNESIUM CHLORIDE. IBYERAL PHS. is Wiel nh [From the Marine Biological Laboratory, Woods Hole.| T is well known that all varieties of muscle suffer gradual loss of irritability when transferred from the normal medium or its equivalent physiological salt solution to pure isotonic solutions of sugar or other indifferent non-electrolyte. The action of these sub- stances is not toxic — dilution of the normal medium with its vol- ume or more of sugar solution does not materially affect the prop- erties of the tissue; the essential condition of the change is the withdrawal of certain electrolytes; when these are restored, irrita- bility returns. Anzsthetics produce a similar temporary and re- versible suspension of function, and the basis of the effect is in all probability the same in both cases. Muscular contractility in the intact organism thus depends on the presence of electrolytes in the external medium; and the researches of J. Loeb and Overton, with others, have shown that sodium salts, usually in association with small quantities of potassium and calcium, have a quite specific and peculiar relation to this property. Other contractile tissues differ from muscle in their relation to the electrolytes of the medium. Many cilia continue their activity in media practically free from salt (fresh-water Protozoa, Vermes, Mollusca); the contractile fibrils or myonemata in the ectosare of many ciliate Protozoa (e. g., Stentor) and the stalk of Vorticella seem also independent of salts in the external medium. In marine animals the cilia show a more evident dependence on the salts and are gradually checked in isotonic sugar solutions (e. g., Arenicola). The above relation of sodium salts to contractility thus appears characteristic of muscle but not of contractile tissues in general. 459 460 Ralph S. Lille. Overton? has made an extensive study of the action of different salts in restoring contractility to frog’s muscle after prolonged im- mersion in isotonic sugar solutions. He finds sodium salts to be the most effective; of the other alkali metals lithium and to a less degree cesium (and ammonium) partly exhibit this power, but not rubidium or potassium; the alkali earth chlorides also show it, but to a still more limited degree. The specific relation of sodium ions to irritability is thus a striking peculiarity of this form of contractile tissue. What is the ground of this specificity? Hoeber ? has answered this question in essentially the following manner: in solutions containing sodium salts the plasma membrane preserves its normal permeability, but not in solutions in which sodium is replaced by other metals. Both Overton and Hoeber refer the toxic action of potassium salts to an alteration of this normal permeability. Assuming with Bernstein? and Brunings + that the normal potential difference between exterior and interior of the resting cell, as shown by the demarcation current, depends on the peculiar permeability of the plasma membrane (free pene- trability to certain cations, but not to anions), Hoeber has investi- gated the influence of ions on the permeability of muscle by study- ing the alterations which the various salts induce in the demarcation current. He finds in general that muscular irritability is restored only by those salts in whose solutions the normal external positivity is found, 7. e., in which, on the membrane theory, the plasma mem- brane shows its peculiar differential permeability in relation to the two classes of ions. The salts affect permeability by their action on the colloids of the plasma membrane; sodium salts affect these colloids in such a manner as to impart that peculiar consistency to which the normal permeability corresponds. The characteristic elec- trical surface polarization, on which the possibility of stimulation depends, is thus restored in favorable solutions of sodium salts. Potassium salts, on the other hand, induce abnormal increase in permeability; of this the local negativity resulting from their action is the sign; the loss of irritability and eventual toxicity are due to ’ Overton: Archiv fiir die gesammte Physiologie, 1902, xcii, p. 346; Ibid., 1904, cv, p. 176. * Horser: Physikalische Chemie der Zelle und der Gewebe, chapters 8, 10. Cf. also, Archiv fiir die gesammte Physiologie, 1907, cxx, p. 492. ° BeRNSTEIN: Archiv fiir die gesammte Physiologie, 1902, xcii, p. 521. * Brtnincs: Ibid., 1903, xcviii, p. 241; 1903, C, p. 367. The Relation of Ions to Contractile Processes. 461 the abolition of the normal permeability. Hence such salts cannot restore irritability. It is essentially because stimulation is associated with a negative variation similar to that resulting from the local action of potassium salts or from other injurious influences — many of which can be shown to increase permeability —that a temporary increase in permeability is assumed to be an essential feature of the stimulation process. Salts thus affect irritability by influencing the condition of the plasma membrane and so determining the readiness with which the permeability change of stimulation is effected. The all- importance of specific external.electrolytes to such a tissue as muscle is therefore evident, since its normal physiological action depends on the preservation of a condition of permeability which only sodium salts — typically in association with smaller quantities of calcium and potassium salts — can give. In a recent paper ® I have attempted to show that this temporary increase in permeability — which Overton, Bernstein, Brtinings, and Hoeber agree in regarding as a fundamental feature of the process —is in itself sufficient to account for the increased liberation of energy which is the essential consequence of stimulation. It is assumed that increased permeability of the plasma membrane will facilitate the escape of metabolic products from the cell; and since the increased energy production following stimulation is associated with an increased loss of carbon dioxide,® the hypothesis is sug- gested that it is the increased rate of escape of this oxidation product from the tissue that determines the increased rate of oxida- tion and so of energy production in the tissue at this time. Rapid removal of the reaction products from a chemical reaction promotes, and their accumulation checks, the progress of the reaction; these corollaries of the law of mass action must apply to the energy- yielding oxidative reactions in living cells. The rate of oxidation and so of energy production in muscle must then have a functional dependence on the rate of escape of the final oxidation products — mainly carbon dioxide; this rate must be profoundly influenced if not altogether determined ‘by the degree of permeability of the plasma membrane. Increased permeability thus means increased evolution of carbon dioxide, hence increased oxidation and energy > Litiik, R.: This journal, 1909, xxiv, p. 14. 8 Not yet satisfactorily demonstrated for nerve, where, however, the energy pro- duction is admittedly very slight. 462 Ralph S. Lillie. production within the tissue; this, on the present hypothesis, is the condition during stimulation. Conversely, decreased surface per- meability means decreased loss of carbon dioxide and hence retarded oxidation and energy production,—in a word, inhibition. The plasma membrane thus constitutes, by virtue of its varying per- meability, perhaps the chief means of regulating the velocity of oxidative and no doubt other metabolic processes within the cell. The exact conditions of the high velocity of these processes in the living cell are not understood; apparently favorable combinations of catalyzers and co-ferments exist, and the structural conditions in the tissue favor their action; under these conditions the progress of reactions to equilibria will be rapid; and disturbances of equilibria, as by the above changes in permeability, will have correspondingly marked and rapid effects. The promptitude and the energy of the response to stimulation may conceivably thus be explained. Just why reaction velocities are so high in living cells constitutes of course a separate problem. It will probably be agreed that increased energy production is the essential and primary change following stimulation. The ques- tion as to the exact means by which the transformed chemical energy is converted into the mechanical energy of contraction is a distinct problem for which various solutions have been suggested. The evidence, in my opinion, favors some form of the general view expressed by various physiologists (d’Arsonval, Bernstein, Imbert, J. Loeb, and others) that the energy of muscular contraction is transformed surface energy. Perhaps the most distinctive feature of colloidal systems is their large. potential surface energy,’ and presumably a portion of this appears as free mechanical energy in contraction. Increase in the surface tension of the colloidal par- ticles forming the fibrille would involve increased coherence of the contiguous particles in the fibril and consequently shortening. Increase of surface tension is a change such as would lead in a simple colloidal solution to coalescence of particles and eventually to coagulation; such a change may therefore be called coagulative, and it is possible that a reversible coagulative change of this kind, under the influence of hydrogen ions freed in oxidation, may be the immediate condition of the contraction. Bernstein’s proof that 7 Cf. WoLFGANG OsTWALD’s article in OPPENHEIMER’S Handbuch der Biochemie, Bd. i, p. 8309. ° Lirue, R.: Loc. cit.; also This journal, 1908, xxii, p. 75. The Relation of Ions to Contractile Processes. 463 the temperature coefficient of contractile energy, with a stimulus of given intensity, is negative, supports strongly the general view that muscular energy is transformed surface energy;® and the only sur- face of area sufficient to account for the quantity of energy trans- formed is the united surface of the colloidal particles forming the contractile elements. Alterations in the permeability of the plasma membrane involvy- ing stimulation or the reverse may be variously induced. In the preceding paper I have considered the stimulating and inhibiting action of electrolytes and fat solvents from this point of view. The present paper describes the results of a study of the action of dif- ferent electrolytes and combinations of electrolytes in restoring contractility to the musculature of Arenicola larve after its removal by the action (1) of sugar solutions and (2) of various electrolyte solutions, chiefly magnesium chloride. EXPERIMENTAL. I. RESTORATION OF CONTRACTION AFTER TREATMENT WITH DEXTROSE SOLUTIONS. In pure solutions of non-electrolytes muscular contractions gradu- ally disappear; ciliary movement is also retarded and eventually ceases, but may continue for some hours before its final disappear- ance. The following is typical of the action of isotonic dextrose solutions : June 8, 1908. — Arenicola larve were collected in watch glasses by heliotro- pism in the usual manner and placed in m-dextrose solution (Kahlbaum’s dextrose in flat solid cakes) at 10.35 A. M.; after the larve had settled the solution was changed to remove all trace of sea water. Observation at five-minute intervals showed the following: 10.40. A.M. Larvae swim slowly and collect in clumps; cilia are much slower than normal; larvae are becoming stiff, but still show sluggish mus- cular contractions. 10.45 A.M. Cilia continue slowly; larve are rigid, but show some slight muscular contraction; when the larve are distributed uniformly through the solution by a pipette, the ciliary action causes them to col- lect at the bottom of the watch glass in groups or clumps. The larval body shows slight shrinkage from the cuticle. ® BERNSTEIN: Archiv fiir die gesammte Physiologie, 1908, cxxii, p. 129. 404 Ralph S. Lillie. 10.50 A.M. Slow ciliary movement; larve are quite rigid, no muscu- lar contractions are seen. There is now distinct shrinkage from the cuticle. 11.00 A.M. Condition as at 10.50. Slow ciliary movement continues; no muscular contractions are seen. : In dextrose solutions ciliary movement may continue for two to three hours, but muscular contractions are always found to have completely or almost completely ceased after twenty to thirty minutes. Some variation has been found in the action of different solutions of dextrose; Kahlbaum’s dextrose is not quite free from chlorides, and a part of the variation may be due to this circum- stance. Thus, while the above record is typical, I have found in a number of experiments occasional muscular contractions after an hour or even more in the dextrose solution. Such movements are slight and close observation is required to detect them. A number of careful comparisons were made between dextrose solutions and solutions of Kahlbaum’s crystallized cane sugar, which appears absolutely free from electrolytes. Thus in three separate deter- minations with m-cane-sugar solutions all ciliary movement was found to have ceased in a little over an hour (a trace of movement remained in one experiment after one hour and seventeen minutes) ; in dextrose solutions some ciliary movement nearly always remains after two hours. Muscular contractions disappear in cane-sugar solutions as in dextrose; in the above three experiments careful search showed feeble contractions in a few larve after thirty, forty- five, and fifteen?° minutes respectively. Solutions of these two non-electrolytes have thus closely similar action; but in dextrose solutions complete disappearance of contractions usually requires somewhat longer. This difference may be due to the slight electro- lyte content of the dextrose solutions. In the following experiments with salts I have used dextrose solutions to deprive the muscles of contractility; these are more satisfactory than cane-sugar solutions on account of their relatively slight viscosity and low specific gray- ity, — which is less than that of the larve, while with cane-sugar the contrary is the case. Action of pure solutions of various sodium salts. —In the experi- ments about to be described I have studied the action of a series of salts in restoring contractility to larvee after its removal by dextrose solutions. Kahlbaum’s salts were used in practically all cases. No observation between fifteen minutes and forty-three minutes in this case. After forty-three minutes prolonged observation showed no trace of contraction. * og dana Ste The Relation of Ions to Contractile Processes. 465 Muscular contractions, having at first an almost normal character, invariably return after transfer from sugar solutions to pure solu- tions of sodium salts. The injurious action of the pure salt solu- tion quickly becomes evident, and contractions, at first energetic, become weaker and soon cease, within an interval which varies considerably with the different salts according to the nature of the anion. Pure solutions of sodium salts are also highly injurious to cilia, most of which, as a rule, are arrested and in large part disin- tegrated within the first few seconds after transfer; a few may remain active for some time longer; this effect also shows signifi- cant variations according to the nature of the anion, and is more rapid (e. g.) with bromide or iodide than with acetate, sulphate, or tartrate. The following record will illustrate: ZEAB ICE. I June 11, 1908. — Larve were transferred from sea water to m-dextrose solution at 2.35 P.M.; this solution was changed for fresh at 2.45. The following solutions of sodium salts in m/2 concentrations were added to portions of larva in watch glasses at the times indicated (the action of KCl, NH,Cl, and LiCl was also studied in this series; these salts will be considered later). The immediate effect of each salt was observed under the mi- croscope with about 69 diameters’ magnification. The intensity of the immediate stimu- lating action is indicated by the promptitude and vigor of the initial contraction or shortening which immediately follows the contact of the solution; this varies in a highly character- istic manner with the nature of the anion; the character of this initial contraction is there- fore indicated in each case. 1. m/2 NaCl. 3.17. Muscular contractions begin at once with a moderate initial con- traction. Cilia are partly disintegrated at once, but many continue slowly. By 3.20 muscular contractions are very feeble; by 3.25 none are seen and most cilia have ceased. At 3.49 there is no muscular movement and cilia have practically ceased. 2. m/2 NaBr. 3.26. Well-marked initial contraction; arrest and disintegration of cilia are more complete than with NaCl; a few cilia remain feebly active. By 3.28 muscu- lar contractions have almost ceased. A trace of ciliary movement remains at 3.47; no movement is seen later. 3. m/2 Nal. 3.30. Marked initial contraction; feeble muscular movements during the succeeding relaxation; cilia nearly all cease and liquefy at once. By 3.32 muscular contractions have practically ceased —a trace of movement is seen. At 3.44 a trace of ciliary movement persists; no muscular movement. No movement seen later. 4. m/2 NaNO,. 3.38. Moderate initial contraction; muscular contractions somewhat slight and cease soon; cilia mostly cease at once. At 3.40 a few contractions at inter- vals; a little ciliary movement. At 3.43 a few contractions. At 3.50 no muscular movement; trace of ciliary. No movement later. 5. m/2 NaClO,. 3.41. Action very similar to m/2 NaNO,; a few contractions at 3.52; none seen later. A faint trace of ciliary movement remains after one hour. 6. m/2 NaCOOCH,. 3.53. Well-marked initial contraction; cilia continue movement in most larve. At 3.58 contractions occur at intervals; cilia slowly active in a large 406 Ralph S. Lilhe. proportion. At 4.31 and 4.44 cilia are still largely active, but no muscular movement is seen. : 7. m/2 Na,SO,. 3.54.5. Initial contraction, followed by well-marked muscular moye- ment; cilia continue in a large proportion. At 4.00 only a few muscular contractions are seen; fair ciliary movement remains. At 4.45 a few feeble contractions are seen; ciliary movement has almost ceased. ; 8. m/2 Na,S,O,. 3.56.5. Immediate effect as in m/2 Na,SO,. At 4.10 occasional con- tractions are seen; most cilia have ceased. At 4.27 a few contractions are seen, and a little ciliary movement. No movement at 4.48. 9. m/2 Na,C,H,O,. 4.02. Slow muscular movements at first. At 4.07 a few contrac- tions occur at intervals; cilia continue in a fair proportion. At 4.24 there is consider- able slow ciliary movement; no contractions are seen. At 4.50 a very few contractions are seen; a little ciliary movement remains. 10. m/2 Na, citrate. 4.03. Muscular contractions are slight and cease soon; cilia con- tinue for a short time. At 4.05 a few occasional contractions are seen; cilia have almost ceased. At 4.22 a trace of ciliary movement remains; no muscular contraction. No movement seen later. 1l. m/2 Na,HPO,. 4.10. Contractions begin somewhat feebly; cilia continue for some little time. At 4.20 occasional contractions are seen; considerable ciliary movement remains. At 4.54 no movement is seen. 12. m/2 Na,Fe(CN),. 4.12. Contractions appear slowly; cilia continue in a fair pro- portion of larvae. At 4.18 a few feeble contractions are seen; cilia have almost ceased at 4.19. No movement seen later. Four other series of experiments with sodium salts gave a similar general result. A certain definite relation of the anions to ciliary movement appears quite uniformly. In solutions of sodium acetate particularly, also in those of sulphate, tartrate, and phosphate, there is less immediate destruction of cilia than in chloride, bromide, iodide, and nitrate, and the average duration of the movement is distinctly greater. If the relative action of the salts is measured by the average maximal duration of ciliary movement in their re- spective solutions, the toxicity of the anions is seen to increase in essentially this order: (COOCH;.< SO,-<* C,AZOp =< BEOw Cl < NO; (and ClO,) < Br (< 33h.) . : 55 m. 17 m 29 m. about 22 h. %) Joy. 50 m.+ 3 m. Stim 2m. about 22 h. 3h 42 m.+ 3°m. 4m 2 m. 29 m. 6m 2 (21 ny (< 17 m.) 4.5 m.-+ 2m. 6 m. (< 6 m.) 4.5 m. none seen 13 m.+ 0 0 none seen 13 m.+ 27 m.+ (< 3$h.) 13 m.+ 12.5 m. <3. 5) 10s U2"5emae (< 24 m.) sugar solution > magnesium chloride solution. The immediate toxic action of a given salt on larve from these media —as shown, for instance, in the destruction of cilia — also shows this order of decrease. Action of pure solutions of sodium salts. — Apart from this differ- ence in the intensity of the immediate stimulating effect, the action of sodium salts is essentially similar to that described above for larvee from sugar solution, and similar differences of action between different salts are seen. The following series of experiments will illustrate : TABLE VII. June 25, 1908. — In each experiment larve were transferred from sea water to m/2 MgCl,, where they remained exactly seven minutes; thence they were transferred to the solution of the experiment. In each case the time of placing in the m/2 MgCl, is indicated in parentheses. 1. m/2 NaCl (10.10 to m/2 MgCl,), 10.17. No muscular contractions are seen. Cilia remain active for some hours (influence of the previous exposure to MgCl,). 2. m/2 NaBr (10.15 to m/2 MgCl,), 10.22. A few slight contractions are seen at first; no contractions are seen at 10.23, 10.26, and 10.36. Cilia remain active. 3. m/2 Nal (10.20 to m/2 MgCl;), 10.27. No contractions are seen at first; within ten to fifteen minutes well-marked contractions appear at intervals in most larve. Cilia remain active after four hours. chloride solutions are described as producing well-marked contractions in larve from m/2 MgCl, a little calcium seems to have been present in the salts used. Kaunt- BAUM’s sodium and magnesium chlorides appear io be practically free from this element. 482 Ralph S. Lillie. 4. m/2 NaNO, (10.25 to m/2 MgCl,), 10.32. Slight contractions appear in a few within two minutes. No contractions are seen later. Cilia remain active. 5. m/2 NaCOOCH, (10.30 to m/2 MgCl,), 10.37. No muscular contractions are seen. Cilia remain active for hours. 6. m/2 NaClO, (10.35 to m/2 MgCl,), 10.42. Almost no contraction results; one or two slight movements are seen at first. Cilia remain active. 7. m/2 Na,SO, (10.40 to m/2 MgCl,), 10.47. A few slight contractions are seen at 10.51. Cilia cease sooner than in the above solutions; have largely ceased within an hour. 8. m/2 Na,C,H,0, (10.45 to m/2 MgCl,), 10.52. Contractions appear in a fair proportion of larvee within two to three minutes. Cilia cease comparatively soon, as in Na,SO,. 9. m/2 Na,HPO, (10.50 to m/2 MgCl,), 10.57. Well-marked contractions appear in a good proportion of larve; effect greater than in any of the preceding solutions. Cilia cease comparatively soon (about one hour). 10. m/2 Na, citrate (10.55 to m/2 MgCl,), 11.02. No contractions are seen. Cilia cease in less than twenty minutes. In a second precisely similar series, with four minutes’ exposure to m/2 MgCl, essentially the same effects were seen, although the differences between the individual salts were somewhat less distinct. Return of contractions in pure solutions of sodium salts is thus, in general, relatively slight as compared with that seen after sugar solutions. The salts show similar though less decided differences in the intensity of their stimulating action. The following is the order of increasing intensity of action for the monovalent anions: COOCH, and Cl < NO,, ClO, and Br <1. Tartrate andisulphate cause slight, and phosphate well-marked, return of contractions; possibly their action as precipitants of calcium is a factor in the effect, — this, however, I have not definitely determined as yet. The differences in toxicity show corresponding relations, as shown by the following experiments: After about four hours in the above solutions a number of larve were transferred from each solution to fresh sea water. There was distinct recovery of contractility in those from acetate and sulphate, and to a less degree from tartrate, but not from the others. In a third series in which larvze, after a variable exposure to magnesium chloride, were brought into m/2 solutions of NaCl, NaBr, NaI, NaNO 3, Na,SO, and Na-citrate, a similar result appeared: on transfer to sea water, after about two and a half hours in these solutions, well-marked and vigorous con- tractions returned in the larve from chloride and sulphate, but from no others. Tartrate, sulphate, acetate, and chloride belong to those anions whose action on colloids, according to Hofmeister and Pauli, is relatively slight; the salts with anions of more pronounced action — Br, NO,, ClO;, I— exceed the others both in immediate stimu- The Relation of [ons to Contractile Processes. 483 lating power and in toxicity. The basis of these characteristic dif- ferences in action has been discussed above (p. 466 seg.) ; in the present experiments the plasma membrane, having been rendered relatively impermeable by the action of the magnesium, apparently does not undergo such prompt and ready alteration of permeability as before, hence the stimulating effect is less marked and the differ- ences between the salts appear less sharply defined. Action of pure solutions of alkali and alkali earth chlorides. — Pure isotonic solutions of the other alkali chlorides, with the exception of potassium (and possibly rubidium, which I have not yet tried), also produce little if any contraction in larve from magnesium chloride. Even potassium chloride causes only slight contractions. Calcium and strontium chlorides, on the other hand, have well- marked action. The following series of experiments will illustrate: TABLE VIII. June “23, 1908. —In this series the larve remained in m/2 MgCl, for five minutes, thence were transferred to the respective solutions. 1. m/2 KCl (10.52 to m/2 MgCl,), 10.57. Slight muscular contractions appear, lasting for about a minute. Cilia continue actively. 2. m/2 NH,Cl (11.01 to m/2 MgCl,), 11.06. No contractions are seen. Cilia have ceased by 11.47. 3. m/2 LiCl (11.09 to m/2 MgCl,), 11.14. No contractions are seen. Cilia remain slowly active at 11.47. 4. m/2 NaCl (11.17 to m/2 MgCl,), 11.22. No contractions are seen. Cilia remain slowly active at 11.50. 5. m/2 CaCl, (11.30 to m/2 MgCl,), 11.35. Well-marked contractions appear at once, which continue slowly at 11.52. Cilia remain active in a fair proportion of larve at 11.52. 6. m/2 SrCl, (11.38 to m/2 MgCl,), 11.43. Vigorous initial shortening results, followed by slight contractions which diminish and cease in about ten minutes. Cilia cease within ten minutes. A second similar series with variable and longer exposure to m/2 MgCl, yielded a practi- cally identical result. The susceptibility to stimulation by pure solutions of alkali chlo- rides has thus been greatly diminished as compared with that seen after sugar solutions. Since the typical strong initial contraction with its associated loss of pigment is absent in these solutions after the treatment with magnesium chloride, it is evident that the latter salt has increased the resistance to the change of permeability which the alkali chlorides normally induce. That this is the case is also shown by the marked decrease in the toxic action of these salts resulting from brief exposure to magnesium chloride solutions (see below, pp. 489-490). 484 Ralph S. Lilhe. On the other hand, the contractions produced by the alkali earth chlorides do not appear to be decreased by previous treatment with magnesium chloride. Calcium chloride in m/2 solution produces littte or no immediate contraction in larve from sea water; 7! but in larve from magnesium chloride solutions well-marked contractions at once appear, as described in Table VIII. Calcium appears thus to antagonize the action of magnesium; the striking effects pro- duced by the addition of a little calcium to solutions of sodium and lithium chlorides (see below, p. 485) are no doubt due partly to this as well as to a simple antitoxic action. Strontium has a some- what similar action to calcium, but is decidedly more toxic. Why calcium should have this antagonistic action to magnesium remains unexplained at present. The physiological contrast between these metals appears to be widespread, especially in relation to muscular contraction (reviving action of calcium on the heart beat, etc.). Action of acid and alkali in low concentrations. —- After treatment with magnesium chloride larve are also much less readily aftected by acid and alkali than after sugar solutions. Thus, in one series of experiments larve were transferred from m/2 MgCl, to a series of m-dextrose solutions containing hydrochloric acid in the con- centrations 2/200, 2/400, n/800, n/1600, /3200, and n/6400; in no solution was any contraction seen.??, A similar negative re- sult appeared in solutions of sodium hydrate (concentrations from n/25 to n/1600). Experiments were also tried with pure m/2 sodium chloride solutions, plus HCl and NaOH from 2/200 to n/6400 as before. In this case the action of the acid is added to that of. the salt (which has a slight action, as already seen), and some effect was seen; slight contractions appeared in acid of n/1600 to n/6400 concentration; alkali, on the contrary, had little or no demonstrable effect, even in the higher concentrations. These results resemble those found after dextrose solutions (cf. p. 476), where also acid produces contractions (more vigorous than after magnesium chloride), while alkali has no such action except in rela- tively high concentrations (1/25 and 1/50). On the present inter- pretation acid in those concentrations would appear to increase, and alkali to decrease, permeability; this inference is confirmed by the contrast in the toxic effects of the two sets of solutions. Thus larve, after one and a half hours in a series of six solutions of *1 Cf. the preceding paper; also below, p. 487. ” Acid in m/2 MgCl, solutions (HCI m/1oo to n/3200) also produces no contractions. The Relation of Ions to Contractile Processes. 485 m/2 NaCl + HCl (between 2/200 and 1/6400, as above), were brought into fresh sea water; no contractions resulted; the animals had an opaque coagulated appearance and were evidently dead. On the contrary, larve from the corresponding alkaline solutions re- covered and showed vigorous contractions in all cases except in n/200 and n/400 NaOH. Alkali and acid in low concentrations thus appear to have opposite influences on the permeability of the plasma membrane in these organisms (see above, p. 479). Action of mixtures of two salts. — Addition of a little potassium chloride to a pure m/2 sodium chloride solution appears somewhat favorable to the production of contractions in larve from mag- nesium chloride solutions. In a series of six solutions in which KCI was added to m/2 NaCl in the proportions m/50, 11/100, 11/200, m/400, m/8o00, and m/1600, slight contractions appeared in every case. These effects, however, were not decidedly different from those found with pure m/2 NaCl in which occasionally slight con- tractions are seen, especially if the previous stay in the m/2 MgCl, has been brief. On the other hand, the effects following the addition of a little calcium chloride to pure sodium chloride solutions are most pro- nounced and highly characteristic. The importance of calcium to muscular contractions, and its power of counteracting the anzs- thetic action of magnesium, are clearly shown by the following experiments : June 27, 1908.— Larve after complete loss of muscular contractility in m/2 MgCl, were transferred to the following solutions: (1) m/2 NaCl, and (2-11) a series of ten solutions of (about) m/2 NaCl containing CaCl, in concentrations ranging from m/s50 (24 volumes m/2 NaCl + 1 volume m/2 CaCl,), m/100, m/200, etc., in regular series to m/51200. In pure m/2 NaCl no contractions were seen; in solutions containing CaCl, in concentrations from m/so to m/8o0o vigorous squirming and bending movements began instantly, and slight contractions remained for three or more hours. In the other solutions similar effects were seen, the contrac- tions growing progressively less pronounced as the calcium content diminished; in m/6400-and m/r12800 CaCl, there was a certain delay in the appearance of the contractions, and these were comparatively slight; while in m/25600 CaCl, only a few contractions were seen, and in m/51200 CaCl, practically none. The favorable influence of the calcium chloride is thus distinctly perceptible in concentrations so low as m/25600. Strontium chloride 486 Ralph S. Lilhe. has a somewhat similar though less favorable action: a series of ex- periments with m/2 NaCl containing SrCl, in concentrations from m/50 to m/3200 showed vigorous contractions in all for some minutes after the transfer from m/2 MgCl,, and slower movements continued for some time (about one hour) afterwards. In similar experiments with sodium and barium chlorides the antitoxic action was obscured by the marked toxicity of the latter salt; in all the solutions — relatively gradually in 7/1600 and m/3200 BaCl, — the larvee slowly underwent permanent shortening and all movement ceased within a few minutes; there were no active bending contrac- tions as with calcium or strontium. Barium thus shows in its physi- ological action a decided contrast to calcium and strontium. Transfer from m/2 MgCl, to mixtures of lithium and calcium chlorides (CaCl, from m/1o to m/160, and from m/50 to m/800) is also followed by well-marked contractions, more siuggish than in the corresponding sodium. chloride solutions and ceasing sooner. No contractions appear in pure m/2 lithium chloride. The opti- mum concentration of calcium appears higher with lithium than with sodium chloride; contractions lasted longest in a mixture of four volumes m/2 LiCl, + one volume m/2 CaCl,, where a few contractions remained after more than an hour. Contractions tend to be sluggish in lithium solutions, and to cease soon, even in the presence of a favorable proportion of calcium; lithium is thus at best an imperfect substitute for sodium. Addition of calcium to pure solutions of potassium chloride does not counteract the toxicity of the latter salt, but the contractions which follow the transfer from magnesium chloride are decidedly more vigorous and last distinctly longer (three to four minutes) than in the pure m/2 KCl. The specific action of the calcium is thus apparent, although even in its presence the potassium quickly de- stroys all muscular contractility. In the experiments thus summar- ized CaCl, was added to m/2 KCI in the concentrations m/50, m/1oo, etc., to m/1600, and, in another series, m/400, m/600, m/800, m/1200, m/1600, and m/2100. III. Mopr or AcTIon oF DIFFERENT PuRE SALT SOLUTIONS. In my preceding paper attention was directed to the contrast be- tween the immediate action of pure solutions of alkali chlorides on The Relation of Ions to Contractile Processes. 487 the one hand, and of magnesium and calcium chlorides on the other. Larve transferred from sea water to the former group of solutions undergo instant shortening with loss of pigment, followed by re- laxation and cessation of movement; while in magnesium and calcium chloride solutions the contractions cease by degrees, with- out muscular shortening and without loss of pigment. On re- transfer to sea water after more or less prolonged stay in the respective solutions contractions are found to return gradually and imperfectly in the case of the former group of salts, and promptly and completely in the latter. This contrast between the two groups of salts is to be referred to the difference in their immediate influence on the permeability of the plasma membrane. Marked and prolonged increase of per- meability involves toxic action, since one of its consequences must be a partial disorganization of the cell through loss of essential constituents. Decreased permeability, on the other hand, simply arrests stimulation and its associated metabolic processes by retard- ing or preventing the escape of metabolic products, especially car- bon dioxide. Arrest of activity in a living cell can thus be induced by two essentially opposite means,?* — one injurious, the other not. This difference of action is further indicated by the following experiments. To confirm the view that the toxic action of the first group of salts is not due primarily to their entrance into the interior of the cell, but to an alteration of surface permeability, the follow- ing experiments were performed. Larvz were transferred to pure m/2 LiCl, KCl, and NaCl (1) directly: from sea water, and (2) from magnesium chloride solution in which they had remained for sev- eral minutes. After a certain stay in the alkali chloride solution they were brought again into sea water. The alkali chlorides, as already seen, have little immediate action on magnesium chloride larve; 1. ¢., on the present view, the permeability remains, for some time at least, practically the same as in the magnesium solution. Such larve should therefore show prompt and complete recovery on return to sea water, in contrast to those which were placed in the alkali chloride solution directly from sea water and which, as already pointed out, undergo immediate marked increase in per- meability and recover contractility slowly and imperfectly on return 3 Cf. HorBer: Physikalische Chemie der Zelle und der Gewebe, 2te Auflage, 1906, pp. 293-294. 488 Ralph S. Lillie. to the normal medium. These expectations were completely real- ized, as the experiments of Table X, p. 490, will illustrate. The following experiments (Table IX) show the differences in promptitude and completeness of recovery between larve exposed Stay in Solution. 1. m/2 LiCl. 54 m. lh; 52m. 4h. 20 m. 2. m/2 NaCl. 54 m. 1h. 50 m. 47h. 17m: 3. m/2 KCl. 53 m. ih: 45>m. 4h. 12 m. 4. m/2 MgCl,. 52 m. lh. 42 m. 4h. 10 m. 5: m/z Ca@),. 52,m. 1h. 40 m. 4h.5 m. 6: m/2 SrCl,. 50 m. 7. m/2 BaCl,. 54 m. 8. m/2MnCl,. 52 m. 3 h. 30 m. TABLE IX Effect of Transfer to Sea Water. No immediate result; larvae remain motionless after forty-five minutes. At two and one-half hours after transfer contrac- tions are well marked. ; No contractions are seen for a long time. At one and one-half hours after transfer larvee show slight contractions at intervals. At one and one-half hours a few slight contractions. At two and one-half hours well-marked contractions at intervals. No immediate effect. At twenty-three minutes a few faint con- tractions are seen. After two and one-half hours contractions are well marked. No contractions are seen for a long time. After one and one-half hours slight contractions occur at infrequent intervals. No contractions are seen at one and one-half hours. Slight movement has returned after about two and one-half hours. No contractions at twenty minutes. At thirty-seven minutes slight jerky contractions are seen in a good proportion. At two and one-half hours contractions are more active than after NaCl. No contractions at twenty minutes. Contractions are well marked after one hour thirty-seven minutes. No contraction at fifty-eight minutes after transfer. At one- and one-half hours a few slight contractions; next day most larve show well-marked contractions, better than after NaCl or LiCl. Vigorous contractions at once. After thirty minutes, movement appears practically normal. Immediate return of vigorous contractions. Immediate contractions. Immediate vigorous contractions. Active contractions return almost immediately. Well-marked contractions return after an interval of a few seconds. No contractions at first or at forty-five minutes after transfer. At two hours and twenty minutes only a few slight infrequent contractions are seen. Practically no return of contractility. No return of contractions. No movements at thirty-three minutes after transfer. At two hours and eleven minutes larve show jerky contractions at intervals; after five hours contractions are vigorous and frequent. No movement is seen for more than two hours. Next day a large proportion of larvee show good contractions. The Relation of Ions to Contractile Processes. 489 for approximately equal periods to solutions of various salts and then returned to sea water. Larvze were brought from sea water into m/2 solutions of the following salts: LiCl, NaCl, KCl, MgCl, CaCl,, SrCl,, BaCl,, and MnCl,. The immediate effect of the alkali chlorides and of strontium and barium chloride is to produce a strong initial contraction with loss of pigment, followed by weaker movements which cease within a few minutes (more rapidly in BaCl, and KCl), while in magnesium and calcium chlorides contrac- tions cease by degrees without the initial contraction or evident in- crease of permeability. Manganese chloride has an immediate action very similar to that of magnesium chloride; there is no initial contraction, and muscular movements cease in a few seconds; ciliary action continues for an hour or more, Larve were trans- ferred from each of these solutions to sea water after varying in- tervals of time with the results indicated in Table [X. The return of muscular contractions in sea water is thus imper- fect and delayed after exposure to lithium, sodium, or potassium chloride, and the more so the longer the stay in the solution. Potas- sium appears the least toxic of these three salts (see below, Table X, p. 490). On the other hand, after several hours in m/2 MgCl, vig- orous contractions return at onces the same is true for m/2 CaCl., although its action is somewhat more toxic. Strontium and barium chloride are both decidedly. toxic, particularly the latter. Manga- nese chloride, however, resembles the second of the above two groups of salts; although more toxic than magnesium, it appears to differ from the latter chiefly in the less ready reversibility of its action. This corresponds to the difference between the reversibility of the precipitating action of alkali earth and of heavy metal cations on colloidal solutions.?# If the larvee have previously been treated with magnesium chlo- ride, the toxic action of the alkali chlorides is much less pronounced and contractions return promptly on transfer to sea water. The following experiments will illustrate (see Table X). The protective influence of previous treatment with magnesium chloride appears clearly from these experiments. By this treatment the plasma membrane seems to have been given a consistency which is afterwards altered only slightly and gradually by the alkali chlo- ride. That the latter salt eventually changes the permeability of the = Cf. Pavtt: Beitrige zur chemischen Physiologie und Pathologie, 1905, vi, Pp. 233- 490 Ralph S. Lilhe. membrane is, however, clearly indicated by the results of another series of experiments; larve treated for twelve minutes with m/2 MgCl, and then left for twelve minutes in m/2 KCl] showed instant recovery in sea water, but after two hours in the m/2 KCl and return TABLE X July 3, 1908. — Larve were placed in m/2 MgCl, at 3.20 p.m. After an interval portions were transferred to m/2 LiCl, NaCl, and KCl, and thence, after a certain interval, to sea water. The effects following transfer to sea water were compared with those found in larve treated for the same length of time with these solutions but without previous ex- posure to m/2 MgCl,. The exposures to m/2 MgCl,, preceding the transfer to m/2 LiCl, NaCl, and KCl, were, respectively, 13.5, 17.5, and 24 minutes. | Action of sea water on larve exposed to m/2 alkali | chloride for period given in first column. | | “Immediate action of salt solution and length of exposure. B.° On larve transferred to solution directly from sea water. A. On larve placed in solution after treatment with m/2 MgCl,. Lien / 20 LAC Z1-5\ mt | In m/2 MgCl, for 13.5 m. | No contractions are seen in sea Marked contractions | in larve from sea| water; none in larve from m/2 MgCl,. Immediate vigorous con- tractions in practically all. Larve are still largely active after 19 h. 2. m/2 NaCl. 24.5 m.|In m/2 MgCl, for 17.5 m. Marked contractions in larvee from sea water; | none after MgCl,. | Sean 2 KCI 924m Marked contraction after sea water, as usual; none seen after m/2 MgCl,. Immediate active and vig- orous contractions. Next day, after 19 h., larvae show vigorous and apparently normal movement. |In m/2 MgCl, for 24 m. Vigorous contractions ap- pear within }m. Next day all larvz are active and ap- parently normal. water for almost an hour, when feeble twitches begin. Next day only a little move-} ment remains — much less than in Experiment 1 A. No contractions appear in sea water for the first 20 m. Slight movements begin in about $ h. Next day (19 h.) most larve are disintegrated and dead; a little intermittent movement remains. Marked contrast to 2 A. No movement appears in sea water for about 10 m. Con- tractions are well marked after +h. Next day all larve are living and active. to sea water there was a delay of over fifteen minutes in the return of contractions. Recovery proved almost complete, nevertheless, while larvee left for about the same time in m/2 KCl, but without previous treatment with m/2 MgCl,, showed only partial and imperfect re- vival of contractions. The potassium chloride during the two hours of its action had evidently produced considerable alteration in the contractile tissues, in spite of the previous decrease of permeability — The Relation of Ions to Contractile Processes. 491 by the magnesium salt. The marked and sudden increase of per- meability in pure alkali chloride solutions, which otherwise results and to which a large part of the toxic action is to be ascribed, is however prevented by previous exposure to magnesium chloride. SUMMARY. I. -Arenicola larve lose muscular contractility gradually in pure isotonic solutions of non-electrolytes (dextrose, cane-sugar), more rapidly in solutions of magnesium salts. The effect resembles anesthesia, and contractility is readily restored on return to sea water or to various electrolyte solutions. 2. Pure solutions of sodium salts produce well-marked contrac- tions in larvee from isotonic sugar solutions, but few or none in those from m/2 MgCl,. The addition of a little calcium chloride to sodium chloride solutions greatly increases the ability of the latter to restore normal contractility; calcium shows perceptible action in dilutions so low as m/25600. 3. In the intensity of their immediate stimulating power sodium salts show a definite relation to the character of the anions; the order of increasing toxicity is, in general, also that of increasing stimulating power. This order is essentially as follows: COOCH, pea elo, <— HPO, 2 Cl< NO, and ClOx;