rite! Pa Tape THE JOURNAL OF BIOLOGICAL CHEMISTRY FOUNDED BY CHRISTIAN A. HERTER AND SUSTAINED IN PART BY THE CHRISTIAN A. HERTER MEMORIAL FUND EDITED BY KIN, New York City. LAFAYETTE B. ME J. J. ABEL, Baltimore, Md. R. H. CHITTENDEN, New Haven, Conn. OTTO FOLIN, Boston, Mass. A. S. LOEVENHART, Madison, Wis. a _ WILLIAM J. GIES, New York. GRAHAM LUSK, New York. os L. J. HENDERSON, Cambridge, Mass. A. B. MACALLUM, Toronto, Canada. 3 REID HUNT, Boston, Mass. J. J. R. MACLEOD, Cleveland, Ohio. ‘ WALTER JONES, Baltimore, Md. JOHN A. MANDEL, New York. a J. H. KASTLE, Lexington, Ky. A. P. MATHEWS, Chicago, Ill. a ‘J, B, LEATHES, Toronto, Canada. F. G. NOVY, Ann Arbor, Mich. “- | } THOMAS B. OSBORNE, New Haven, Conn. Bs i t T. BRAILSFORD ROBERTSON, Berkeley, Cal. * : P. A. SHAFFER, St. Louis, Mo. ; } 4 A. E. TAYLOR, Philadelphia, Pa. F. P. UNDERHILL, New Haven, Conn. | : V. C. VAUGHAN, Ann Arbor, Mich. 4 ALFRED J. WAKEMAN, New Haven, Conn. r HENRY L. WHEELER, New Haven, Conn. 2, }Y Ja \ ‘ 524 VOLUME xvi V*%~{b | BALTIMORE afi Y 1913-14 COMPOSED AND PLINTED AT THE :" ~ WAVERLY PRESS hh / By rae Wietrame & Wiuwins Courant Bacrimonn, U.8, A, x CONTENTS OF VOLUME XVI. GrorGe Peirce: The partial purification of the esterase in pig's liver GerorGce Petrce: The compound formed between esterase and sodium GTO ak bo cn es Ieee ed: Larayetre B. Menpet and Rosert C. Lawanl The rate of elimina- ‘ tion of nitrogen as influenced by diet factors. I. The influ- ence-of the:texture of the diet.... ..,.<) hues ee ee LarayetTre B. MenpDEL and Ropert C. Lewis: The rate of elimina- tion of nitrogen as influenced by diet facters. Il. The influ- LarayetTe B. Menpev and Rosert C. Lewis: ‘The rate of eli tion of nitrogen as influenced by diet factors. IIT. ence of the character of the ingested protein. _ Baa ee N and B. a +h P. A. Levene and Sona D. Vor LYKE: The separation of d-ala- ee el ee a Donatp D. Van Stryke: The gasometric determination of aliphatic amino nitrogen in minute quantities......................... Dona.tp D. Van Stryke: Improved methods in the gasometric deter- mination of free and conjugated amino-acid nitrogen in the q Car. O. Jonuns and Emit J. BAuMANN: Researches on purines. XIII. i On 2,8-dioxy-1,6-dimethylpurine and 2,6-dioxy-3,4-dimethy]-5- nitropyrimidine (a-dimethylnitrouracil)...................... f Ray E. Nerpie: Polyatomic alcohols as sources of carbon for lower BG tema er, Mt i, ee ak Epwarp B. Meres and Howarp L. Marsu: The comparative compo- ‘ sition of human milk and of cow’s milk....................... Victor C. Myers and Morris 8. Fine: The influence of the admin- istration of creatine and creatinine on the creatine content of Donatp D. Van Stryke: The fate of protein digestion products in the body. II. Determination of amino nitrogen in the tissues Donatp D. Van StYKe and Gustave M. Meyer: The fate of protein digestion products in the body. III. The absorption of amino- acids from the blood by the tissues........................... Donaup D. Van SLYKE and Gustave M. Meyer: The fate of protein digestion products in the body. IV. The locus of chemical transformation of absorbed amino-acids. ..................... iii ence of carbohydrates and fats in the diet......... #4 vaca 19 121 iv Contents Donatp D. Van StyYKeE and Gustave M. Meyer: The fate of protein digestion products in the body. V. The effects of feeding and fasting on the amino-acid content of the tissues........... 231 K. Mryaxe: The influence of salts common in alkali soils upon the growtl of the rice plant...[ogiimee es... 5... eee 235 Paitie A. SHarrer and W. McKim Marriotr: The determination OF Geymputyric acid......... Signe... devas 265 W. M. Marriotr: The determination of acetone.................... 281 W. M. Marriorr: Nephelometric determination of minute quantities Gr acetone... ...... 2. RRR s Gus 289 W. M. Marriorr: The determination of 8-oxybutyric acid in blood and tiases. :........: Ge ie a oe ad aN ae 293 E. V. McCottum and D. R. HoaGianp: Studies of the endogenous metabolism of the pig as modified by various factors. I. The effects of acid and basic salts, and of free mineral acids on the endogenous nitrogen metabolism.....................64.02005 299 otLuUM and D. R. HoaGianp: Studies of the endogenous ism of the pig as modified by various factors. II. ence of fat feeding on —— a metab ee ee | The influence of ie 10 (i a lia A) 321 tabolism.............. ¥ Jacos RoseNBLooM and 8. Rov isis: The non-interference of “‘ptomaines”’ with certain tests for morphine................. 327 M. E. Penninoton, J. 8. Hepsurn, E. Q. Sr. Joun, E. Witmer, M. O. Srarrorp and J. I. Burrett: Bacterial and enzymic changes Th milk QMG'CROAM AGM 6 dvs ic 6 eee stolen. so cee Das 331 Howarp B. Lewis and Bren H. Nicoter: The reaction of some pu- rine, pyrimidine, and hydantoin derivatives with the uric acid and phenol reagents of Folin and Denis....................... 369 Istpor GREENWALD: The formation of glucose from propionic acid in dismeses mellitus: dees ide cis cPibee ccc dey come es ts Cae 375° EB. K. Marswary, Jr. and L. G. Rowntree: The action of radium emanation on lipases... ae.) eae de ee. . a 379 S. R. Beneprer and J. R. Muruin: Note on the determination of amino-acid nitrogen im urine..................ceee eee ceeeeees 385 W. Denis: Metabolism studies on cold-blooded animals. II. The blood and urine of fights... 05)... 5: dues... ss. SS. ee 389 W. Dents: Note on the tolerance shown by elasmobranch fish towards Certain nephrotoxic agents..........sccesesvceawale-coumueeee 395 Cyrus H. Fiske and Howarp T. Karsner: Urea formation in the liver. A study of the urea-forming function by perfusion with fluids containing (@) ammonium carbonate and (b) glycocoll.,. 399 P. A. Levene and C, J, Weer; The saturated fatty acid of kephalin, 419 Contents Vv Tsomas B. Ossporne and Larayetre B. MENDEL (with the codpera- tion of Epna L. Ferry and ALFRED J. WAKEMAN): The influ- ence of butter-fat on growth...) .. jiavaeesasee tenes scacsee-e- 423 J. Du P. OosrHuizEN and O. M. Suepp: The effect of ferments and other substances on the growth of Burley tobacco.. awake 3 ee ' J. R. Greer, E. J. WirzemMann and R. T. Woopyartt: Stuilies on the theory of diabetes. II. Glycid and acetole in the normal and phlorhizinized animal........ 2... .cigepe ene cs die eeeseeis 455 | A. T. Campron: The iodine content of the thyroid and of some bran- Gnier Clete OFeONsS,.. 6. ............ Hie os wknd tole e 465 P. A. Levene and C. J. West: A general methied for the conversion of fatty acids into their lower homologues..................... 475 Artuur W. Dox: Autolysis of mold cultures. II. Influence of ex- haustion of the medium upon the rate of autolysis of Aspergil- TE te a ee aS 479 Suiro Tasutro: Carbon dioxide apparatus III. Another special ap paratus for the estimation of very minute quantities of carbon Gigxidey Ue ee bwawaaess......... P eee On the rate of absorption of chol ae ee eee eee eee Snithistive of Fa ge qaelees upor eae i are Ot a Se 515 W.R. Bioor: On fat absorption. III. Changes in fat during absorp- ee ee es a ae Donautp D. Van Stryke: The hexone bases of casein................. 531 Donaup D. Van Stryke and Freperick J. Bircnarp: The nature of the free amino groups in proteins...................-......... 539 P. A. Levene and C. J. West: On sphingosine. II. The oxidation of sphingosine and dihydrosphingosine.....................-.. 549 _ P. A. Levens and Gustave M. Meyer: On the action of leucocytes ! and of kidney tissue on amino-acids.......................40. 555 _ R. Leprne: “Sucre virtuel’’ and blood glycolysis.................... 559 A. I. Rineer and E. M. Franxet: The chemistry of gluconeogenesis. . VI. The effects of acetaldehyde and propylaldehyde on sugar ‘ formation and acidosis in the diabetic organism............... 563 SY OFM: OO VR cis cca Wiaa'ss . s dein = 1.5. The concentration of the free sodium fluoride cannot be obtained directly, but it will be taken as equal to the total sodium fluoride present. This is, of course, not absolutely true; but we shall see in Table IV that very little of the total sodium fluoride present is combined with the enzyme, especially in weak enzyme mixtures. No appreciable error is therefore in- volved. The headings in the following table are all self-explanatory. The data are derived from Table I. TABLE Il. 5 ec. enzyme solution B in a total volume of 280 ce. Ey NaF CONC. OF ENZYME og y aici PERCENTAGES Free Ens. M 5 actpiry ‘‘10”* ) NaF. Enz. X10 iter | Normality Free | NaF. Enz. ea . minutes 0.00000 0.000 0.00893 0.213 x 10-* a 3 0.0268 0.638X10-* 92.8 42.1 | 1.38 0.0893 2.13 X10-* 190.0 71.7 | 0.305 0.268 |6.38 X10-* 450.0 88.06 | 0.136 * Acidity “10’’ = 10 ce. ay acid in 50 ce. mixture. As an additional example a similar experiment with a crude 10 per cent extract is included. Only the final results are given. k is different in the two tables. This is due in part to slightly TABLE III. 5 ce. enzyme solution used to 250 ce. ethyl butyrate solution. NaF CONC. OF ENZYME aE te a TIME TO REACH + alba ae = axe .—— ) Normality | = Free Ens. | NaF. Ens. ae Ear | minutes mer a 0,0008 0.000 | 47.9 100 | 0 0.0008 |0.233K10-" 53.9 89 11) Sa | 0.0195 |0.464X10-" 58.5 82 18 | 4.56 | 2.12 0.0481 |1.15 X10-*) 77.1 62 eo eee 0.0043 2.25 X10-*) 105.4 46 | 54 | 0.852 | 1.92 0.189 |4.50 X10-* = 158.3" 30 | 7 | 0.420 | 1.98 EES George Peirce _ II different conditions in the two experiments, but mainly to the fact that the enzyme solutions used were entirely different. Other experiments, performed with purified enzyme solutions, gave con- stants nearly equal to the ones in Table II. In every series, moreover, the values of k are constant within the limits of error of the experiment, so that the data agree satisfactorily with equa- tion [3]. No other values for m, n and p in equations [1] and [2] are so consistent with the observations. If the equations as given are true the ratio® portional to the free sodium fluoride. If a large amount of enzyme is used we should expect so much of the sodium fluoride to be combined that the concentration of the free sodium fluoride would be appreciably diminished. In this event, a given con- ' centration of sodium fluoride would have less inhibiting effect in ti strong enzyme solutions than in weak ones. On testing this view 3 inhibition was apparently less in very strong - i the i 80. erence was so slight as to be within the limits of experimental error. Unfortunately no great quantity of uniform purified enzyme remained for experiments in dupli- cate, and as it was quite evident that the enzyme was far from pure, it. did not seem advisable to repeat the experiment until a much purer enzyme could be obtained. The experiment did, however, show that very little sodium fluoride was bound even in enzyme solutions of considerable strength (five times the con- centration in Table II), so that the assumption made in that ~ experiment, that the free sodium fluoride was very nearly equal to the total sodium fluoride, is justified. The following table gives tlhe data on which the preceding conclusion is based. Only the last column requires any explana- tion. This is obtained as follows: The top figures of the first five columns are obtained by extrapolation, and thus a series of © figures is obtained in the fifth column giving an irregularly de- scending series. The total amount of sodium fluoride present is 0.030 mg. per liter, and, starting from this figure, the last column gives a regularly descending series almost directly proportional to the figures in the next to the last column. 6 Note inversion of this ratio. This is done for convenience of presen- tation. ee 12 Esterase and Sodium Fluoride TABLE IV. 20 per cent purified enzyme B. Total volume 280 cc. 0.714 X 10-5N. 0.030 mgm. per liter. CONCENTRATION OF ENZYME CONC. OF ENZYME ABSOLUTE AMOUNTS PERCENTAGES CALCULATED* NaF. ene COMER ate cc-gol. Bin | Mem. dried sub | ging. |Naw.Ems.| | | tents Pmm ts) 280 cc. reaction mixture Amounts ap-| Amounts ap-| (52.6) | (47.4) | (0.90) | 0.030 (total proaching proaching amount zero zero present) 5 §.7 53. 47. 0.89 | 0.030 - 10 11.4 52. 48. 0.93 | 0.029 25 28 .6 55. 45. 0.82 | 0.027 * The last column gives a uniform series, although the figures in the next to the last column diminish regularly. 16 showed the sodium fluoride was free, but fai t.e., it did not show how riment was successful in its primary purpose; 7.e., it | in mixtures of low enzyme concentration almost all, with a given amount of e e. | hee ae One of the most important Points Shout this reneuon in. thal it is reversible. Loevenhart and Peirce? mixed esterase and so- dium fluoride and dialyzed the mixture. After dialysis the solu- tion had regained its original activity. The experiment was conclusive evidence for dissociation of the inactive compound, provided an inactive compound was formed under those conditions (¢.e., mixture of the enzyme with sodium fluoride). It is, however, possible, and indeed probable, that the presence of ethyl butyrate or aleohol or butyric acid or even two or three of these substances is necessary for the formation of the inactive compound.* The evidence for the exact nature of this inactive compound will be presented in a succeeding paper, but the question does not concern us here. The reversibility of its formation is, however, easily demonstrated. For instance, in a 250 ec. mixture containing 153.4 ce. } ethyl butyrate, 10 ec. of enzyme and 1: 6,000,000 sodium fluoride the action proceeded as if only 27.6 per cent of the enzyme present ’ This Journal, ii, p. 406, 1907. * For comment on these points, see Conclusion 7 at the end of this paper. George Peirce 13 were acting. Ata given time (fixed by a preliminary experiment) 5 ce. 75 butyric acid had been produced per 50 ce., so that 50 ce. of the mixture then contained 5 cc. % butyric acid, 5 cc. 7 alco- hol and 25.68 cc. 7 ethyl butyrate. Fifty ec. of this solution were now added to 200 cc. of a mixture containing the same amount of ethyl butyrate, alcohol and butyric acid, but free from enzyme and sodium fluoride. In so doing, the enzyme and sodium fluor- ide were diluted five times, leaving all other factors unchanged. Two possibilities were now open for the further course of the reaction. In the first place, it might have proceeded one-fifth as fast as it did before dilution (where only 27.6 per cent of the enzyme was acting) or it might have produced acid at the same _ rate as a solution originally made up with 2 ce. enzyme in 250 ce. containing sodium fluoride 1:30,000,000. A control soluti made up in this way worked as if about 59 per cent of the enzyme were rie and alee to what was actually ob- niga v out 41 per cent Fok : present in n4 inactive form. The difference was” great enough to be unmistakable, and gave good evidence ' for the fact that the reaction is reversible, whatever the nature _ of the inactive compound. The data in the following experiment were obtained in the usual way, with two exceptions. First: the 50 cc. of solution to be titrated were run into 25 ce. of neutralized 80 per cent alcohol. This stopped the action more effectively than strong sodium fluor- ide. Second: 25 cc. instead of 50 ec. were in several instances used for a titration on account of lack of material. This accounts to a certain degree for divergence of the controls, as the titration errors must be multiplied by two. A partial discussion of the results in the following table has just been given and the full data will now be presented. ae 2 14 Esterase and Sodium Fluoride TABLE V a. 200 ec. Ethyl butyrate solution (50 ce. = 38.35 ec. 35 solution). 10 cc. Enzyme solution C. 25 cc. Sodium fluoride 1:600,000. 15 ec. Water. A (OBSERVED) A (AVERAGE) a (CALCULATED) a(cc.X-actd)| ¢ (min.) 2 a bt | 5t-118.7 2.01 13 .08 2.11 13.94 2:21 14.80 : 3.68 29 .70 3.74 30.29 3.80 30.88 (4.88) (43.7) 218.5 99.8 (5 .00) (45 .2) Fh, 5.32 49.12 5.41 50.48 252.4 138.7 5.49 50.83 ¢ 8.85 | 98 .43 8.94 100.01 500.0 381.3 9.03 | 100.58 (10.00) (114.7) 573.5 10.34 | 120.08 10.43 | 120,78 603.9 | 10.52 121 .67 5 Bo, Pal ie * For explanation of columns 5 and 6, see description of Figure 2. A second similar solution was made up (also in duplicate), and at the end of approximately 44 minutes, 50 ce. of it were added to the following solution: 134 cc. Ethyl butyrate solution 20 cc. § Butyric acid. 5 ce, ¥ Alcohol. 41 cc. Water. The first titration was made within 45 seconds of mixing and the time taken as 0 at this point. en George Peirce “15 TABLE V Bs. OBSERVED AVERAGE x t ree t 4.88 | 0.00 | 4.88 | 0.00 4.87 0.00 5.39 | 14.62 | 5.44 | 14.85 5.49 | 15.08 6.42 | 41.37 6.52 | 43.59 6.61 | 45.80 7.36 | 74.75 | 7.50 | 75.65 7.64 | 76.55 ss 8.84 1118.72 | 8.87 | 119.90 8.90 | 119.68 | 10.42 | 163.80 | 10.28 | 164.63 10.14 | 165.45 Toone 9 VcAND V bp. 9 eee butyl fholition: *% 500 cc. Ethyl butyre 10 ec. Enzyme solution C. 5 cc. Enz - Occ. Sodium fluoride. 12.5 ec. Sodium fluoride 1: 600,000. 40 cc. Water 107.5 cc. Water. OBSERVED AVERAGE OBSERVED AVERAGE z t x t z t z t 4.79 | 12.25 4.91 12.76 2.50 42.28 2.46 41.41 5.03 | 13.27 2.41 41.53 a2. 7.55) Suoee 7.57 21 .87 3.93 73 62 3.87 73.31 | 7.58 | 22.37 3.80 | 73,00 : (4.88) | (99.8) 9.50 | 29.07 9.42 29 .43 5.21 | 106.87 5.13 | 106.73 - 9.34 | 29.78 5.04 | 106.58 (10.00) | (31.6) 10.92 | 35.20 10.69 34.70 6.28 | 135.47 6.17 | 135.61 10.45 | 34.21 ‘ 6.05 | 135.75 7.53 | 175.58 7.39 | 175.29 7.24 | 175.00 8.92 | 222.03 8.75 | 222.20 8.57 | 222.87 (10.00) |(267.7) 10.41 76 .28 10.24 | 276.33 10.06 | 276.38 16 Esterase and Sodium Fluoride The results are also expressed graphically in the following dia- gram. The letters of the curves refer to the preceding table. * represents a reaction going - one-fifth as fast as A. ‘A, C ey D start from the origin. On the curve D, 4.88 ce. acid were pro- duced in 99.8 minutes; but B, as observed, begins at « = 4.88 and t= 0. To make the points = 4.88 on the two curves coincide, 99.8 is added to the values of ¢ in plotting the curve B. In plotting + the values of ¢, for curve A, are multiplied by 5. For az = 4.88,t = 218.5. In order to make the point x = 4.88 on this curve coincide with the corresponding points on B and D, 118.7 is subtracted from the values of ¢. In spite of the apparent complexity of this experiment, the point that it makes is very simple. It shows that a given mix- ture of enzyme, sodium fluoride, ethyl butyrate, alcohol and bu- tyric acid, if diluted five times, with the proper mixture of ethyl butyrate, alcohol and butyric acid, is more than one-fifth as active as it was before dilution. Since an enzyme solution that con- tained no fluoride would have been only one-fifth as active, the additional activity must have come from the partial dissociation of some sort of inactive compound present in the solution. This reversibility of the formation of the inactive compound absolutely excludes destruction of the enzyme by the sodium fluoride. In George Peirce © 17 q addition, the fact that curves B and D so nearly coincide, shows _ that the point of equilibrium demanded by equation [3] is reached _ almost instantly from both directions. CONCLUSIONS, 1. Sodium fluoride forms a compound with the esterase from _ pig’s liver. This compound has little, if any, hydrolytic action on ethyl butyrate. 2. The formation of this compound is reversible. _ 3. When the concentration of the sodium fluoride is varied from - 0.00893 mgm. per liter to 0.268 mgm. per liter, the inhibition . increases from 20.8 per cent to 88.06 per cent. 4, Although theoretically we should expect a given amount of _ sodium fluoride to have less inhibiting effect in mixtures contain- ing a large amount of enzyme, than in weaker enzyme mixtures, the eeeronce actually found was very slight. cates that in the weal , at least, very little of al sodium s into e formation of the inactive com- = : 5. "The following gixtion, based on the supposition that one molecule of the inactive compound contains one molecule of en- " zyme and one molecule of sodium fluoride, agrees with the obser- vations: Cone. free Enzyme X Conc. free NaF = k Cone. (NaF. Enz.) 6. The observations will not agree with an equation based on |. any other supposition as to the number of molecules of sodium - fluoride or enzyme entering into the formation of the inactive ' compound. For this reason it is justifiable to conclude for the |. present that one molecule of the inactive compound contains only one molecule of enzyme and one molecule of sodium fluoride. | 7. It is possible that ethyl.butyrate, alcohol or butyric acid | are also constituents of the inactive compound. This does not affect the argument in any way: It is merely necessary to con- k sider that the “free enzyme’ ” of the above equation represents ® Journ. Amer. Chem. Soc., xxxi, p. 1528, 1910. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 1. 18 Esterase and Sodium Fluoride “free enzyme” is present in the form of a compound with ethyl butyrate, so that only a part of the so-called “free enzyme” is actually free. Whether the sodium fluoride combines with some compound of the enzyme or with the enzyme actually free is immaterial, provided the experiments are so arranged that — the concentration of the substance with which the sodium fluoride combines is proportional to some quantity that we know. This is done by never comparing any two solutions unless the con- centrations of the ethyl butyrate, aleohol, butyric acid and hydro- gen ion are the same at the stage of the reaction where the two solutions are compared. Under such conditions the concentra- tions of all enzyme compounds, except the fluoride compound, are proportional'® to the “free enzyme”’ of the equation, so that the mathematical treatment is justified. 1° This statement must be modified if any enzyme compound that con- tains two molecules of enzyme is present in large amounts, Practically, there is no evidence that such compounds occur, so that for a preliminary — investigation such as the present, the possibility of their existence may — be neglected. ONE LD SEY eee lad THE RATE OF ELIMINATION OF NITROGEN AS INFLUENCED BY DIET FACTORS.' I. THE INFLUENCE OF THE TEXTURE OF THE DIET. By LAFAYETTE B. MENDEL ann ROBERT C. LEWIS. (From the Sheffield Laboratory of Physiological Chemistry, Yale University ,- New Haven, Connecticut.) (Received for publication, August 4, 1913.) INTRODUCTION. The changing view as to the extent of digestion before absorp- tion and the probability that proteins are split to amino-acids has _ raised the question whether all amino-acids are utilized alike. Are not some more resistant in metabolism than others? Does not deaminization take place with greater difficulty in some cases? If it does, one can easily conceive how different may be the behavior of the various proteins in nutrition. Bearing in mind the well- known fact that proteins vary widely in chemical composition, it is evident that the products of absorption after the ingestion of one protein may be much unlike those when another is fed. Thus, if the assumption that amino-acids are of variable resist- ance is correct, we may have an entirely different metabolic picture in the two cases. A study of the rate of elimination of nitrogen? in’ the urine sug- gests itself as a means of ascertaining whether or not the amino- acids behave alike in metabolism. It is obvious that any variation in the ease of deaminization of the amino-acids may lead to a ! The experimental data embodied in the papers of this series are taken from the dissertation submitted by Robert C. Lewis for the degree of Doctor of Philosophy, Yale University, 1912. 2 For a review of the literature on the rate of elimination of nitrogen see: Graffenberger: Zeitschr. f. Biol., xxviii, p. 318, 1891; Hawk: Amer. Journ. of Physiol., x, p. 115, 1903; Stauber: Biochem. Zeitschr., xxv, p. 187, 1910; Wolf: zbid., xl, p. 193, 1912. 19 20 Rate of Nitrogen Elimination change in the nitrogen-output curve after the ingestion of differ- ent amino-acids or of proteins of unlike composition. Other points, however, must be taken into consideration. Will a change in the rate of elimination of nitrogen necessarily be due to a differ- ence in the metabolic behavior of the amino-acids? Certainly several other factors may play a part in this connection. Varia- tions in the rate of the different processes of alimentation—gastric digestion, discharge of the food residues from the stomach and their passage along the digestive tract, pancreatic and intestinal — digestion, absorption—may have a decided influence on the rapidity with which nitrogen leaves the body after a protein meal. Furthermore, metabolic processes distinct from deaminization, such as the behavior of the non-nitrogenous foodstuffs in influenc- protein metabolism, are not without significance in this con- ection. All these factors must be considered as having a bearing | in a study of the rate of elimination of nitrogen. It seerhs quite probable that the lack of concordance in the nitrogen-output curves found by pr ious inv tors may — si due to a variation in incidental factors of the diet, s as the form in which the protein was taken, the amount of car-— bohydrate and fat ingested along with the protein, the water — intake with the meal, and finally the proportion of indigestible material. Since the initiation of these investigations Benedickt — and Roth’ have suggested comparable explanations for the dis- © crepancies in the results of earlier workers. That indigestible materials have an influence on alimentary processes is a familiar fact. Hedblom and Cannon‘ have observed — that coarse branny foods in the diet cause a more rapid discharge of the stomach contents. Recently Mendel and Fine’ have shown that indigestible substances added to the daily meal even in small quantities cause a poorer utilization of protein. It seems quite probable, then, that the rate of elimination of nitrogen in the urine — may be affected by the texture of the diet. . * Benedickt and Roth: Zeittschr. f. klin. Med., Ixxiv, p. 74, 1911. *Hedblom and Cannon: Amer, Journ. Med. Sci., exxxviii, p. 1, 1909. * Mendel and Fine: this Journal, xi, p. 5, 1912. . Lafayette B. Mendel and Robert C. Lewis 21 METHODS. The method employed in the present series of investigations for studying the rate of elimination of nitrogen in the urine has been to collect the urine at definite intervals after the ingestion of pro- tein, to determine its content of nitrogen for the different periods, and to obtain from the results a curve of nitrogen output. Bitches were used as subjects of investigation, the urine being obtained by catheterization. The elimination of individuality was secured by the use of more than one animal for each type of experiment. In the present paper, however, it will be necessary to limit our- selves to the report of a single experiment illustrating each point. The dogs used are designated by a specific letter in the number of the experiment. _ each morning in a single meal a definite ration—the ‘Standard _ Diet.” On experimental days the meal differed from this “Stand- ard Diet’’ either by having something added to it | or by having (or more) of its constituents replaced. The same amount of iz en was always given, however; and in the replacement of F non-nitrogenous constituents isodynamic quantities of some other foodstuff were substituted. Preceding an experiment there was ’ always one day with the “Standard Diet’’ and generally there were two or three. Three of these preliminary days were intro- duced at the beginning of each series so that the animal might have plenty of time for adjustment to the new régime. Thus _ day after day throughout the whole series the food contained the same amount of nitrogen and, except in a few cases, approximately the same calorie value. _ On an experimental day the animal was catheterized in the morning and fed fifteen minutes after the beginning of catheter- ization. To avoid possible secretory disturbances, the tempera- ture of the well mixed food was always the same—20°C.—at the time of ingestion. Usually the animal ate the entire meal greedily. At times, however, the food had to be forced. In all cases feeding was complete in ten minutes or less. Collections of urine were made three, six, nine, twelve, fifteen and twenty-four hours after the beginning of the experiment. In some experiments a control specimen of urine was collected for an hourly period before the While a series of experiments was in progress the animal received 22 Rate of Nitrogen Elimination commencement of the experiment; but in the majority of cases a collection for the twenty-fifth hour was made. Catheterization was planned so as to take exactly ten minutes, the expiration of that time being awaited where necessary, before the completion of the final washing. Nitrogen was determined by the Kjeldahl- Gunning method. The “Standard Diet”’ was, of course, arbitrarily chosen. The only requirements in selecting such a ration were that it must furnish at least the minimal protein requisite and sufficient calo- ries for the needs of the body. ° It seemed desirable, however, to use more than the minimal protein requirement in order to have a liberal output of nitrogen in the urine. The ration adopted consisted of meat,® lard, sucrose, bone ash’ (5 grams), and water, with the addition in some cases of NaCl (2 grams). The water was calculated on a basis of the dry constituents, three times as much water being added as there was dry material. In other words the water was present in the whole ration in about the same pro- portion as it is found in meat. The sugar and lard were given in quantities approximately isodynamie to each other. The meal always contained 0.6 gram of nitrogen per kilo of body weight — and furnished about 70 calories per kilo. The exact calorific value of the diets is not known because the fat content of the meat was not determined. By employing a fixed “Standard Diet” which was easy to duplicate, the rates of elimination of nitrogen under the various experimental conditions with the same and different animals were readily comparable. Inasmuch as the addition of indigestible materials to the ‘““Stand- ard Diet’? suggests itself as a means of studying the influence of the texture of the diet on the character of the nitrogen-output curve, mineral oil, vaseline, paraffin, filter paper, ground cork, agar-agar, bone ash, and sand were added in various experiments. There is little question that these materials pass through the intestine chemically unchanged. Bone ash may possibly be dis- solved to some extent in the hydrochloric acid of the gastric juice. With respect to the mineral oil Bradley and Gasser* have reported * Preserved frozen, according to the method of Gies. ? For the use of bone ash see Steele and Gies, Amer. Journ. of Physiol., xx, p. 343, 1907. * Bradley and Gasser: Proceedings of the American Society of Biological Chemists, December 1911, this Journal, xi, p. xx, 1912, Lafayette B. Mendel and Robert C. Lewis 23 that an emulsified mixture of olive and petroleum oils fed by sound to a dog leads to absorption of both fat and hydrocarbon—a result not in accord with the later experience of Bloor.’ The indigesti- bility of the other materials used is well established. An amount of water equivalent to three times the weight of the superimposed substance was also added with two-fold purpose in the cases of the water-absorbing materials: filter paper, cork, and agar-agar. In the first place, without this extra water the added material could not have been soaked up and well mixed with the food. Secondly, the amount of water carried out through the bowel when these substances were used was relatively very great; and, in order to insure against a large loss of water from the tissues, it was necessary to give an added amount of water with the meal. CONTROL EXPERIMENTS WITH THE “STANDARD DIET.” Before attempting to determine the relation of the different _ diet factors to the rate of nitrogen elimination in the urine, it was necessary to ascertain the nature of the nitrogen-output curve after the ingestion of the “Standard Diet” and to see whether a characteristic curve always followed. In the text each type of experiment is illustrated by a curve, plotted from the data obtained. Curve I, a typical graphic illustration,’ shows the agreement of the nitrogen-output curves of two experiments with the “Standard Diet.” The abscissae represent equal increments of time; the ordinates, grams of nitrogen. Thus at a glance the average hourly rate of elimination of nitrogen for a single period is shown by the value of the ordinate. It is readily seen that after the ingestion of the “Standard Diet” there is a rise in the nitrogen output dur- ing the first period, reaching a maximum in the second three hours, followed by a fall to the initial level early the next morning. In the present work it has always been possible with the same animal to get “standard’’ experiments which agree within reasonably close limits (Curve I). Furthermore “standard’’ curves of duplicate character have been obtained repeatedly with different animals. ® Bloor: This Journal, xv, p. 105, 1913. 10 All curves show the rate of nitrogen output in two experiments, a “standard’”’ experiment (broken line) being plotted for purposes of com- parison. 24 Rate of Nitrogen Elimination Curve I. To illustrate the agreement of duplicate experiments after the ingestion of the ‘‘Standard Diet.’ N gm. O6 ; ta 1 BS. ; "Standard Diet" 1 ss aienihiees 225 gm. meat (N= 7.72 gm.) a4 ; 35" lard 1 is fs 70" sucrose roo : - 4 Sai ee, 5" done ash ' 325 " water pr 0.2 ag | tl ale —_——_—9 f . L On ee A =: pas hrs. ----- Experiment D VII, ‘‘Standard Diet.’’ ———— Experiment D XIII,4‘‘Standard Diet.’’ Lafayette B. Mendel and Robert C. Lewis 25 EXPERIMENTS WITH INDIGESTIBLE MATERIALS. Mineral oil (Curve II). When mineral oil was added to the diet the nitrogen output in the second period was notably less than in the corresponding period after the ingestion of the ‘Standard Diet” alone. Evidently mineral oil causes a slower rate of elimination of nitrogen. Curve II. To illustrate the effect of an addition of mineral oil to the “Standard Diet’’ on the rate of elimination of nitrogen. N g™- o "Standard Diet” ~~ 256 gm. meat (H =. | gm.) ‘7 “ae : ; ' “a Pek, 740". lard go * sucrose iin NaCl hy bone as ---4 soc " water i q — - #8 —_— & = oa See eure. . ide US 2h25 hrs. ----- Experiment G VIII, ‘‘Standard Diet.”’ —— Experiment G XII, ‘‘Standard Diet’? + 75 grams mineral oil. 1 A colorless, purified product sold under the trade name of ‘“‘Alboline.’’ 26 Rate of Nitrogen Elimination Vaseline (Curve ITT). The effect of vaseline on the curve of nitrogen elimination is similar to, but more marked than that of mineral oil. The nitro- gen output in each of the first three periods is smaller than in the “standard” experiment; afterwards the two curves are almost identical. There is a delay in the excretion of nitrogen. Curve III. To illustrate the effect of an addition of vaseline to the “Standard Diet’’ on the rate of elimination of nitrogen. N gm. 0.6 2:0 | -> ; i 0.5 : : oe Standard Diet” ‘ 225 gm. meat (N= 7.27 gm.) Rtentd i. 85" lara ; 1 70" sucrose | bone ash 325 " water Ti 2 a aa: has hs. w---- Experiment D IV, ‘‘Standard Diet.’’ -——— Experiment D VI, ‘Standard Diet’’ + 75 grams vaseline. ? Yellow petroleum jelly (M.P.=38°C.). —— Lafayette B. Mendel and Robert C. Lewis 27 Paraffin (Curve IV). The experiment with paraffin shows a much more decided flat- tening of the nitrogen-output curve, 7.e., a preliminary’ delay in excretion of nitrogen, than was the case with either of the softer | petroleum products. The rate of elimination of nitrogen is not i only lower in the earlier periods than in the “standard” experiment, but also higher during the latter part of the day. | Curve IV. To illustrate the effect of an addition of paraffin to the | “Standard Diet’’ on the rate of elimination of nitrogen. N gm. | 07 aaah "Standerd Diet" ' 2 ! 500 gm. meat (N = 10.27 gm.) 0.6 | Lae 50" lara ‘ BS i er : 100" — sucrose 705) - I 2" Wecl H 5" bone ash 04 450" water | i ! eae leks en 03 }---—- } t ! i] u i a ay | 0.2 |i ena t a a 0.1 ; 5 Wie Gag. (Raais 425 hrs. ----- Experiment F V, ‘‘Standard Diet.’’ —— Experiment F IX, “Standard Diet’’ + 75 grams of paraffin.“ 18 Fine shavings, obtained by scraping a cake of paraffin (M.P.=51°C.) with a knife. 14 Large quantity of paraffin feces during the night (15-24 hour) period. 28 Rate of Nitrogen Elimination Filter Paper® (Curve V). The rate of elimination of nitrogen during the earlier periods after the ‘ingestion of the “Standard Diet’’ plus filter paper is lower than in the “standard” experiment; during the later hours it is higher than normal. Thus, as was the case with paraffin, there results a very decided flattening of the nitrogen-output curve. Curve V. To illustrate the effect of an addition of filter paper to the ‘Standard Diet’’ on the rate of elimination of nitrogen. N g™. a ='=5 ' ‘ . : "Standerd Diet" 1 ' \ 0.6 | { ous S00 gm. meat (l= 10.27 gm.) ' 50" lara i . 0.5 | 1 gag 100 he sucrose .. a 2" Yeci 0.4 Bi gad 5" done ash L 450 * weter ° | a -——_—— qi = am o2h os 1 i 1 | ieee et Ol ; u o 5 — = a ie 2425 hrs. ----- Experiment F V, ‘‘Standard Diet.’’ Experiment F VIII, “Standard Diet’’ + 75 grams filter paper'® and 225 grams water. * Cut up in small pieces. * Large quantities of paper feces during the 4th and 5th three-hour periods and during the night (nine-hour) period. Lafayette B. Mendel and Robert C. Lewis 29 Cork” (Curve VI). With cork there is likewise a much slower elimination of nitrogen than with the “Standard Diet’’ alone. In this case the total nitrogen output for the entire day is lower than in the “‘standard”’ experiment. This, however, is not the only cause of the sub- normal elimination of nitrogen in the early periods of the day; _ for the character of the nitrogen-output curve is radically different _ from the “standard’’—lower during the early periods and higher during the later periods of the day. Curve VI. To illustrate the effect of an addition of cork to the “‘Stan- dard Diet’’ on the rate of elimination of nitrogen. oN gn o.1 "Standerd Diet™ a ® os 250 gm- meat (N= 8.68 gm.) 40 " lard a _J rt B™ wacl wae ---4 pea om ae ' 1 ' i} ; >, * bone ash 1 <4 350 " water eR 5 ont i l 1 Ss a6 eo) Ne als 2has hes. ----- Experiment G VIII, “Standard Diet.”’ Experiment G XV, “Standard Diet”? + 50 grams cork!8 and 150 grams water. 1 Finely ground in a coffee mill. 8 Large quantities of cork feces during the 4th and 5th three-hour periods and during the night (nine-hour) period. 30 Rate of Nitrogen Elimination Agar-agar® (Curve VII). Of all the indigestible materials used the agar-agar caused the most pronounced delay in the nitrogen output. In the experiment here reported there was a rise in the second period over the value — of the first three hours and then very little change for twelve hours; in other experiments there was a similar though slightly less marked effect. Curve VII. To illustrate the effect of an addition of agar-agar to the “Standard Diet’’ on the rate of elimination of nitrogen. gm "Standerd Diet" of 200 gm. meat (N= 5.58 gm.) rae 30" lara 0.3) ! 50 " © sucrose ee L 5" bone esh 0.2. ee 250" water 1 fe 0.! Pisin on call a eee Ss .s & te 2425 hrs. w--- Experiment B I, ‘‘Standard Diet.”’ Experiment B III, ‘‘Standard Diet”’ + 75 grams agar-agar®® and 225 grams water. Very finely chopped. 2 First agar-agar feces during first three-hour period. No note kept of subsequent defecations; feces in almost every period, however. Lafayette B. Mendel and Robert C. Lewis 31 Bone Ash (Curve VIII). With the addition of bone ash to the ‘Standard Diet” there is a flattening of the curve, but by no means to such an extent as with any of the previously mentioned indigestible materials except the softer petroleum products. There is a delayed excretion of nitrogen in the first two periods, followed by a slight compensatory rise during the next six hours, the curve afterward running parallel to that of the ‘‘standard’”’ experiment. Curve VIII. To illustrate the effect of an addition of bone ash to the “Standard Diet’’ on the rate of elimination of nitrogen. gm. ro-c-5 : "Standard Diet" { é oe 225 em. meat (X= 7.79 on.) oan $5" lard of ' 70" sucrose 1 - 5" bone ash ee | + O35 325" water ' o2h ee ; 22 @& Out } ! J Bs f 9 He as 2425 hrs. ----- Experiment D VII, “Standard Diet.”’ —— Experiment D IX, ‘Standard Diet’’ + 75 grams bone ash. 32 Rate of Nitrogen Elimination Sand (Curve IX). The effect on the nitrogen-output curve of an addition of very fine sand to the ‘‘Standard Diet” is entirely different from anything so far reported. The rate of elimination of nitrogen during the first two periods is notably higher than in the corresponding periods of the “standard” experiment; afterward the two curves run © parallel. . ee Curve IX. To illustrate the effect of an addition of sand to the “Stan- dard Diet” on the rate of elimination of nitrogen. | : N i 0.7) | 06 Pe “— “Standard Die niet a eo Os ae , ~ 225 em- meat (N= 7.66 en.) id mt ; 35" lera 0.4 Q St 70 "sucrose ! 5" tone ash * i =. 325." water 0.3) | | Stabe. " ! ' 02h | i] pi 0.1} | 1 ‘ ! : ! 5 6 Pp Kk aS hrs «s--- Experiment D XIII, “Standard Diet.” Experiment D XVI, “Standard Diet’’ +°75 grams sand. Lafayette B. Mendel and Robert C. Lewis 33 DISCUSSION. The experiments with a variety of indigestible materials have shown a slower rate of elimination of nitrogen after the addition of these substances to the “Standard Diet” except in the case of sand. Obviously there has been some delay in the processes of alimentation; for, excepting differences in the amount of water absorbed,” the purely metabolic conditions are the same when an indigestible substance is included in the daily meal as when the “Standard Diet” alone is fed. The prime factor in this delay must have been a slower rate of absorption, whether induced by a retardation of the discharge of the gastric contents, a delay in digestion, an adsorption of digestive products by the indigestible material, or a loss of absorbable material by an early evacuation of the bowel. Let us consider the bearing of each of these con- tributory factors on the present work. In all probability no delay in the discharge from the stomach occurred when indigestible materials were added to the diet. With mineral oil, vaseline, paraffin, and bone ash the passage of _ . food onward must have been as rapid as under normal conditions; for during the first periods with these materials there was no - marked decrease in the nitrogen output below that of the ‘“stand- _ ard” experiment. The early appearance of the added indigestible material in the feces following the ingestion of filter paper, cork, and agar-agar” suggests an acceleration rather than a retardation of the normal gastric discharge. This is in harmony with the report of Hedblom and Cannon” that branny foods cause a more rapid emptying of the stomach. A delayed absorption on account of a sub-normal rate of diges- tion in the experiments with indigestible materials is quite possible. *t With several of the indigestible substances—cork, filter paper, and agar-agar—a large amount of water was excreted through the bowel. As a result the volume of urine, and hence the water absorbed, was very small. That the lack of water was not an important factor in causing a retardation of the rate of elimination of nitrogen has been shown by an experiment with agar-agar in which a very large amount of water was given. In this case there was a normal flow of urine, but the same effect was obtained as when the secretion of urine was small. *2 The influence of these substances on the emptying of the bowel is indicated in the reports of the experiments. *8 Hedblom and Cannon: Amer. Journ. Med. Sci., exxxviii, p. 1, 1909. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. l. 34 Rate of Nitrogen Elimination It was pointed out in the preceding paragraph that some of the indigestible materials caused a more rapid discharge from the stomach. This might very well lead to a slower rate of digestion because the preliminary gastric proteolysis was inhibited before it had advanced very far, and the intestine thus received much more complex residues than under normal conditions. This pos- sibility of an unfavorable result from sparing the stomach at the expense of the intestine has been previously mentioned by Cohn- heim™ and Cannon.” It is interesting to note that it was in the — cases where there was an early emptying of the stomach that the greatest delay in elimination of nitrogen, and hence in absorption, 7 occurred. Furthermore, digestion may have been delayed be- — cause the food residues, mixed as they must have been with the — indigestible materials, are rendered more or less inaccessible to the action of the digestive enzymes. Another possible explanation for the delay in absorption when indigestible substances are added to the diet is that the products — of digestion are adsorbed by the added indigestible material. It .— is quite conceivable that Agar, for example, might adsorb the. — soluble intestinal contents. An examination of the data shows — that the indigestible materials exerted a retarding influence c the rate of absorption, as measured by the rate of elimination nitrogen. This retardation rogressively greater in the follo ing order—mineral oil, vaseline, bone ash, paraffin, filter paper, a cork, agar-agar—corresponding with the comparative adsorptive — power of the same substances. This parallelism between delayed absorption and power of adsorption seems to be more than a coinci- dence. Adsorption of intestinal contents by the indigeStible adju- vants might cause a slower rate of absorption in two ways: in the first place, the partially digested protein residues might im part be rendered less accessible to enzyme action; secondly, a smaller — proportion of the completely digested protein residues would come in contact with the intestinal wall, and for this reason absorption would be hindered. It is unlikely that the delayed absorption in the experiments with indigestible materials is attributable to an bod emptying ~ Pe ee) ee ee eee Bo *Cohnheim: Quoted by Cannon. * Cannon: The Mechanical Factors of Digestion, International Medical — Monographs I (Longmans, Green and Company), 1911, p. 123. Lafayette B. Mendel and Robert C. Lewis = 35 of the bowel and a consequent loss of available nitrogen. The more than compensatory rise in the urinary nitrogen output above the normal which always occurs in the later periods with the substances causing early evacuation makes such an explanation questionable. There remains a consideration of the results obtained with sand. The data presented emphasize an increased elimination of nitrogen in the first periods after the meal when sand is added to the “Stand- ard Diet.’’ Can this rise be caused by a more rapid discharge of the gastric contents? No evidence has been found in the litera- ture to warrant such a conclusion. On the contrary the results of Hedblom and Cannon,” showing that hard irregular pieces of dried starch paste in the diet caused a slower discharge of the stomach, appear to speak against such an explanation. It is ' unlikely that sand has mechanically stimulated an increased _ secretion, the reabsorption”’ of which has raised the nitrogen out- put of these first two periods; for Pawlow®* has demonstrated conclusively that the blowing of sand with force against the walls of the inactive stomach does not stimulate gastric secretion. The - secretion is f utherae by the follow- q ” Dann the middle of the fourth day of a fast the urine was col- lected for a three-hour control period. ‘ A quantity of sand (and a little water) was then given; and the urine was subsequently collected at three-hour intervals for nine hours. There was no increase in the nitrogen output of these later periods over that of the control period.” °® Hedblom and Cannon: Amer. Journ. Med. Set., exxxviii, p. 1, 1909. *7 Mosenthal (Journ. Exp. Med., xiii, p. 319, 1911) has attributed to the intestinal secretion a considerable source of absorbable nitrogen. From experiments on dogs with isolated loops of intestine he estimated that the nitrogen content of the succus entericus secreted in twenty-four hours was equivalent to about 35 per cent of the nitrogen intake. Inasmuch as the | feces contained nitrogen equivalent to only 10 per cent of the intake, the _ major part of the intestinal secretion must have been reabsorbed. ** Pawlow: The Work of the Digestive Glands, translated by W. H. Thomp- son (Charles Griffin and Company), 1902, pp. 86-90. Ina previous experiment we had ascertained that the eget 36 Rate of Nitrogen Elimination SUMMARY. The typical curve of nitrogen elimination on a selected mixed diet shows a rise in the first period, reaching a maximum in the second three hours, followed by a fall to the initial level early the next day. With a definite diet it has always been possible to duplicate experiments on the same animal. Different animals on the same type of diet have given parallel curves. - A delay in the elimination of nitrogen is caused by the addition to the diet of such indigestible materials as mineral oil, vaseline, — bone ash, paraffin, filter paper, cork, and agar-agar—substances which act in a purely mechanical as contrasted with a chemical manner. Invariably there is a subnormal rate of nitrogen output — in the first periods following ingestion of the meal; with paraffin, — filter paper, cork, and agar-agar this is followed by a higher rate — in the later periods. The effect of the indigestible materials is — progressively greater in the order in which they are given above. A delayed absorption of the nitrogen intake is presumedly responsible for the slower rate of elimination of nitrogen. As possible causes of this retardation of absorption the following have ! been suggested: (1) a slower rate of digestion caused by an early emptying of the stomach and a consequent early exclusion of 4 gastric proteolysis, with the possibility of a more prolonged in- testinal digestion; (2) a slower rate of digestion caused by an adsorption of partially digested protein residues by the added — indigestible material, making them less readily accessible to the action of the digestive enzymes; (8) an adsorption ‘of the final digestive products by the indigestible substance whereby their absorption from the intestine is hindered. Sand gives an exception to the results obtained with the other indigestible materials studied, as it causes an elimination of nitro- — gen above the normal in the first six hours. This rise is presumably — not caused by an increased secretion and subsequent reabsorption — of digestive juices, for the ingestion of sand during starvation has — no effect on the nitrogen-output curve. o> jae el . THE RATE OF ELIMINATION OF NITROGEN AS INFLUENCED BY DIET FACTORS. II. THE INFLUENCE OF CARBOHYDRATES AND FATS IN THE DIET. By LAFAYETTE B. MENDEL ann ROBERT CG. LEWIS. (From the Sheffield Laboratory of Physiological Chemistry, Yale University, New Haven, Connecticut.) (Received for publication, August 4, 1913.) Although the general effect of carbohydrates in the diet on the rate of elimination of nitrogen in the urine seems to be well estab- lished, no study of the comparative behavior of different car- _ bohydrates has been attempted. In the past the method of in- E vestigation employed has been to superimpose the carbohydrate to be studied on a “standard” diet, to determine the rate of elimination of nitrogen after this augmented meal, and to ascer- tain the effect of the carbohydrate on the nitrogen-output curve by contrast with the rate when the “standard” diet alone was fed. _ The objection to this method is that on the day when carbohy- | drate is given the diet has a greater calorific value than the ‘‘stand- ard” diet, so that the conditions on the two experimental days are not comparable from the standpoint of energy intake. In the present investigation' an isodynamic quantity of carbohydrate ) was substituted for the non-nitrogenous constituents of the 1 “Standard Diet,” and thus the calorie value of the food of all '. days remained the same. The following carbohydrates were _ chosen for study: the polysaccharides, starch and soluble starch; "the disaccharide, sucrose; and the monosaccharide, dextrose. 1 The methods employed were those outlined in our first paper (cf. this Journal, xvi, p. 19, 1913). 37 38 Rate of Nitrogen Elimination EXPERIMENTS WITH CARBOHYDRATES. Starch? (Curve I). When starch was substituted for the non-nitrogenous constit- uents of the “Standard Diet” there was a distinct delay in the elimination of nitrogen, the amount of nitrogen excreted being smaller than in the ‘“‘standard’”’ experiment in the earlier periods of the day and larger in the later periods. . Curve I. To illustrate the effect on the rate of nitrogen elimination ee. oe ee of substituting s‘arch for the non-nitrogenous constituents of the “‘Standard Diet.”’ N gm "Standard Diet” 0.4 ; | aa ea 200 gm. neat (N= 5.58 gm.) i | 30 " lerd Oo. 3 Beet, —— 50" sucrose ic f 5° bone ash a2 4 a =f shee 250 " water Ht 1 0.1 \ ! 1 5 6 lg 2 5 2|4 hrs, ---- Experiment A IX, “Standard Diet.’’ Experiment A XIV, lard and sucrose of ‘Standard Diet”’ replaced by starch (118 grams). —o — — _ — _ ee * Arrowroot starch was mixed with water and heated in an autoclave for fifteen minutes in order to rupture the starch grains. le ee a Lafayette B. Mendel and Robert C. Lewis 39 Soluble starch (Curve II). With the use of soluble starch in place of starch the retarding effect on the nitrogen excretion was even more marked. { Curve II. To illustrate the effect on the rate of nitrogen elimination of substituting soluble starch for the non-nitrogenous constituents of the “Standard Diet.” N gm "Standard Diet” 225 em. meat (N= , a 35 " lard a a a 70 " guerose rf a eo" Bete LL ——+ 5" bone ash 325." water a 1 ! ! Reem y Rie. ee Poe ee ee ae of se Be ap Rey is 2has hrs. ---- Experiment D XXIV, ‘‘Standard Diet.’ _ —— Experiment D XXVII, lard and sucrose of ‘Standard Diet’’ _ replaced by soluble starch (150 grams). 40 Rate of Nitrogen Elimination Sucrose (Gurve IED: When sucrose was substituted for the lard of the “Standard Diet” the maximum output of nitrogen did not occur until the third three-hour period, as contrasted with the second period in the ‘‘standard”’ experiment and in experiments with the carbohy- drates above reported. The nitrogen output was relatively much smaller in the earlier periods of the day and much larger in the later periods than was the case with either of the polysaccharides. In other words the flattening effect of sucrose on the nitrogen- output curve was much more pronounced than that of starch or of soluble starch. Curve II’. To illustrate the effect on the rate of nitrogen elimination of substituting sucrose for the lard of the “‘Standard Diet.”’ N g™: a OL ' } i "Standard Diet" 225 gm meat (N= 7.61 gm.) ! 0.6 35“ lard : 70 " gsucrose os : 2" Wacl ers — 5" bone ash of $25 " water \ f 0.3) . 4 i t=, o.2! - =e : io ag oe Cf a = oi} Sau we cite “als daas hrs ~--+=- Experiment D XXIV, “Standard Diet.’’ Experiment D XXVI, lard of ‘Standard Diet” replaced by suc- rose (80 grams), Lafayette B. Mendel and Robert C. Lewis = 41 Dextrose (Curve IV). With dextrose replacing the non-nitrogenous constituents of the “Standard Diet”’ the preliminary delay in nitrogen excretion was much greater than with any of the other carbohydrates studied, the nitrogen-output curve having a much more flattened aspect. Curve IV. To illustrate the effect on the rate of nitrogen elimination of substituting dextrose for the non-nitrogenous constituents of the “‘Stan- dard Diet.” x "Standard Diet” 0.5 | reocf 170 gm. meat (DN = 6.41 gm.) : 25" lara 0.4 : 60 " sucrose } | 2" acl 0.3| — ae) 5" bone ash ie mal a 250 " water 02 o ! t — © 2. ees 2425 hrs. ---- Experiment H I, “Standard Diet.” Experiment H ITI, lard and sucrose of ‘Standard Diet”’ replaced by dextrose (117 grams). Sucrose; Lard being present in the diet. In the experiments above reported the diets were fat-free and in this respect not comparable with the “Standard Diet.’’ What effect would an increased amount of sucrose, for example, have on the rate of nitrogen elimination if lard were also present in the diet? In a further experiment the meal contained an amount of sucrose isodynamic to the non-nitrogenous constituents of the 42 Rate of Nitrogen Elimination “Standard Diet’’ and enough fat (94 grams of lard) to make the ratio of sugar to fat approximately the same as in the “Standard Diet.””. With such a procedure an increased number of calories was given, but to accomplish the desired end this was necessary. There was the same preliminary delay in the nitrogen excretion in this case with sucrose as when no fat was present in the diet. DISCUSSION. The substitution of the different carbohydrates for the non- — nitrogenous constituents of the “Standard Diet’’ resulted in a slower rate of elimination of nitrogen, a flattening of the nitrogen- output curve. The carbohydrates studied were progressively — more effective in the following order: starch, soluble starch, sucrose, and dextrose. With the method employed in the present study with carbohy- drates two important changes have been made in the diet: (a) fat — has been removed; (b) an added amount of carbohydrate has been given. Which of these two changes is of paramount importance in causing the results above reported? An experiment in which fat as well as sucrose was added to the “Standard Diet” shows that. the removal of the fat was not the causal factor; for here, as in the cases where sucrose replaced lard, the curve of nitrogen elimination — is considerably flattened despite the presence of an abundance of — fat. The added carbohydrate must have been directly respon- — sible, then, for the slower rate of nitrogen elimination. . The results of the present investigation with carbohydrates are completely in harmony with those of previous investigators. Vogt — (1906) found that the addition of rice or rice flour to a meat diet — caused a slower rate of elimination of nitrogen than meat alone. Levene and Kober (1909) reported that with the addition of — starch to a “standard’’ diet containing plasmon, cracker meal, : and lard the course of nitrogen elimination did “not differ mate- 7 rially from that of the standard diet.’’ A careful examination of — the data presented by these authors discloses, however, a slight — flattening of the nitrogen-output curve when starch was added to j Eee the diet. According to Van Slyke and White (1911) starch super- — imposed on a diet of fish, cracker meal, and lard caused a delayed — elimination of nitrogen of about the same magnitude as in the — starch experiments of the present work. Falta and Gigon (1908) Lafayette B. Mendel and Robert C. Lewis 43 found a delayed excretion of nitrogen after the addition of either wheat flour or levulose to a meat diet, the effect being more marked in the case of the latter carbohydrate. This result with levulose agrees with that of Falta, Grote, and Staehelin (1907). Pari (1908) reported that with the addition of sucrose to a meat diet there was a retardation of the nitrogen excretion. Interesting in this _ connection are the experiments of Boettcher and Vogt (1909) in _ which subcutaneous injections of dextrose (5-10 grams) caused a flattening of the nitrogen-output curve. All of these investiga- tors worked with dogs.* Lusk (1912) fed dextrose alone to dogs _ 24 hours after the last meal and found a nitrogen output ‘ower than the fasting level during the hour following the dextrose in- take. Subsequently there was a compensatory rise in the nitrogen elimination. Concerning the manner in which carbohydrate may be responsi-~ ble for a delay in nitrogen excretion several possibilities must be considered, viz: a subnormal rate of discharge of the stomach contents, a retardation of digestion, a delayed absorption, altered - metabolic processes. Van Slyke and White (1911) have given no _ experimental proof for their conclusions that the retardation of nitrogen elimination when starch is added to the diet is caused by a delay in digestion and absorption. The expefiments of Boett- " cher and Vogt (1909), showing a delay in absorption after sub- cutaneous or intravenous dextrose injections in five out of seven eases, are hardly comparable with the experiments of the present _ series where the carbohydrate was given per os. In fact, no con- _ clusive evidence has been found in the literature to the effect that - a subnormal rate of any of the alimentary processes is caused by an addition of carbohydrate to the diet. On the contrary, the report of Cannon‘ that a mixture of carbohydrate and protein foods leaves the stomach more rapidly than protein alone, whereas f fat has a retarding action on the emptying of the stomach, makes _ it probable that the addition of carbohydrate to, and the removal of fat from the diet in the present experiments is, if anything, | followed by a more rapid discharge of the gastric contents than in the “standard” experiment. 3 Wolf (1912) studied the rate of elimination of nitrogen after the inges- ' tion of starch by a fasting man. The results have no bearing on the ex- periments here reported. 4Cannon: Amer. Journ. of Physiol., xii, p. 387, 1904. 44 Rate of Nitrogen Elimination There is some evidence in the literature that variations in meta- bolic processes are responsible for the slower rate of nitrogen elimination under the influence of carbohydrates. Falta and Gigon (1908) and Par. (1908) have attributed: this delay to the protein-sparing action of carbohydrates; for after a fast, when the glycogen depots are almost depleted, the carbohydrates no longer — exert a retarding action on the nitrogen-output curve. The reason for this, according to these authors, is that the carbohy- drates now go to make up the depleted glycogen supply in preference — to being immediately burned. Boettcher and Vogt (1909) think — that a disturbance of intermediary metabolism is in part responsi- — ble for the delay in nitrogen excretion obtained after subcutaneous — dextrose injections, although they offer no experimental proof for — their contention. The consensus of opinion, then, seems to favor — a disturbance of metabolic processes, rather than a delay in ali- mentation, as the causal factor in the retardation of nitrogen ex- — cretion when carbohydrate is present in the diet. Although the experimental data obtained in this study do not — warrant the adoption of a final theory as to how the carbohy- drates act to retard nitrogen excretion, the writer is inclined to the — belief that the protein-sparing action of carbohydrate causes this — delay. When carbohydrate is present in the diet, it is digested, — absorbed, and burned; while the protein residues, which are simul- taneously absorbed, are temporarily spared to some extent and — are only completely metabolized when carbohydrate is no longer available, a preliminary delay in nitrogen excretion thus occurring. If such a theory holds, the physiological, six-carbon sugar dex- — trose should be more efficient than the polysaccharide starch in causing a retardation of nitrogen excretion; for the more nearly — the carbohydrate is prepared for absorption when ingested, the sooner should its sparing action come to expression, and so the — greater should be the delay in nitrogen elimination. As a matter — of fact the carbohydrates studied did show a progressively greater retarding effect in the order: starch, soluble starch, sucrose, dex- trose. Thus the theory accounts for all the results obtained. In conclusion reference should be made to the effect of indigesti- i ble carbohydrates in the diet on the rate of elimination of nitrogen. — In the previous paper® it was shown that cellulose—filter paper — * Mendel and Lewis: This Journal, xvi, p. 19, 1918. ee ee ee Lafayette B. Mendel and Robert C. Lewis 45 and cork—and agar-agar caused a delay in nitrogen excretion. It is obvious, however, that the explanations of the similar effect on the rate of nitrogen elimination after the ingestion of digestible and indigestible carbohydrates, respectively, are radically different. SUMMARY OF RESULTS WITH CARBOHYDRATES. Carbohydrates in the diet cause a slower rate of elimination of nitrogen after a protein meal, the various carbohydrates studied having a progressively greater effect in the following order: starch, soluble starch, sucrose, dextrose. The experimental data do not warrant the adoption of more than a tentative theory as to the explanation of the retardation of nitro- gen excretion when carbohydrates are present in the diet. It seems quite probable, however, that the protein-sparing action of carbohydrate is responsible for this delay. At any rate all the results obtained in the experiments with carbohydrates may be explained by such a theory. Tue RELATION oF Fats IN THE DIET TO THE RATE OF ELIMINATION OF NITROGEN. | Nocomparative study of fats of different texture—soft or hard— has been attempted in the few previous investigations of the influ- ence of fat in the diet on the rate of nitrogen elimination. Most of the work in the past has been done by superimposing the fat on a “standard” diet, and determining its effect on the nitrogen-out- put curve. Inasmuch as this procedure is open to the objection that with the addition of fat to the diet an increased number of calories is given, a method similar to that employed in the study of carbohydrates® was adopted for fats, the non-nitrogenous con- ‘stituents of the “Standard Diet” being replaced by the fat to be _ studied. The following fats of widely different textures were used: cotton-seed oil, lard, and “Oleo-stearin’’? (M.P.=53°C.). ® See the first part of this paper. 7 Armour and Company kindly furnished this product. 46 Rate of Nitrogen Elimination EXPERIMENTS WITH FATS. Cotton-seed oil (Curve V). The substitution of cotton-seed oil for the non-nitrogenous constituents of the “Standard Diet” had very little effect on the course of the nitrogen-output curve, the only variation from the “standard” occurring in the second three-hour period where the nitrogen output was decreased. Curve V. To illustrate the effect on the rate of nitrogen elimination of substituting cotton-seed oil for the non-nitrogenous constituents of the ‘Standard Diet.’’ N gm. 0.6 a "Standard Diet” i 250 gm meat (N= 6.68 gm ) 1 40" lard ibe. 0.5 fl on on all /. l 80 " sucrose ‘ae ! ‘ 0.4) 4 2 NaCl = 5 " bone ash UJ ” 0.3 a 350 water t U U 0.2] 5" bone ash 125" water U Se 1 9 Ie is 214 ----- Experiment D XXXIII, meat of ‘Standard Diet”’ replaced by ’ 60 grams extracted meat (N=8.05 grams) and 170 grams water. 4 Experiment D XXXVIII, meat of ‘‘Standard Diet’”’ replaced by 58 grams ovovitellin (N=7.99 grams) and 170 grams water. 64 Rate of Nitrogen Elimination Edestin (Curve VI). The nitrogen-output curve with edestin was not of such a flat- tened aspect as that with extracted meat. The nitrogen excre- tion in the earlier periods was larger; in the later periods, smaller than with extracted meat. The edestin curve, however, was very much the same as that with fresh meat. - | Curve VI. Comparison of the rates of nitrogen elimination on diets containing meat, extracted meat, and edestin, respectively. N g™ 0.1 "Standard Diet” 0.6) 200 gm. meet (Ne@ 6.84 gm.) 50 " lerd 0.5 30" sucrose = §6Racl 04 5" bone ash . 250" water a 4 h t a t———., 0.2 od i oo en oo i eo A 4 Ol aso a ee ; 1S. 6 ape Sle -ouls 24 hrs —--— Experiment H XIV, “Standard Diet,’ --+-- Experiment H VII, meat of “Standard Diet’’ replaced by 58 grams extracted meat (N=7,11 grams). Experiment H XI, meat of ‘Standard Diet” replaced by 44 grams _ edestin (N=7,14 grams), Lafayette B. Mendel and Robert C. Lewis 65 “Glidine” (Curve VII). When “Glidine” constituted the protein intake, the nitrogen excretion was larger during the earlier periods of the day than with extracted meat. The character of the nitrogen-output curve was practically the same as that with fresh meat; the two curves ran parallel, that with ‘Glidine” being at a higher level. ea ae eee ™ aT) Curve VII. Comparison of the rates of nitrogen elimination on diets containing meat, extracted meat, and ‘‘Glidine,’’ respectively. N gm 0.1) ss "Standard Diet” 0.6 | 200 gm. meat (N= 6.84 gm.) ‘ —_ ' a, | | 50" lerd — OS! | | "sucrose ¢.| , NeCl . ca} " bone ash a i Bsa : " water iJ ames 02 “™s bP we wis 214 hrs. —-— Experiment H XIV, “Standard Diet.’ weeee Experiment H VII, meat of “Standard Diet’’ replaced by 53 grams extracted meat (N=7.11 grams). Experiment H IX, meat of “Standard Diet’’ replaced by 48 grams “‘Glidine’’ (N=7.10 grams). THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL, XVI, NO. l. 66 Rate of Nitrogen Elimination e Gelatin (Curve VIII). Again with gelatin the nitrogen-output curve was not as flat- tened as that with extracted meat, but identical in character with the “standard” (fresh meat) curve. There was a negative balance with gelatin and the nitrogen excretion for all the periods was higher than with meat. Curve VIII.- Comparison of the rates of nitrogen elimination on diets containing meat, extracted meat, and gelatin, respectively. gm. 0.7 "Standard Diet" ee 200 gm. meat (N= 6.84 gm.) 06 | | 50" ara : l 30" sucrose 0.5 | | 2" NaCl oant n b tr Ii 5 one ash | 04 Ii HT 250" water —" 1 Biiee i} ‘hes 4 andslened Lu 0. 0. 0.! 3 | 9 le as 4 hrs —--— Experiment H XIV, “Standard Diet.’ w<-0- Experiment H VII, meat of “Standard Diet” replaced by 53 grams extracted meat (N=7,11 grams). gelatin (N=7,08 grams), Experiment H X, meat of ‘Standard Diet” replaced by 47 grams” Lafayette B. Mendel and Robert C. Lewis 67 Soy bean (Curve IX). On account of the comparatively poor utilization of soy bean!’ _ its ingestion was followed by a smaller nitrogen output in all the _ periods of the day than when extracted meat was fed. Further- more, this lower curve was not parallel to that with extracted meat; its character was quite different. In the earlier periods of _ the experiment here reported a smaller percentage of the 24-hour nitrogen output was excreted than when meat was fed; in the later periods, a larger percentage. In a second experiment the maximum nitrogen excretion did not occur until the third three- hour period, instead of the second. In both cases, then, there was a delay in the elimination of nitrogen independent of the poorer utilization of the soy bean. Curve IX. Comparison of the rates of nitrogen elimination on diets containing extracted meat and soy bean, respectively. . N ie 7 -~ J gm. "Standard Diet" 200 gm. meat (N= 7.15 gm.) co lard sucrose bone ash Nacl water SS; Se Se U ! t l i a fl CS zz. 214 hrs, _ ----- Experiment H VII, meat of ‘Standard Diet’”’ replaced by 53 rams extracted meat (N=7.11 grams) and 150 grams water. Experiment H XII, meat of ‘‘Standard Diet’”’ replaced by 99 rams soy bean (N=7.18 grams) and 150 grams water. 17 Mendel and Fine (this Journal, x, p. 345, 1911) found soy bean nitrogen ‘0 be poorly utilized. 68 Rate of Nitrogen Elimination Uncoagulated egg-white (Curve X). On the days when native egg-white was fed the nitrogen out- put was only about half as large as the intake. It is probable, as the discussion to follow will show, that this material was poorly utilized. The character of the nitrogen-output curve with un- coagulated egg-white was somewhat different from that with extracted meat, the maximum excretion occurring during an earlier period—in two experiments during the first three-hour period, Curve X. Comparison of the rates of nitrogen elimination on diets containing extracted meat and liquid egg-white, respectively. N gm. 0.6 oo. "Standard Diet” a i 225 em. meat (N= 8.04 gm.) OS) : 55" lara : ; 35" ousreae . 0.4 Experiment D XXXIII, meat of ‘Standard Diet’’ replaced by 60 grams extracted meat (N=8.05 grams) and 170 grams water. Experiment D XXXII, meat and water of ‘Standard Diet’ replaced by 412 cc. liquid egg-white (N =8,03 grams).'* ania * Diarrhea during 5th three-hour period; urine contaminated, Thi ingestion of dried native egg-white was also followed by diarrhea, a ad y Lafayette B. Mendel and Robert C. Lewis 69 instead of during the second; in a third experiment during the second period, instead of during the third. The results with liquid egg-white and dried egg-white, respectively, were concor- dant. Coagulated egg-white (Curve XI). Coagulated egg-white was evidently well utilized; but the nitro- gen-output curve after its ingestion was more flattened than that when extracted meat was fed. In other words there was a com- parative delay in the elimination of nitrogen when coagulated egg-white constituted the protein intake. Curve XI. Comparison of the rates of nitrogen elimination on diets | containing extracted meat and coagulated egg-white, respectively. gm. "Standard Diet" Pt on 225 gn. meat (N= GimeneeD 0.6 | 55" lard 35" sucrose a =~ 6 NaCl " bone ash water . a b6 BP RR Ws 24 hrs. ----- Experiment D XXXIII, meat of ‘Standard Diet’’ replaced by 60 grams extracted meat (N=8.05 grams) and 170 grams water. Experiment D XXXVII, meat of ‘‘Standard Diet’’ replaced by 419 grams coagulated egg-white (N=8.04 grams). 70 Rate of Nitrogen Elimination Ovalbumin® (Curve XII). When ovalbumin was fed the urinary nitrogen-output was much smaller than the intake, the marked diarrhea about nine hours after the meal suggesting a poor utilization of the ovalbumin as the cause of the smaller nitrogen excretion. , Furthermore, the rate of nitrogen elimination with this material was quite different from that with extracted meat. There was a rise to a maximum in the second three-hour period, then a slight fall during the third three hours, followed by a second rise to a maximum in the fifth Curve XII. Comparison of the rates of nitrogen elimination on diets containing extracted meat and ovalbumin, respectively: - > N "Standard Diet" gm. 200 gm. meat (N = 6.84 gm. 50 " lard Z0-* sucrose nue eam Neal Nacl . hy bone ash — --4 260 be water 3 6 9 te As 24 hrs ----- Experiment H XV, meat of ‘Standard Diet” replaced by 51 grams extracted meat (N =6.84 grams ). Experiment H XVII, meat of “Standard Diet’’ replaced by 50 grams ovalbumin (N=6,73 grams).?° 1” We are greatly indebted to Mr. R. L. Kahn for performing this and several other experiments of the present series. 2° Voluminous diarrhea during third three-hour period; animal ate some feces, Lafayette B. Mendel and Robert C. Lewis 71 period. The fact that the animal ate a portion of the diarrheal feces at the beginning of the third period may account in part for the second rise. DISCUSSION. The nitrogen-output curves following the ingestion of unaltered meat and extracted meat powder, respectively, were quite differ- ent. The slower rate of elimination of nitrogen when extracted meat was fed cannot be explained by the dry condition of this material; for the nitrogen-output curve was unchanged by a pre- vious drying of the meat of the “Standard Diet.” The absence of extractives in the extracted meat can only account in part for the delayed nitrogen excretion. This product presumably contains proportionately more connective tissue than the fresh meat used and thus is digested more slowly. The nitrogen-output curves following the ingestion of casein and ovovitellin are practically identical with that of extracted meat; the character of the curves with edestin, “‘Glidine,” and gelatin is the same as that with fresh meat. . Soy bean, egg-white, and isolated ovalbumin gave nitrogen- output curves radically different from those of either of the meat products studied. The comparative delay in nitrogen elimination independent of the poor utilization of soy bean may be explained _in part by a greater difficulty of digestion of this product, and in part by the presence of sucrose in soy bean, it having been shown in a previous paper that the presence of carbohydrate in the diet delays nitrogen excretion. The comparatively small excretion of nitrogen after the ingestion of native egg-white and ovalbumin is caused in all probability by a poor utilization of these materials, the early diarrhea following their ingestion making such an ex- planation quite likely. That uncoagulated ovalbumin is poorly __ utilized was reported by Falta (1906), who found that the coefficient of utilization of this material in man was only about 70 per cent. Wolf (1912b) fed a large quantity of native egg-white to man and reported that only about half was utilized. When liquid egg-white” was fed it is probable that very little gastric proteolysis occurred; 21 Mendel and Lewis: this Journal, xvi, p. 37, 1913. 2 The dried egg-white mixed with water would be essentially the same as the natural product. 72 Rate of Nitrogen Elimination for Cannon," and London and Sulima,” working with cats and dogs, respectively, have reported that this material begins to leave the stomach almost immediately after ingestion. The early dis- charge of the stomach, the comparatively early emptying of the bowel, and the unfavorable character of liquid egg-white for the action of digestive enzymes, together with a possible resistance of native protein to digestion, may all contribute to a poor utiliza- tion of this material. Coagulated egg-white was well utilized; but following its ingestion there was a comparative delay in nitro- gen excretion. The slower rate of elimination of nitrogen with this source of protein cannot be accounted for by a delay of gastric discharge; for Cannon, and London and Sulima have demon- strated that egg-white coagulated by heat leaves the stomach more rapidly than most proteins. It is probable that the delayed excretion of nitrogen may be caused in part, at least, by a com- _ difficulty of digestion of coagulated egg-white on account of the compact and impermeable character of fine particles of coaguium. A few reports in the literature are in harmiaal with this view that such changes as do occur in the rate of elimination of nitrogen after the ingestion of different protein materials may be explained by variations in alimentary, rather than metabolic processes. ’Van Slyke and White (1911), using the method of the present — work, demonstrated that different nitrogen-output curves were obtained after feeding various boiled fish meats to a dog; and attributed this result to a variation in the rate of digestion. The validity of such an explanation is made very clear by a comparison of the results obtained by these authors with fresh and salt cod, there being as much difference in the nitrogen-output curves of the two preparations of this one fish as in those of any of the different fish. Vogt (1906) investigated the effect on the rate of elimination of nitrogen of superimposing various proteins in con- siderable quantities on a standard diet, finding that both coagu- lated and uncoagulated ovalbumin caused a delay in nitrogen excretion whereas edestin and a casein preparation (Nutrose) gave a nitrogen-output curve of approximately the same character #1 Cannon: Amer. Journ. of Physiol., xii, p. 387, 1904. *? London and Sulima: Zeitschr. f. physiol. Chem., xlvi, p. 282, 1905, ‘ Lafayette B. Mendel and Robert C. Lewis 73 as meat. This author believed that the delay with ovalbumin was caused by a comparatively slow rate of digestion of this material. Loeb (1911) studied the rate of elimination of nitrogen after replacing about one-half of the meat of a standard diet by another form of protein; and demonstrated that there was only a very slight change when meat and casein, respectively, were fed, although considerable difference existed between the curves of these proteins on the one hand and those of their hydrolyzed prod- ucts on the other. In experiments where the urine was collected only for twelve-hour periods Falta, Grote, and Staehelin (1907) found approximately the same rate of nitrogen excretion with casein as with meat. All of these investigators worked on dogs. Wolf (1912a, 1912b) added various proteins and non-proteins to a “standard” diet in man and collected the urine in hourly periods, studying among other things the rate of elimination of nitrogen... He found little difference in the nitrogen-output curves following — the ingestion of gelatin and plasmon, respectively. With veal, however, there was a somewhat slower rate of nitrogen output. Native egg-white and coagulated egg-white gave results quite similar to those reported in the present paper. Wolf (1912c) also studied the rate of nitrogen elimination in dogs after feeding cooked and raw meat, respectively, obtaining practically identical results in the two cases. In considering the significance-of the results of the present study attention should be given to Falta’s conclusions from his work on the rate of metabolism of proteins. The method of in- vestigation employed by this author was to determine the average daily nitrogen-output of a dog in nitrogen equilibrium for a period of several days, then to superimpose the protein to be studied on the “standard” diet, and to ascertain how long a time was required for a reappearance of an excess of nitrogen in the urine equivalent to the nitrogen of the superimposed material. Falta (1904 and 1906) studied different proteins on man and found that with most of these more than half of the excess nitrogen reappeared on the first day, about three days being required for the entire amount to show up. With casein, for example, about two-thirds of the excess nitrogen reappeared during the first day, and most of the remainder on the second day. A few exceptions occurred, however, 74 Rate of Nitrogen Elimination the most striking being with ovalbumin and ovovitellin. In these cases only about 27 per cent of the excess nitrogen appeared the first day; and five days were required for all to reappear, although all but a very small amount had come out in three days. When coagulated ovalbumin was the added protein no longer time was required for the reappearance of the excess nitrogen than was the case with casein. With dogs the results with ovalbumin and casein were the same as for casein with man. These observa- tions on man were confirmed by Himiiliinen and Helme (1907), who demonstrated that a longer time was required for the reappear- ance of the excess nitrogen with egg-white than with a casein prep- aration (Proton) or roast veal. Cathcart and Green (1913), employing the Falta method of superimposition on man, reported that with egg-white, both coagulated and uncoagulated, only a small part of the extra nitrogen appeared in the urine even after several days. There was little difference in the rate of elimina- Aion of the extra nitrogen after adding veal and gelatin, respec- tively; to the diet, greater differences being obtained with the same sample of gelatin in two experiments where the basal rations varied considerably. Vogt (1906) used Falta’s method of study on dogs and found that a longer time was required for the reappearance of the excess nitrogen when egg-white, both uncoagulated and coagulated, was superimposed on the standard diet than when edestin or a casein preparation, Nutrose, was added. All of these communications are in harmony with that of Graffen- berger (1891), who showed by a somewhat different method that when gelatin or fibrin was superimposed on a standard diet the excess nitrogen reappeared more rapidly than when peptone con- stituted the increased nitrogen intake. From the results of his experiments Falta (1906) has concluded that the longer time required for the reappearance of the excess nitrogen after superimposing ovalbumin on the standard diet was the result of the absorption of comparatively large cleavage prod- ucts of this protein, a longer time being required for the catab- olism of these higher protein residues, Hiimiiliiinen and Helme — (1907) held a similar view; and Levene (1909a, 1909b, 1909¢, 1910) and his co-workers from a series of studies of an entirely different *” Only one experiment with ovovitellin is reported and the author says that no definite conclusion should be drawn from a single experiment. ee —, ———————————=————EEE== cere errr erm ree LT PTS Lafayette B. Mendel and Robert C. Lewis 75 nature likewise came to the conclusion that the higher protein cleavage products are catabolized more slowly than the simple amino-acids. Vogt (1906) was not inclined to favor such an ex- planation, maintaining that a slower rate of digestion and absorp- tion might account in part for the results obtained by him with egg-white, and that certain unknown factors of intermediary metabolism might play a part. Although Falta’s experiments are not directly comparable with those of the present study, yet if one recalls how readily texture of the diet influences the rate of digestion and absorption inde- pendently of the character of the protein, it seems quite likely that Falta’s results were caused in part by a difference in the rate of digestion of the various materials studied. Let us consider what would be the effect of a markedly delayed digestion and absorption in studies of the type that Falta made. In the ex- periments of this author on man the superimposed protein as well as the standard diet was fed in four portions distributed over the day. Under such conditions it is quite likely that absorption of the digestion products of a difficultly digestible protein would not be complete until after the beginning of the second day. It is not surprising, then, that the excess nitrogen eliminated on the day when the protein was added to the diet should be smaller than when the superimposed protein was readily digestible; nor that it should be greater on the following day, the amount of nitro- gen absorbed on this second day being greater than that usual on a normal day. The fact that, when Falta fed a single meal to dogs at the beginning of the day, he obtained a result with oval- bumin similar to that with casein makes such an explanation more probable; for in this case digestion and absorption would certainly be complete during the first day. SUMMARY OF RESULTS WITH PROTEINS. The nitrogen-output curves following the ingestion of meat and extracted meat, respectively, differ considerably, that with the latter product being more flattened. This slower rate of elimina- tion of nitrogen cannot be explained by the dried condition of the extracted meat; and only in part by the absence of extractives in this material. It is suggested that the extracted meat may 76 Rate of Nitrogen Elimination have contained proportionately more connective tissue than the fresh meat used and thus have been less readily digestible. The nitrogen-output curves following the ingestion of most of the proteins studied—casein, ovovitellin, edestin, ‘“Glidine,” gela- tin—differ in character to no greater extent than those obtained by feeding the two meat products employed. With egg-white, ovalbumin, and soy bean, however, curves of a character radically different from that of either of the meats were obtained. These results may be explained to a great extent by a difference in the rate and completeness of digestion and absorption of these mate- rials; while the sucrose in the soy bean will also.account in part for the delay in nitrogen elimination with this product. When these alimentary differences are duly taken into account, the conclu- sion seems justified that proteins do not differ materially in their rate of metabolism. Falta’s conclusions from his work on the rate of protein metabo- lism and the similar conclusions of Hiimiiliinen and Helme, and of Levene and his co-workers were discussed. From the results of the present study it seems quite probable that the findings of these authors may be explained by other factors than an assumed difference in the rate of metabolism of proteins caused by an ab- sorption of larger or smaller cleavage products. The results of the experiments reported in the papers of this series show that apart from the character of the protein ingested a large number of diet factors—the water intake, the presence and nature of indigestible materials in the diet, the amount and character of the carbohydrate fed, and to some extent the pres- ence of fat in the diet—play a réle in modifying the rate of elimi- nation of nitrogen after a meal containing protein. With most of the proteins studied the nitrogen-output curves differed to only a slight extent from one another; and in no case did the nature of the protein have a greater effect on the rate of nitrogen elimina- tion than some of the non-protein diet factors mentioned above. Lafayette B. Mendel and Robert C. Lewis 77 BIBLIOGRAPHY. Catucart and GREEN: 1913, Biochem. Journ., vii, p. 1. Faura: 1904, Deutsch. Arch. f. klin. Med., |xxxi, p. 231;1906, ibid., Ixxxvi, p. 517. Fara, Grote, and STAEHELIN: 1907, Beitr. z. chem. Physiol. u. Path., ix, p. 333. : GRAFFENBERGER: 1891, Zeitschr. f. Biol., xxviii, p. 318. HAMALAINEN and Heume: 1907, Skand. Arch. f. Physiol., xix, p. 182. LEVENE: 1909a (Levene and Kober), Amer. Journ. of Physiol., xxiii, p. 324; 1909b (Levene and Meyer), ibid., xxv, p. 214; 1909¢ (Levin, Manson, and Levene), ibid., xxv, p. 231; 1910 (Carrel, Meyer, and Levene), ibid., xxv, p. 439. Logs: 1911, Zeitschr. f. Biol., lv, p. 167. Van SLYKE and Waite: 1911, Journ. Biol. Chem., ix, p. 219. Voar: 1906, Beitr. z. chem. Physiol. u. Path., viii, p. 409. _ Wor: 1912a, Biochem. Zeitschr., xl, p. 193; 1912b, ibid., xl, p. 234;1912¢c, ibid., xli, p. 111. ae * 5 al THE CARBON DIOXIDE AND OXYGEN CONTENT OF THE BLOOD AFTER CLAMPING THE ABDOMINAL AORTA AND INFERIOR VENA CAVA BELOW THE DIAPHRAGM. By J. R. MURLIN, LEO EDELMANN anp B. KRAMER. (From the Physiological Laboratory of Cornell University ‘Medical College, New York City.) (Received for publication, August 22, 1913.) * In search of experimental support for the over-production theory of diabetes mellitus Porges' and Porges and Salomon® have found that ligation of the abdominal aorta and inferior vena cava just below the diaphragm, both in normal rabbits and in depanereatized dogs, causes a rise in the external*® respiratory quotient. They interpret this result as proof (1) that the organism is dependent upon the liver for its power to oxidize protein and fat, carbohy- drate only or carbohydrate chiefly being oxidized when the liver is excluded; and (2) that the depancreatized animal retains its power to oxidize sugar. A rise in respiratory quotient after this radical interference with the circulation has been confirmed by Rolly,* who, however, finds the rise not at all constant and gives an altogether different explanation. Verzdr® likewise witnessed a sudden change in the R. Q., but a change in the same direction, when the liver was partially excluded by anastomosis of the portal vein with the lower 1 Porges: Biochem. Zeitschr., xxvii, p. 131, 1910. * Porges and Salomon: Jbid, p. 148. ’ The term external R.Q. is used here in order to emphasize the fact that the exchange of gases between blood and outside air does not under all cir- cumstances take place at the same rate as the exchange between blood and tissue. The volumetric relations between the CO, gain of venous blood and O, loss from arterial may be called the internal R.Q. ‘Rolly: Deutsch. Arch. f. klin. Med., ev, p. 494, 1912; Miinch. med. Woch- enschr., 1912, Nos. 22 and 23. 5 Verzdr: Biochem. Zeitschr., xxxiv, p. 52, 1912. 79 80 ° Metabolism after Clamping Abdominal Vessels part of the inferior vena’ cava (Queirolo operation). Fischler and Grafe® ligated the hepatic artery of dogs which, some weeks before, had been successfully operated for the Eck fistula, and found in two out of six cases a distinct rise in the R.Q. Béhm’ reports but a very slight rise “after exclusion of the abdominal organs even in depancreatized dogs.” In ‘all these experiments showing a higher R. Q.° there is, as would be expected, a reduction in the total respiratory exchange, depending in amount upon the kind and amount of tissue excluded from the circulation. The reduction in the absorption of oxygen is greater than that for the elimination of carbon dioxide. Hence the higher R. Q. In other words, after the crucial operation there is, relatively, a greater output of COs, There are the following possible ways of explaining this result: (1) A greater production of CO, with no change in the rate of @limi- nation. If carbohydrate or carbohydrate-like bodies should be oxidized instead of protein or fat, moré CO: would be produced. This is the explanation adopted by the von Noorden school. : (2) Greater elimination of CO, from the blood with no essential change in the rate of production. The influences which may be con- ceived of as driving out more CO. may be (a) chemical or (b) mechanical. If more acids were produced, or if acids produced as usual were not neutralized, after exclusion of the liver, more CO, would be liberated from its combination in the tissues and the blood, and would eseape. This is the explanation adopted by Rolly, and approved of by Fischler and Grafe. Rolly has actually found in the serum of his operated animals a lower de- gree of alkalescence than in that of normal animals and Porges in a recent paper’ has himself shown that acidification of the blood by intravenous infusion of sodium dihydrogen phosphate will raise the respiratory quotient, though not so much as occurred in his earlier experiments. The mechanical factors have not been sufficiently emphasized. *Fischler and Grafe: Deutsch. Arch. f. klin. Med., eviii, p. 516, 1912, 7 Bohm: Zentralbl. f. Physiol., xxvii, p. 120, 19138, * Except one animal which had convulsions in Fischler and Grafe's series. Béhm’s complete paper is not accessible and it is possible that his series may contain other exceptions, * Porges: Biochem. Zeitschr., xivi, p. 1, 1912. : : : J. R. Murlin, Leo Edelmann and B. Kramer 81 Porges, in his original article assumes that any change due to over- ventilation which might result would be equalized in fifteen min- utes. Presumably he means by over-ventilation only exaggerated breathing for he cites in support of his view, the work of Born- stein and Gartzen’® on the effects of over-ventilation by voluntary effort in human subjects, showing that after fifty minutes no more CO. can be pumped out in this manner! He also cites one of his own experiments in which the R. Q. in the second period was slightly higher than in the first period after ligation of the vessels. According to Porges’ view, the quotient in the second period should be smaller if any factor of over-ventilation were operative. Neither of these citations offers any convincing evidence which would exclude the mechanical factors; for in the experiments of Bornstein and Gartzen the circulation was in no way disturbed, while the forced breathing was maximal, and Porges’ own experi- ment proves only that, whatever was the controlling cause of the higher quotient after ligation of the vessels, the conditions were the same in the second period as in the first. Moreover it should be borne in mind that over-ventilation may mean something more than exaggerated breathing: there may be over-aération due solely to a disturbance to the pulmonary circulation. Fischler and Grafe appreciated the possible effect of the dis- turbance to the circulation resulting from the Porges procedure, saying, ‘‘One does not know to what extent the results may be due to the direct consequences of this alteration.’ In the writers’ opinion the work of Fischler and Grafe is suffi- cient refutation of the position taken by the von Noorden school as to the réle of the liver in the metabolism of the food-stuffs. Excluding the liver by ligation of the hepatic artery after Eck fistula did not cause a permanent rise in the respiratory quotient in dogs which survived from six to twenty hours. On the other hand, there is no doubt, in certain cases at least, about the rise of quotient after ligation of -the abdominal aorta and inferior vena cava just below the diaphragm. It remains to give a satisfactory explanation of this phenomenon. It seems almost incredible that the purely mechanical effects of so radical a procedure as one which excludes at a stroke fully 10 Bornstein and Gartzen: Pfliger’s Archiv, cix, p. 628, 1905. 1 Fischler and Grafe: loc. cit., p. 519. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 1 82. Metabolism after Clamping Abdominal Vessels. one-half of the blood and considerably more than one-half of the animal’s weight from the circulation should not have been more seriously considered. What effect would it have on the heart rate, on the blood pressure, on the rate of blood flow through the lungs? Neither Porges nor Porges and Salomon gives any data as to the pulse, blood pressure, rate of respiration or volume of respiration, to say nothing of the gaseous content of the blood before and after ligation of the vessels, and yet the results are presented as proof that, by turning a valve, so to speak, the meta- bolic processes are suddenly changed so that one fuel and one only can now supply the body’s energies! The first and most obvious control which one would think of in connection with so radical a change in the R. Q. would be the gaseous content of the blood. There are numerous experiments in the work of Zuntz,!* Krogh," Barcroft,'* Henderson® and others showing that the O2 and CO: contents of the blood are subject to considerable variations, particularly under operative conditions. It is a commonplace laboratory exercise to clamp off the abdom- inal aorta below the diaphragm and witness the enormous rise in systemic pressure (carotid) which results. Simultaneous clamping of the inferior vena cava with the aorta will likewise produce the rise in systemic pressure. But if the two vessels are not clamped simultaneously, the mechanical effect will depend on the order in which the two are clamped. Clamping the inferior vena cava without clamping the aorta will produce a great fall in blood pres- sure for very obvious reasons. Hence, if the vena cava be clamped even as much as fifteen seconds before the aorta, the rise in carotid pressure is not so great as when the two are clamped together. Vice versa should the aorta be clamped first and even a small interval of time intervene before the vena cava is clamped, the blood from the abdominal viscera will continue to flow into the thorax until the vis a tergo is exhausted and the pressure on clamp- ing the aorta will mount even higher, ofttimes so high as to cause heart failure. These mechanical effects must, of necessity, affect the circula- '* Zuntz, N.: Deutsch. med, Wochenschr., 1892, p. 109. ' Krogh: Skand. Arch. f. Physiol., xxiii, p. 179 et 8eq., 1910, “ Bareroft: Lrgeb. d. Physiol., vii, p. 699, 1908. '* Henderson: Amer. Journ. of Physiol., xxi, p. 126, 1908; xxiv, p. 66, 1909. J. R. Murlin, Leo Edelmann and B. Kramer 83 tion through the lungs, the aération of the blood in other words, and in turn the exchange of oxygen and carbon dioxide between blood and outside air. Suppose the two vessels be clamped exactly at the same moment. If the heart remains competent to empty itself against the increased pressure, the blood will of necessity circulate more rapidly through the lungs, unless the heart compensates by beating more slowly. On the other hand, if the heart be not competent, it may go into fibrillations or beat imperfectly, in which case blood will accumu- late in the lungs, producing passive congestion. The former set of conditions should result in a decrease in the carbon dioxide in the blood because, the blood being exposed more often to the alveolar air, the carbon dioxide has more oppor- tunity to escape. The latter set of conditions should result in an interference with oxygen absorption with or without a decrease in the carbon dioxide. EXPERIMENTAL PART. , Reasoning along these lines, the writers have undertaken to determine to what extent clamping of the abdominal aorta and inferior vena cava would alter the carbon dioxide content of the blood as it leaves the heart (carotid artery). Method. The method of procedure in the earlier, orienting experiments was as follows. Normal dogs were anaesthetized with chloretone. Urethane, which Porges and Porges and Salomon employed, has been avoided because it has been the experience in this laboratory that this drug excites the respiratory center of dogs much more than does chloretone. Morphine has likewise been avoided, except in one experiment, because it tends to increase the CO, in the blood.'® When anaesthesia was fully established, the abdominal incision was made, the abdominal aorta exposed just above the origin of _ the coeliac axis, and hemostatic clamps were adjusted all ready to be closed at a signal. Before clamping, the pulse and respira- tion were usually counted and the control sample of blood was 16 Cushny: Textbook of Pharmacology, 1910, p. 221. 84 Metabolism after Clamping Abdominal Vessels drawn. By these precautions one had knowledge of the condition of the animal just before the crucial operation. Then at a signal the two clamps were closed simultaneously, the closure of the vessels being immediately verified by examination. After the lapse of a varying interval of time, during which the pulse and respiration were counted frequently, the second sample of blood for analysis was drawn in the same manner as the first.!” The blood analyses were made by the chemical method of Hal- dane'® using the apparatus devised by Brodie.’® All determina- tions were made in duplicate. The results of the preliminary experiments on five animals are presented in Table I.?° It is evident, from an examination of these results, that a very great change in the aération of the blood is brought about by the occlusion of these vessels. How serious a matter this change is for the life of the animal is seen in the fact that Dog III survived for only twenty minutes after the obstruc- tion was accomplished. In all probability death was due to failure of the left ventricle. ; Out of the four experiments in which a second analysis of blood was made, three exhibit a marked fall in the carbon dioxide of the ‘arterial blood. Two show, in addition, a material fall in the oxygen content. It is not to be supposed that the blood alone loses carbon dioxide. Arterial blood, containing less than the usual percentage of COs, will carry away CO, by diffusion and this will continue the more rapidly the more the tension in the tissues exceeds that of the 17 The clamps which have been used for obstructing the vena cava are those known to surgeons as gastero-enterostomy clamps, fitted with rubber. Much difficulty has been experienced in placing a clamp on the aorta above the origin of the coeliac axis without serious rupture of the diaphragm and a number of animals were killed prematurely in this way. After this experi- ence, however, it was found that by exposing the coeliac axis itself and just above it applying the clamp in such a way as to include the arcuate fibres about the aorta within the clamp it was possible to effect a complete ob- struction without injury to other structures, Later a heavy wrapping cord was passed about the aorta at this point by means of a ligature carrier. '* Haldane: Journ. of Physiol., xxii, p. 465, 1897-8; Haldane and Bar- croft: Ibid, xxxii, p. 232, 1902. 1° Brodie: Tbid, xxxix, p. 391, 1910. 10 Of, Proe, Soc. Exp. Biol. and Med., x, p. 174, 1918. J. R. Murlin, Leo Edelmann and B. Kramer 85 arterial blood. The total amount of extra carbon dioxide appear- ing in the respiration after the clamps are applied, therefore, will depend upon the amount of the gas stored in the tissues. There is no perfectly satisfactory method of estimating the total carbon dioxide stored in the body at one time; hence it is impossible, from the percentages in the blood, to say just how much would escape in a given time. There is scarcely any doubt, however, that sufficient CO, has escaped from the animal’s body in each of the three experiments, to cause a considerable rise in the respira- tory quotient had it been determined for the period during which the vessels were clamped. Respiration experiments. Proof of the correctness of this view could only be had by repeti- tion of the experiments of Porges accompanied by blood-gas analy- ses. Should the carbon dioxide of the arterial blood fall as in the foregoing experiments, during a respiration period occurring imme- diately after clamping of the vessels, and showing a higher R. Q., the conclusion would be irresistible that the higher quotient was due simply to an alteration in the rate of discharge and not an alteration in the rate of production of this gas. Again if the factor of over-ventilation, in the sense of increased breathing, were a controlling one, a period of exaggerated breathing preceding _ the clamping off of the vessels should nullify the effect of clamping on the respiratory quotient. It was desired also to make blood- pressure determinations before and after clamping to ascertain if possible what change is produced in the rate of flow through the lungs. Method of respiration experiments. The apparatus used was a respiration incubator constructed for the special purpose of studying the respiratory metabolism in new-born infants. It consists of a copper chamber (303276 em.) placed inside a Freas electric incubator," by means of which the chamber can be kept at a constant temperature, and connected *1 The Freas incubator can be purchased from Eimer and Amend, New York. The assemblage of apparatus as used in these experiments will be described in detail soon. 86 Metabolism after Clamping Abdominal Vessels to a small Benedict” respiration machine, by means of which it is ventilated. The chamber will accommodate a dog of 6-10 kgm. The entire cubic contents of the air circuit is about 80 liters and, with the subject inside, the air space is correspondingly less. This fact permits of the determination of the R. Q. by the well-known method of Benedict”* with an unusual degree of accuracy; for the reason that a small variation of temperature or of barometric pressure makes but a slight error in the oxygen determination. By making residual analyses at the beginning and end of each Jespiration period, even these errors may be eliminated. The apparatus has been thoroughly tested by burning alcohol inside it, with results, for the R. Q., very close to the theoretical value (0.666); namely, 0.661, 0.654, 0.662, 0.667. For the blood-pressure readings the pulse-pressure instrument of Dr. C. J. Wiggers,** who very kindly instructed us in its use, was employed. Instead of the usual levers for graphic records, the maximal and minimal pressure tubes were connected directly to long mercury manometers and the maximum and minimum pressures were read off the millimeter scale directly. It was .* necessary .to have unusually long manometers on account of the very great rise in pressure which often, though not always, takes place on obstructing the vessels. In one of the earlier experiments of this serie’ the mercury was blown out by the excessive pressure and some 50 cc. of blood escaped from the carotid artery in the confusion which followed. The pressure readings were satisfac- torily obtained in only two experiments. In order to obtain the true pulse pressure and not the maximo-minimal pressures®* the readings were made while the respirations of the animal were inhibited temporarily by central stimulation of the vagus nerve. All of the animals were anaesthetized with chloretone (usually given by stomach) and were tied lightly on an ordinary laboratory dog-board which had been sawed off to fit the respiration chamber, Once the dog is in the chamber and the latter sealed air-tight * Benedict, F.G.: Amer. Journ. of Physiol., xxiv, p. 345, 1909, * U.S. Dept. of Agriculture, Office of Experiment Stations, Bulletin 175, 1907. “ Wiggers, C.J.: Amer. Journ. of Physiol., xxx, p. 233, 1912. *% Wiggers, C.J.: loc. cit. J. R. Murlin, Leo Edelmann and B. Kramer 87 the respiration experiment begins (after a short preliminary period of 10 to 25 minutes during which the chamber is being ventilated by the Benedict machine) on the second of the minute, by throwing a switch and turning a valve excluding the absorbers. Oxygen is admitted automatically throughout the period, but at the end the pressure is brought to the starting-point pressure by hand. The period was usually one hour in length. I, Experiment on depancreatized dog showing higher R. Q. The first subject of this series was a depancreatized dog (Table II). That the animal was thoroughly diabetic is seen by the R. Q. obtained in two successive hours at the beginning of the experi- ments. The blood sample drawn at 3.12 p.m., just three minutes before clamping the vessels, showed a rather low percentage of both gases. It is well known that in diabetic subjects the CO, tension falls at times to a very low point. Another sample drawn fifteen minutes after clamping however shows both gases still farther reduceed—the carbon dioxide more than 10 per cent. Twenty-five minutes after taking this sample of klood the second respiration experiment began and continued for nearly two hours. The R. Q.s are decidedly increased, although the total respiratory exchange is very much reduced. Accompanying the higher R. Q. is a very great depression in the carbon-dioxide content of the blood. In view of the elevated blood pressure, which continued for twelve minutes at least after clamping the vessels, and the increased percentage of oxygen it seems likely that the true explanation of the lower percentage of CO. and hence of the higher R. Q. in this case is an increased aération of the blood by more rapid cir- culation through the lungs. II. Experiment in which the R. Q. remained the same after clamp- ing the vessels. This was a normal dog, anaesthetized as usual. Blood pressures were determined before the respiration experi- ment. The dog had been fed the day before on dog biscuit, con- taining a high percentage of carbohydrate, and may have eaten some of it left over from the previous day, on the morning of the experiment. The R. Q. is rather high (probably for this reason) the first hour, but fell the second hour to a point more nearly within the range of a true niichtern value. *6 Beddard, Pembrey and Spriggs: Lancet, 1903, I, p. 1366. 88 Metabolism after Clamping Abdominal Vessels Upon clamping, the blood gases fell rapidly within the next fifteen minutes and the blood pressures which were high at first fell rather suddenly to very low levels. The R. Q. did not rise in the second respiration period and the CO: did not fall as in the previous experi- ment. The explanation of the low blood pressure was found at autopsy in the fact that the clamp on the aorta had caught a bit of the stomach and for this reason was not quite competent to hold the arterial pressure. It was possible, after sectioning the aorta below the clamp, to squeeze blood through. The animal therefore must have bled into his abdominal vessels until the arterial pressure reached a level which could no longer pass the obstruction. This experiment proves the relatively greater importance of CO, than of O, in the blood in determining the external R. Q. Oxygen is not stored in the tissues in any quantity as is carbon dioxide; consequently a considerable change in ‘the O, content of the circulating medium does not affect the R. Q. materially. III, Experiments in which the R. Q. fell after clamping the vessels. In the next experiment (Table IV) the dog exhibited a high R. Q. in the first period but instead of falling as is usual the further the time from feeding, it rose. Unfortunately the respiration rate was not recorded during these preliminary periods. It must be supposed however that there had been a considerable over- ventilation of the lungs and a consequent Auspumpung of COs; for upon clamping the vessels there was no fall, to speak of, in the CO; content of the blood within the first twenty minutes, and during the subsequent respiration period the CO, rose to 54.8 per cent while the oxygen fell. Again the CO, proves to be the determining factor; for its rise in the blood denotes a very considerable storage in the tissues and it is the holding back of this CO, which causes the R. Q. to fall to the extremely low level of 0.61 in the second period. The respiration apparatus was thoroughly tested imme- diately after this experiment and proved to be absolutely correct, giving a R. Q. with the alcohol flame of 0.667. The results of the preceding experiments are fully confirmed in the following one (Table V) which was more complete. The dog had had no food since the previous day. The respiration rate was recorded during the preliminary respiration periods and established the cause of the high R. Q.s unquestionably to be the J. R. Murlin, Leo Edelmann and B. Kramer 89 Auspumpung of CO:. The temperature of the respiration chamber during these periods was very close to the critical temperature at which dogs begin to pant. This fact together with a rather light state of anaesthesia probably accounts for the high rate of breathing. The pumping out of CO, in this case was so complete that upon clamping the vessels there was no reduction of the CO, in the blood, but instead a slight rise. In all probability this rise started from a still lower level the moment the dog was removed from the respiration chamber, for the respiratory rate fell at once to normal. In the respiration experiment which followed clamping of the vessels the R. Q., instead of rising, fell to an abnormally low point. The clamps were absolutely competent. Severe congestion of the lungs with oedema was found at autopsy, a circumstance which explains the very low percentage of oxygen found at the end. The carbon dioxide was not so high, however, as in the previous expe iment. The rate of respiration declined rapidly and the dog was near death when removed from the chamber. a. One other experiment, not reported in detail, was performed on a normal fasting dog, in which the R. Q.s in the preliminary periods were 0.72 and 0.85, while after clamping it was 0.67. The same explanations probably apply. Two other experiments on depancreatized dogs were attempted _ but both died upon clamping of the vessels. Porges and Salomon succeeded in obtaining respiratory periods after ligation of the vessels in only four depancreatized dogs out of fifteen. There are obvious reasons why the animals do not survive longer. The strain upon the heart is tremendous. In several dogs, both normal and depancreatized, of this series, the heart failed at once and could not be revived. Aside from this the very rapid fall in the CO, percentage, which cannot be entirely compensated for by reduced rate (see Dog V, Table I) must produce a profound effect on all the higher brain centers. When, added to this, we consider 4 that the congestion of the lungs is such as to interfere with the _ absorption of oxygen, the wonder becomes that so many animals survive as long as they do. 90 ©Metabolism after Clamping Abdominal Vessels Alkalinity of the blood. Rolly?’ has established, by a new and much improved method, the fact that in dogs operated after the Porges procedure, the H-ion concentration of the blood is increased and the OH-ion concentration is diminished. This observation has been confirmed in a single examination of the blood reaction made in these experi- ments. From Dog IX, 20 ce. of carotid blood were drawn (10 ce. into each of two centrifuge tubes containing 0.5 cc. each of 0.1 per cent hirudin solution) before clamping the vessels and again just after drawing the last sample of blood for gas analysis. Ten ec. of the hirudin plasma titrated to the first pink color of phenol- phthalein with 4; NaOH required for the first sample 4.2 ec. and for the second 7 ce. The acidity, in other words, had nearly doubled, and yet in spite of this change the CO, was held back coincident- ly so as to reduce the R. Q. to 0.633! From this single observa- tion the indications are that this greater acidity (H-ion concen- tration) cannot be the only cause of the extra elimination of COs. DISCUSSION OF THE FACTOR OF EXAGGERATED BREATHING. This series of experiments was undertaken in the full expecta- tion of finding mechanical factors adequate to explain any altera- tion in the R. Q. which could result from sudden obstruction of the main vessels leading to and from the abdominal organs One such factor, exaggerated breathing, unquestionably is; for in these experiments it has been shown (Dogs VIII and IX) that increased respiratory activity may keep the quotient far above normal for at least two hours. That over-ventilation (exaggerated breathing) was present in the experiments of Porges and of Porges and Salomon may be inferred, in the absence of direct data, from the expressed assumption of Porges that after fifteen minutes of exaggerated breathing no more CO, could be pumped out. Fur- thermore it is the experience of this laboratory that urethane, which Porges and Salomon used, always excites the respiratory center (in dogs) and that it cannot always be controlled with moderate doses of morphine. In a former series of experiments in which the respiration apparatus was attached directly to the *7 Rolly; Minch. med. Wochenschr., 1912, Nos, 22 and 23, J. R. Murlin, Leo Edelmann and B. Kramer 1 trachea* urethane was tried and was given up for this very reason. In the original experiments of Porges and of Porges and Salomon, the higher quotients are doubtless due in part to this form of over- ventilation. DISCUSSION OF THE FACTOR OF BLOOD FLOW THROUGH THE LUNGS. That some other factor than exaggerated breathing may account for a great reduction in the CO, of the arterial blood and there- fore for a rise in the respiratory quotient after clamping of the ves- sels, is seen from the experiments with Dogs I, II and V (Table I) and Dog VI (Table II). In none of these experiments was any in- creased breathing observed. The blood-pressure determination with Dog VI gave a clue which it was hoped would lead to definite conclusions on the matter of blood-flow when the pulse pressures were more accurately determined in the experiments with Dogs VII and IX. Unluckily the leak in Experiment VII invalidated the blood-pressure findings, as a criterion of blood-flow in that experiment; for the mean pressure changed. In Experiment IX however it may be seen that the minute volume of blood-flow through the heart, and therefore through*the lungs has changed greatly after clamping and that this change is consistent with the change in blood gases. According to the law of von Recklinghausen®® the amplitude of the pulse wave (pulse-pressure) at any given mean pressure is a measure of the systolic output, provided the distensibility of the arterial wall is constant. The product of the pulse-pressure | by the pulse-frequency is then a measure of the minute-volume. There is no reason to suppose that the distensibility coefficient per se of the arterial system is in any way altered by the clamping of the aorta and vena cava. Therefore if the mean pressure remains about the same the product of pulse-pressure into pulse- frequency would afford a criterion of the effect of the operation on the blood-flow. Referring to Table V it is seen that the pulse pressure just before clamping is three and one-half times as great as just after clamping. The mean pressure hasrisen slightly but not sufficiently 28 Murlin and Greer: Amer. Journ. of Physiol., xxvii, p. xviii, 1911. 29v. Recklinghausen: Arch. f. exp. Path. u. Pharm., lvi, p. 1, 1906. 92 Metabolism after Clamping Abdominal Vessels to offset the difference in pulse pressure.*° The pulse frequency is considerably higher before the operation than after it. Hence the blood-flow through the lungs has been greatly reduced by the operation. The surprising thing is that such a change in the blood-flow should not have produced a greater effect on the ex- change of gases. Two facts then stand out with some significance in the matter of blood-flow. In Experiment VI where the CO, in the blood fell rapidly after clamping of the vessels (while the O, rose), and the R. Q. as a consequence rose, the pulse pressure was maintained. Since there is no reason to believe that the pulse rate suffered any diminution (see Experiments I-V), the minute volume after clamp- ing was at least as great as before. In Experiment IX where the CO, in the blood rose slightly (while the O, fell) and the R. Q. as a consequence was falling (after the previous over-ventilation) the minute volume was distinctly less. These two facts are offered not as final proof but as evidence, consistent as far as it goes, that R 8° y. Recklinghausen’s formula is A = ——_. X 1/k where A is ampli- . er) an dp/ p tude or pulse pressure, R is pulse volume, the expression (2), denotes distensibility of the arterial wall, at the mean pressure and k is a constant determined by viscosity, diameter of vessels, etc. The pulse A (distensibility) I/k ; Making substitutions from Table V the pulse volume before clamping 22 x distensibility at 54 I/k volume R then would be expressed by the formula would be ; after clamping it would be 6 X distensibility at 72 “it Vk ’ Supposing the distensibility and the value of k to be the same the pulse volume before clamping is more than three times the value after clamping. The minute volume would be found by multiplying the value of the pulse- volume, or systolic output, by the pulse-frequency. Taking 210 as pulse- frequency just before clamping and 180 just after, the minute volume proves to be less than one-third its former value. In all probability this differ- ence is too great; the point is to show that distensibility or the value of k would have to change a great deal to offset the difference in pulse pres- sure observed. J. R. Murlin, Leo Edelmann and B. Kramer 93 the altered rate of blood-flow through the lungs is an important factor in determining the CO, (and O2) content of the blood and therefore in explaining the altered respiratory exchange. CONCLUSION. Whether one or both of the factors discussed above are control- ‘ing, there can be no doubt as to the significance of the blood-gas analyses. In each instance the blood-gas changes are consistent with the mechanical explanation of the altered respiratory quo- tients after clamping the vessels. Where the R. Q. rose (Experi- ment VI) the CO, of the blood fell; where the quotient remained stationary (Experiment VII), the CO, did not change; and where the R. Q. fell (Experiments VIII and IX), the CO, in the arterial blood rose. Clamping off the blood from the abdominal organs therefore does not alter the character of the metabolism, and the © experiments of Porges and of Porges and Salomon have no bear- ing on the problem of the oxidation of sugar. . 94 Metabolism after Clamping Abdominal Vessels TABLE I. Dog. I. 8Skgm. March 22, 1913. Chloretone per rectum. BLOOD TIME EVENT PULSE bagi Se Oz CO: p.m. per cent| per cent 3.20 | 4.3 cc. carotid blood drawn 15.48 438.6 3.25 | Vessels clamped simultaneously 3.30 96 36 3.40 144 30 3.45 138 30 3.54 | 4.2 cc. carotid blood drawn ' 15.52) 22.53 3.55 | Clamps removed 4.00 120 35 Dog IIT. 12kgm. April 10,1913. Chloretone intraperitoneally. 2.15 | 4.4 cc. carotid blood drawn 17.10) 38.35 2.17 | Vessels clamped simultaneously 66 | 35 2.20 120 24 2.35 % 102 |, 12 2.40 ue 102 30 2.41 | 3.2 ce. carotid blood drawn 17.16) 37.47 2.47 | Clamps removed ht Dog III. 7.5kgm. April 12, 1913. Chloretone anaesthesia. 2.08 108 | 30 2.20 114 24 2.30 114 24 2.35 120 | 15 2.36 | 4.5 cc. carotid blood drawn 18.85, 42.01 2.40 120 | 24 2.46 | Vessels clamped simultaneously; heart stopped 2.58 24 2.59 | Artificial respiration ' 3.00 96 | 30 3.06 120 | 54 3.10 | Dog died; clamps on only 20 minutes; cause of death not apparent ane - — en — J. R. Murlin, Leo Edelmann and B. Kramer 95 TABLE I.—Continued. Dog IV. 9kgm. April 19, 1913. Chloretone by stomach. BLOOD TIME i EVENT PULSE * ante 6 vara Oz COz p.m. per cent) per cent 1.50 138 35 2.25 — | 120 72 ‘2.32 | 4.4 cc. carotid blood drawn 138 66 | 19.52) 39.42 2.38 132 60 2.40 | Vessels clamped simultaneously 2.43 104 96 2.52 126 64 3.02 120 78 8.15 120 80 3.25 120 72 3.35 126 72 3.43 | 4.35 cc. carotid blood drawn 17.09) 24.28 3.44 | Clamps removed 3.45 120 60 | Dog V. 10 kgm. May 10, 1913. Morphine subcutaneously. Chloretone by stomach. 2.238 96 34 3.15 120 32 3.17 | 4.15 ce. carotid blood drawn 13.32 51.06 3.20 | Vessels clamped simultaneously 3.22 102 16 3.27 120 16 3.33 120 14 3.48 120 14 4.03 120 14 4.13 120 18 4.18 - 120 16 4.22 | 4.3 ec. carotid blood drawn 11.60, 34.16 4.23 | Clamps removed 120 | 24 4.29 108 24 4.39 108 24 Metabolism after Clamping Abdominal Vessels ee sSuny jo one oa -u0d ou {zyuazeduI00 sdureya ‘Asdoyny — 90°9 86° ZT OL ET UMBIP POOTq 99 OF ) cre F'0f z8°0 jusunsiedxe uojesdsay 8686p S-Co F 0°08 | €8°0 quourtiedxe uonyendsay oo Foo € _—- snqeredde uoreiidser 07 powinqey ces 10'¥% (98°6 x UMBIP POO] “99 EEF og’ € sz | zit | 6 | 08 Be ney %% |901 | 26 | 08 0z° "Paes PLI | O&T LV€ GLI | SFI 9l¢ ; Apenoouss nuns podure]o sjesse 4 cre GL FE 90° IT #8 99 UABIp poolq “99 9° zie 0s OL oo€ snyeivdde oinssoid Poold jo juoweZuv1e pu uoT}TsedG | 0g -08 I 39°0 snyeivdde uonemdser uy = 2Z 1-22 1 mad 89°0 snyeivdde uornesndses Uy | 22° ZI-22 IT juss sad 7090 40d |< “we 709 | 40 ‘cm | 38 | “OOK % “az |. 3A eg ate ugd INGAS ani SISA/IVNV ag1od aunssaud ' aoo1d aoola } oa 96 “"yaDUo}s hq auoja.oqy) “Sie ‘6 Ane quaunrsdazg “grel ‘4 Ane pozivossundeg “mby Tg “JA boq es ‘TT TEV ; —_ a OFI | OE Ajsnoouez[nuis podureyo sjosso 4 ¢0' 1 tL | +9 00°T = SE" LE gL SZ —— plyo1v9 WIOIJ POo]q °99 FF 0G ZI & 6 | 92 | | 801 A][wijUoo poye[NUys sNBCA 4JoryT 8h SI S 96 os Gel A][B1jue0 pozeNuirys sn3va 4yjory LY ZI v; +6 | O08 OST ATyeazues poye[nurnys snBva 4jor] OF ZI : 0S | SL : T€° Or ~ ry Awojyoredery 1a ZI oO ued & 6 'ZE |Z18"0 |690'F \Le9°F OF-S8Z quourtiodxe uolyedsey | ce ZI-OF IT ae 9 8E /P98"0 |468°€ 0E9"F 8I quourtiedxe uoryerdsoy | OF TI-So' Ol a snjeredde uoivaidses ut pooetg of OL 5 OIL | 98 Ze AT[V1}WIO poye|NUIys SNFVA 4 jor] 62'0T FE one OIL | +6 921 A]jwaques poyepnurrys sn3vA 4 Jor] 82;0l 7 P21 | SOI 9Z1 Ajjesyuso poye[nuirys sn3vAa 4yjory zor 8 (x) 91 OTT | 2 } zel | Zit . 96° O1 fi Bs rel | OTT : ’ GOl 4 OT | OFT | sn3va 4jo] ND wor 8 a jail ——- SNBVA 4J9] POL, wOL 3 = Z| OST pryores posodxsy oz or 8 = *Q ‘bap sunsb | swpsb |yuao sad |7U90 sod as uD g . oar — | 00 | to | ‘xen | “UN $ pe y — bf “ava “asa SISATVNV auassaud este Gee pip seer ; ban aoola aoo1d 8 “yoowoys fig auojatoyy) “SI6I ‘6e Ain ‘jowson “why gg “TIA S0q g ‘II AITAVL fed o ” n vo a ‘aginsseid poojq ul S [18j surejdxq -yve] WyF1s poygruned > e durejo vyi08 ynq “juezeduI00 durejo E | eavo vusa pomoys Asdoyny -q3¥9q ore = ve | O08 ese a ce | 8% cel woe < 9% | 0 PPI I¢Z op leh €% [88° FT UMBIP POO] PI}OIVd °99 ZF os'Z oy : snzeivdde w01j poaouiey Crs Ss LSE |bI8'O COEF TIS F 2-08 quountssdxe uonesdsey | OF Z-22 1 = snjyeiedde uornvidses 07 pouinjoy SII oO IT's 28°61 UMBIP Poo] Prj}0IBd “99 Z°¢ | cll 2 ve | 8% | OC Al[vsqu99 poyE[NUnys sNBwvAa 3jo"] rd oa | v - gg | #9 A]jesyues poye[nunys snsva 330] | 801 “a 0) eee sunsb | sups0 yuad sad quad sad i ud ; 5 od io uno. | | ee oO F pect O°U |-aviam) “asa SISA'IVNV auoassaud Raat feos wai ere 2 aoo1a aoo1a : © ‘ponuyu0g—TIT AIAVL = ie 98 99 J. R. Murlin, Leo Edelmann and B. Kramer in pe sZuny] jo uo1sesuo0o [31s {yuezodwi09 sdurvjo pomoys Asdoyne { pop oq 00°F 08 #o | TL ad UMBIP POOTY 99 g'E oo ¢ 9°6z | 19°0 | S0'T | 88°0 |hG-8S quounsedxe uoryesidsoy | 09° -99°% 0°08 \602°0 |FIE T 082° T #9 m7 quounts0dxe uolzeiidsey | ¢g gZ-Ge' TL snyeivdde 07 pouinjzory eet 6° €P Sh 02 UMBIP POOlq *99 FF oe T cS | OFT Lz 1 89 | OFT att Ajshooueynuns pedurv[o sjosse 4 Or I&' FF |22° &% UMBIP POO|q *99 OF so'T 9¢ | $I co’ 1 Sh | él 6h ZI poyeiedo oq OF ZI ‘ud T' TE |[F6'0 992° 698° F quouriedxe uoryesidsoy | Cg’ ZI-S¢" IT 0 1g |198°0 960° ZIS"F quoursodxe uoryestdsey | og TI-S€ OT ‘OD bap supi6 | suvsb uae sad yUua0 sad “uD hd ‘09 0 “NIN “NIA Oca | 20, | 00, | = susaivNY uaa | uaa INGAG awIL *InaL aoo 1g “d8au | Ge10d arlLouyvo “fyoouonsadsjur auonesony) “s16t ‘ge Amp “youson “why og “IITA 5d ‘AI ATAVL ce |ost | ae 100 Metabolism after Clamping Abdominal Vessels C2 ss gngeaedde uonraidses yo yno Soq | ¢o°z Ge | FO'T (869° F \ez¢°9 : potied puoses jo pug | 0° FIL 2 et OL CRI 9L cel 89 ‘i hee iim ree | 96'0 Sze F |10L'S potied ysug jo pug | 101 CF So zr Ce st pe}18ys poriod ysi1q | £0 ZI Ss “urd 00°Z1 snyeindde uoneidses ut 30q | OS IT 9 8 Ajyeaques pozepnuiys snsva yyArRy | ST IT 89 GP Aj[viques pezepnuitys sn3va 4y3ny | ZT II OL | OF qno snzea 4qary | ZT I 09 | OS snjvivdde ainsseid poojq YIM UOr}eUUOD | OT TT Cm a promo pesodxg | 1¢°01 If OI *O “bap sunib | suns \ua9 sad \yuad sad : “xu uw ; : ? asa ae “een 109 *O9 me) Ww WwW IK ‘NIK sh SISA'IVNV auossaud d8au | asiod “aNGL aoo1d — aoo18 ‘yoouoys fig auojaLojy “ST6T ‘Tasnény -jouion “why og “XI Sd : “A TIAVL Be , a IOI J. R. Murlin, Leo Edelmann and B. Kramer 6 1€ €89°0 Isls 568° T 1G LE oF 4So] cL Ol ve 92 92 co OL OL cv gREe c61 OST — O8T OIG quajeduios Ajaitjue sdure[o ‘saqoy] 10M0] yjoq jo vurapoo pomoys Asdoyny ‘perp soq HO®PN {99 0'Z poyvszty Buse d “00 OL A1QOUIT[VY]V 10J UMBAP POOT PI}OIBd *99 OZ UMBIP POO|G PIjOIvd “90 Gf snjvivdde uoresidses jo yno 30q spus potied uoreidses pay y, suiseq pored pary.y, snyeredde uoresdsei ut 30q UMBIP POO[ PIjOIvd “99 G*F A][BAQUeD pozyEpNUIIys SNBvA 4jorT Aj[eajueo poye[nuys snsva 4yjory A[[B1}U99 poye[NUITZs SNSVA 4jo'T Ajsnoouv}[nuis poedureys sjasse A HO®N & °99 Sh Pores}} °99 OT £480} AIJOULI[VY]S 1OJ UMBIP POOT PI}OIVd *99 OZ UMBIP POO][G PIJOIvO “99 FE pezye[NUITys SNSBA 4JorT at poyeNuiys pus yNd snBVA 4Jo'T PERRSSRIRRSSS SF NN AN OD OD OD OD OD OD SH SH ~~ 33S NAN ARRE NANANN THE SEPARATION OF ¢d-ALANINE AND d-VALINE. By P. A. LEVENE anp DONALD D. VAN SLYKE, (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, August 23, 1913.) In the ester method of protein hydrolysis the esterifiable amino- acids are separated by distillation into two fractions, a higher boiling containing aspartic and glutaminic acids, phenyl alanine, and serine, and a lower boiling fraction containing proline, /-leu- cine, d-isoleucine, d-valine, d-alanine, and glycocoll. For several years we have been trying to devise methods to approximate as nearly as possible a quantitative separation or determination of the six amino-acids composing the latter mixture. Proline, unlike the other members of this fraction, is very sol- uble in alcohol,’ and is partially separated from them by alco- holic extraction. The extract,,however, usually consists of about two-thirds proline and one-third of a mixture of other amino- acids which have gone with the proline into solution in the alcohol. - Proline, however, contains no primary amino nitrogen, while all the nitrogen of the other acids of this ester fraction is in the form of primary amino groups. Therefore, a determination of the total and the primary amino nitrogen,’ respectively, in the extract per- mit one to calculate accurately the amount of proline, which is indicated by that of the non-amino nitrogen. The other five amino-acids can be distributed by fractional crys- tallization among subfractions the composition of which varies , greatly according to the proportions in which the different acids are present. As glycocoll and alanine dissolve at room tempera- ture in only four parts of water, while the other three, particularly 1 Fischer: Ber. d..deutsch. chem. Gesellsch., xxxix, p. 530, 1906. *Van Slyke: Quantitative Determination of Proline obtained by the 4! Ester Method in Protein Hydrolysis, this Journal, ix, p. 205, 1911; Quan- titative Determination of Aliphatic Amino Groups, this Journal, ix, p. 185, 1911 and xii, p. 275, 1912. 103 104 Separation of d-Alanine and d-Valine the leucine and isoleucine, are much less soluble, one can usually obtain by crystallization the greater part of the mixture in two fractions, a comparatively insoluble one consisting of the leucine and isoleucine, together with much of the valine, and a very solu- ble fraction containing glycocoll and alanine. For the quantitative determination of the proportions in which leucine, isoleucine, and valine are present in the less soluble fraction we have already published methods which have been utilized with “satisfactory results.’ More recently we have described the separation of gly- cocoll from alanine in the more soluble subfraction by means of glycocoll picrate, which is difficultly soluble in cold water.‘ Besides the leucine-isoleucine-valine and the glycocoll-alanine crystallized fractions, however, one usually obtains another, in- termediate between these two, containing alanine and valine in such proportions that they cannot be separated by crystalliza- tion. This paper presents a method for the separation of the alanine and valine of this intermediate fraction. One can now determine all the six amino-acids from the lower boiling ester fraction with a fair degree of accuracy. This does not mean that they are completely regained in the amounts in which they are present in the proteins. Losses which prevent this still occur in the esterification and distillation “of the esters. The uncertain- ties, however, which were formerly connected with the separation — of these amino-acids after the distillation, are now reduced comparatively small proportions. We have determined the following data, on which is based the method for separating valine from alanine, and from glyeocoll in — case this also should occur in the intermediate fraction. Data on which the separation is based, d-Alanine in the presence of 10 per cent sulphurie acid is pre- cipitated by phosphotungstic acid as a crystalline salt which contains approximately 14 parts of phosphotungstic acid to 1 of | * Levene and Van Slyke: this Journal, vi, p. 391, 1909. Abderhalden and — Weil have recently isolated from nerve tissue a third leucine isomer. We — did not find evidence of it in casein or edestin; but if it proves to be a general constituent of the proteins still further development of special methods for this fraction will be necessary, Zeitschr. f. ec rat Chem., Ixxxiv, p. 39, 1913. ‘Levene and Van Slyke: this Journal, xii, p, 285, 1912. P. A. Levene and D. D. Van Slyke 105 alanine. At 0° about twenty-four hours are required for precipi- tation of the maximum amount of alanine. The presence in solu- tion of about 20 grams of phosphotungstie acid (in excess of the amount precipitated with the alanine) per 100 cc. of solution is required to insure most complete precipitation. Under these conditions the amount of alanine left in solution at 0° in 100 cc. of mother liquor is 0.15 gram. The concentration of free phos- photungstic acid can be increased up to at least 70 grams per 100 ec. of solution without either increasing or diminishing to a sig- nificant extent the solubility of alanine phosphotungstate. d-Valine has under the same conditions the much greater solu- bility of 1.2 grams per 100 cc. Valine phosphotungstate shows, under proper conditions, very little tendency to form mixed crys- tals with alanine phosphotungstate. In case a mixture of the two is obtained, one can readily separate them by reerystallization from a solution containing 10 per cent of sulphurie and 20 per cent of phosphotungstie acid. The solubilities of the phosphotungstates of both alanine and valine are very dependent upon the concentration of sulphuric acid present. Glycocoll is precipitated under the same conditions as d-alanine, only 0.2 gram of glycocoll remaining in 100 ce. of mother liquor. Lead acetate, recently recommended by Benedict and Murlin® for the removal of phosphotungstic acid from solutions containing amino-acids, is the most satisfactory reagent which we have found for freeing both alanine and valine from sulphuric and phospho- tungstic acids. The precipitation of phosphotungstic acid is quan- titative, and the small amount of lead sulphate remaining dis- solved in the filtrate is reatlily removed by addition of an equal volume of alcohol. Five per cent, and sometimes even more, of the amino-acid present are usually adsorbed by the heavy pre- cipitate, but the loss is less than when barium hydrate is used, and the amino-acid regained after removing the excess lead as sulphide and concentrating the solution to dryness contains less than 1 per cent of ash. Natural leucine is precipitated by concentrated solutions of phosphotungstic acid, the precipitate being redissolved by suffi- cient excess of the acid, as found by Levene and Beatty. Leucine 5 Proc. Soc. Exp. Biol. and Med., 1912. 106 Separation of d-Alanine and d-Valine may interfere with the purification of alanine as the phospho- tungstate, however, and should be removed, either by crystalliza- tion or by precipitation as the lead salt® before the separation described below is begun. Dilute methyl] and ethyl alcohol are unsuitable solvents for the recrystallization of valine when even a small proportion of alanine is present; because the relative solubilities of the two amino-acids in water are reversed in both alcohols, in which alanine is much less soluble than valine. This is the case to a less marked extent with acetone, and it is, therefore, better suited to throw valine out of water solution in the presence of alanine. If to 100 ec. of water at 20° one adds 200 ce. of 80 per cent acetone, the resulting solution will dissolve 3.2 grams of alanine and 3.4 of valine. The solu- bility relations are such that one can add 3, 4, 5, 6, or 7 volumes of 80 per cent acetone with nearly the same effect. A mixture of 100 cc. of water and 700 ce. of 80 per cent acetone dissolves 2.5 grams of alanirle and 3.4 of valine. Consequently, as the results are within a wide range independent of the volume of solu- tion added, 80 per cent acetone affords a convenient means for throwing valine out of water solution in the presence of small amounts of alanine. Because of the fact that alanine is much less soluble than valine in ethyl and methyl] alcohol, especially the latter, it was thought that valine could, perhaps, be extracted from a mixture of the two amino-acids by means of methyl alcohol. It was found, however, — that it was impossible to extract all the valine without also dis- * solving a large proportion of the alanine. Precipitation and purification of alanine as phosphotungstate. The mixture of valine and alanine should preferably contain not over 50 per cent of valine. If more is present, part can readily be removed by recrystallizing from water, in which valine is much less soluble than alanine. It is advisable, because of the appreciable solubility of alanine phosphotungstate, to precipitate it from as small a volume of 10 per cent sulphuric acid as will hold the valine in solution. In order to obtain ut once alanine phosphotungstate free from valingy * Levene and Van Slyke: this Journal, vi, p. 391, 1909. bi eral. P. A. Levene and D. D. Van Slyke 107 the volume of solution must be as great as 100 cc. for each gram of valine present. If the alanine phosphotungstate is recrystallized, however, one need use but 30 to 40 ec. for each gram of valine, recrystallizing once from a similar volume of fresh solution. One thus completes the separation, using in all only 60 to 80 per cent of the volume of solution required when one does not recrystallize, and one is also somewhat more certain of the absolute purity of the alanine. The process which gives the most satisfactory separa- tion is the following: The mixture of alanine and valine is dissolved in a hot solution which contains 10 grams of sulphuric acid per 100 cc. The volume of this 10 per cent sulphuric acid used should be 30-40 ec. for each gram of valine which analysis of the mixture indicates can, as a maximum, be present. In the hot solution one further dissolves enough purified phosphotungstic acid to combine in the ratio of 14:1 with the maximum amount of alanine which previous anal- ysis has indicated can be present in the mixture, and in addition leave 1 gram of excess phosphotungstic acid for every 5 ee. of the 10 per cent sulphuric acid used. The use of a greater excess of phosphotungstic acid does not interfere with the separation, but leaves one an unnecessarily large amount to remove at the end of the operation. The solution prepared as above directed is placed in a refrigerator at 0° and allowed to remain there for at least ___ twenty-four hours.’ In case the volume of the solution is large, time must be allowed for it to cool before beginning to count the period allowed for crystallization. The precipitate separates in large, transparent crystals, which form a solid layer about the walls and bottom of the flask. When sufficient time has been allowed for the separation, the supernatant solution is decanted off as completely as possiblé. The crystals are then redissolved by heating with a volume of 10 per cent sulphuric acid equal to that originally used. Phosphotungstic acid, in the ratio of 1 gram to each 4 or 5 cc. of 10 per cent sulphuric acid used, is then dissolved in the hot solution, and the alanine phosphotungstate is again allowed twenty-four hours at 0° to crystallize. The supernatant solution is again decanted, and the crystals are washed with suction with a small volume of an ice-cold solution containing 10 per cent of sulphuric and 20 per cent of phosphotungstic acid. 7 If only an ordinary ice box, which usually gives a temperature of 8°, is available, the flask should be immersed in ice water. 108 Separation of d-Alanine and d-Valine Determination and isolation of the precipitated alanine. The alanine phosphotungstate is at once dissolved in hot water, where it forms a solution that is usually somewhat turbid. It is diluted in a measuring flask to such a volume that 10 cc. contain from 50 to 100 mgms. of alanine, and aliquot parts are used for determination of the nitrogen present. The determination is most conveniently performed by the nitrous acid method for determina- tion of amino nitrogen.’ If the micro-apparatus (cf. p. 121) is used 2 ce. of solution are sufficient; with the larger apparatus one uses 10 cc. The determination can also be done according to Kjeldahl, — although in this case it is necessary to draw air through the mix- _ ture, while it is digesting with sulphuric acid, in order to prevent the violent bumping which the precipitated tungstic acid causes.° It is preferable to base the calculation of the amount of alanine present on the nitrogen determination rather than on the sub- stance actually isolated, because, when the phosphotungstic acid is removed with lead, the bulky precipitate of lead phosphotung- state adsorbs several per centof the alanine present, and the amount actually recovered is only 90-95 per cent ofthat in solution before the removal of the mineral acids. To the amount of alanine calculated from the nitrogen determination one may add a solu- bility correction for the amount dissolved in the total volume of solution from which the alanine was precipitated and recrystal- — lized. This amount is calculated on the basis of a solubility of — 0.15 gram of alanine per 100 ce. The remainder of the solution, after the portion for the analysis has been removed, is washed into a Jena beaker and heated to boiling. A 20 per cent solution of neutral lead acetate is added in portions until an excess is present, and can be detected, by means of the sulphuric acid test, in a drop removed from the sur- face of the solution in the beaker. The heavy precipitate of lead sulphate and phosphotungstate is filtered with suction and washed thoroughly with water. The filtrate is concentrated to a volume * This Journal, ix, p. 185, 1911, and xii, p. 275, 1912. As there is so much mineral acid present, it is advisable to add, to the nitrous acid solution in the apparatus, enough 4 or 5 Nn NaOH to nearly neutralize the sulphuric acid before the solution containing the latter is run in. * Denis: this Journal, viii, p. 427. bir a P. A. Levene and D. D. Van Slyke 109 of about 50 ec. for each gram of alanine present, and mixed with an equal volume of 95 per cent alcohol. This precipitates a small amount of lead sulphate which had remained, owing to its slight but appreciable solubility in water. The solution is allowed to stand on the water bath for an hour or more to complete the pre- cipitation, the sulphate is filtered off, and the excess of lead in the filtrate is removed with hydrogen sulphide. The lead sulphide is washed with water through which H.S has been bubbled, and the filtrate is concentrated, preferably in vacuum, to a small volume. It is then transferred to a Jena glass evaporating dish and the concentration continued on the water bath until all the visible liquid has been evaporated. The drying is completed in a vacuum desiccator over sulphuric acid and potassium hydrate. It is not advisable to try to drive off with heat the last traces of water and acetic acid, for this is likely to somewhat discolor the substance. The product, dried in vacuum, is perfectly colorless, nearly ash- free (if pure reagents have been used), and free from valine. Besides the alanine isolated as above described, a small amount, left in solution when the alanine phosphotungstate was precipitated, is later obtained from the mother liquors of the valine. In case the original valine-alanine mixture contained glycocoll, the latter will now be found with the alanine, from which it can be separated as the picrate, according to the method described by us.?? Determination and isolation of the valine. The decanted filtrates and the washings from the alanine phos- photungstate are diluted to a definite volume and the amino nitro- gen determined in an aliquot part, in the manner described for the alanine solution. A special blank determination to ascertain the correction for the reagents should be made, using as the con- trol solution 10 per cent sulphuric acid instead of water, as the presence of so much mineral acid increases the correction. The phosphotungstic and sulphuric acids are removed with lead acetate, as in the isolation of alanine, and the valine solution, free from mineral acids and bases, is concentrated on the water bath until the valine begins to erystallize at the surface. Two or three 10 Levene and Van Slyke: this Journal, xii, p. 285, 1912. 110 Separation of d-Alanine and d-Valine volumes of 80 per cent acetone are then added, and the mixture is rinsed, using more 80 per cent acetone, into a flask. This is stoppered to prevent evaporation of the acetone, and allowed to stand over night while the valine crystallizes. The latter is filtered, washed with 80 per cent acetone, and thus obtained free from alanine in a yield of 80 to 85 per cent of the amount present. The filtrate from the valine contains the small amount of ala- nine which escaped precipitation by phosphotungstic acid, and an amount, usually about equal, of valine, which remained in solution in the dilute acetone. The filtrate is concentrated to dryness, weighed, and the alanine and valine separated with phosphotung- stic acid as before. This second crystallization makes the separa- tion practically quantitative. When refrigeration facilities do not enable one to keep the solu- tions at 0° during the entire period while the alanine is being pre- cipitated, one can let the solutions stand over night at room temperature, and then place them in ice water for several hours, stirring them occasionally to complete the crystallization at 0°. The precipitation is nearly, though not quite, so complete as when the solution is kept at 0° for the entire period. Working at room temperature entirely, one can precipitate at, least 75 per cent of the alanine in purity, using one-half the vol- ume of solutions given in the above directions. Purity of reagents. a Because of the large amounts of lead acetate and phospho- tungstic acid used, both reagents must be pure or the amino- acids obtained after their use will be accompanied by ash. The lead acetate should leave no residue after precipitation of a solu- tion with hydrogen sulphide and evaporation of the filtrate to dryness. We have had no difficulty in obtaining good lead acetate from the manufacturers. The phosphotungstic acid should leave no residue after precipitation with pure lead acetate and evapora- tion of the filtrate. We,purify the commercial phosphotungstic acid by Winterstein’s method. ‘The acid is dissolved in a small amount of water, from which itis shaken out with ether. With the latter it forms an oily solution much heavier than water. The ether solution is washed several times with water, and the ether is driven off on the water bath. The produet is not hydroscopie, and forms a colorless solution, P. A. Levene and D. D. Van Slyke Ii EXPERIMENTAL. Analysis of materials. d-Alanine was obtained from hydrolyzed silk by the ester method. The glycocoll accompanying the alanine in the amino-acids ob- tained from the low boiling fraction of esters was removed with picric acid," and the d-alanine was purified by recrystallization from dilute alcohol. It gave the following figures on analysis. Substance, 0.1195 gram; CO:, 0.1764 gram; H2O, 0.0825 gram. Substance, 0.0909 gram; nitrogen gas at 21°, 763 mm. (nitrous acid method), 25.20 cc. Substance, 0.1817 gram; solution (containing 1.3 mols. HCl), 2.4910 grams; concentration, 7.29 per cent; sp.gr., 1.03; rotation in 2 dm. tube with yellow light from a spectroscope, +2.07° +0.01°. Substance, 0.1422 gram; solution in 20 per cent HCl, 2.5810 grams; con- centration, 5.51 per cent; sp. gr., 1.087 at 25°; rotation in 2 dm. tube, +1.64° *(0.01°. Calculated for Found CesH70.:N: Se RUS 40.25 40.41 Be... a ae es 7.42 7.92 TS See 15.74 15.73 [a], with 1.3 mols. HCl.. .. +9.77° +10.30° (Calculated for HCl salt. )!? [a], with 1.3 mols. HCl.. ..+13.78° Calculated for amino- acid. [a] with 20 per cent HCl. .. +9.72° Calculated for HCl salt. [a]> with 20 per cent HCl. . +13.69° Calculated for amino- acid. From the above figures it is apparent that the d-alanine was analytically pure and as free from dl-alanine as one can usually prepare it from hydrolyzed protein. The rotation is, as stated by Emil Fischer, practically unaffected by the amount of excess hydro- chloric acid present. d-Valine was prepared from casein by esterification and the use of our lead method. The preparation gave the following figures on analysis: _ . Substance, 0.1203 gram; COs, 0.2260 gram; HO:, 0.1013 gram. Substance, 0.1081 gram; nitrogen, 22.9 cc. at 25°, 762 mm. (nitrous acid method). 1 Levene and Van Slyke: this Journal, xii, p. 285, 1912. 2 E. Fischer: Ber. d. deutsch. chem. Gesellsch., xxxix, p. 464. 8 Levene and Van Slyke: this Journal, vi, p. 391, 1909. 112 Separation of d-Alanine and d-Valine Substance, 0.1510 gram; solution in 20 per cent HCl, 2.6240 grams; sp. gr., 1.10; rotation in 2 dm. tube with yellow light, +3.28° +0.01°. Calculated for Found: CsHi102N: ora ecko esc cnc Once 0 bcs eee pi 21 51.24 lt Lp II 9.43 9 .47 |. ee a eee eh 11.97 11.96 MR +25 .93° +28 .80° The valine was analytically pure. The rotation was lower than that obtained by Fischer for synthetic d-valine,“ but is as high as one usually obtains in the natural product after acid hydrolysis. As the valine obtained by acid hydrolysis of proteins usually has a rotation of +24° to +26°, the use of the above material gives one more nearly the conditions actually met in hydrolysis work than would employment of the optically pure synthetic substance. Composition of alanine phosphotungstate. Levene and Beatty found that alanine combines.with phospho- tungstic acid to form a crystalline salt." We have prepared the salt as nearly pure as possible in order to determine its composi- tion. Preliminary preparations showed that the ratio of alanine to phosphotungstic acid was approximately 1:14. We dissolved the two constituents in this ratio (0.5 gram of alanine and 7 grams of phosphotungstie acid) in 15 cc. of normal hydrochloric _ acid, and let the solution stand over night while the salt crystal- lized. The crystals were filtered on a clay plate and dried over solid potassium hydrate until the chloride reaction disappeared. The product was further dried in a vacuum at 100°. The pro- portion of alanine was then determined by estimation of the amino nitrogen with nitrous acid. The results were: a RRS. SS “D 1.036 per cent. PEED, +. cots 5 MEMES + 6 SMEs ss MME ov aEWe 6.57 per cent. RRR SSSR OB SRR SA 93.33 per cent. pemeumlaning: PT Ags... .uss+-cctuesscddaes =1:14,1 The salt forms with water of crystallization, The air-dried substance loses 3.8 per cent of its weight when dried in vacuum at 100°, and the anhydrous salt when exposed to air takes up a sim- 4 Ber, d. deutach. chem, Gesellach., xxxix, p. 2320. * Levene and Beatty: Zeilschr. f. physiol, Chem., xlvii, p. 149, 1906, " P. A. Levene and D. D. Van Slyke 113 ilar weight of moisture. This corresponds to approximately 3 molecules of water for 1 of alanine, the ratio, 1 alanine: 3 H,O, requiring 3.99 per cent water. Solubility of d-alanine phosphotungstate in varying concentrations of sulphuric acid. Solutions each containing 0.250 gram of d-alanine, 5 grams of phosphotungstic acid, and varying amounts of sulphuric acid were made up to 10 cc. volume and left at 0° for forty-eight hours. The solutions were then decanted through dry filter papers into the 10 cc. burette of the aminometer (apparatus for determination of amino nitrogen) described in this Journal, xii, p. 275. The nitro- gen in the measured volume of filtrate was determined by the nitrous acid method, and from the result the amount of alanine present in 100 cc. of filtrate was calculated. The percentages of sulphuric acid indicate grams per 100 ce. of solution. TABLE I. CONCENTRATION He2SO« | ALANINE IN 100 cc. OF FILTRATE per cent grams 3 0.56 4 0.38 5 0.36 6 0.30 8 0.19 10 0.14 10 0.15 12 0.16 14 0.18 16 0.18 As 0.250 gram of alanine combines with 3.5 grams of phospho- tungstic acid, the excess of the latter in solution was 1.5 grams, or 15 grams per 100 ec. ‘The above table indicates that, in the - presence of this excess of phosphotungstic acid, sulphuric acid decreases the solubility of alanine phosphotungstate, the maximum effect of the sulphuric acid being exerted in 10 per cent concen- tration. Under these conditions the solubility of alanine at 0° is only 1 gram per 700 ec. of solution. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 1. 114 | Separation of d-Alanine and d-Valine Effect of the concentration of free phosphotungstic acid on the solu- bility of d-alanine phosphotungstate in 10 per cent sulphuric acid at 0°. Portions of 50 mgm. of d-alanine were dissolved in 5 ce. each of 20 per cent sulphuric acid in test tubes, and varying amounts of a solution containing 2 grams of phosphotungstic acid per cubic — centimeter were added to the different solutions, all of which were then made up to 10 cc. with water allowed to stand thirty hours at 0°. The amounts of alanine remaining in solution were then determined as described in the preceding section. The excess phosphotungstic acid was estimated by subtracting from the amount added the 0.7 gram combining with 0.05 gram of alanine, TABLE II. P PTA ADDED PER 100 cc. ee ee mii ae sindhond = grams . grams 15 8 0.22 20 13 0.21 25 18 0.18 30 >. 2 0.15 40 Bh 33 0.14 60 53 0.15 80 73 0.13 It is evident that about 20 per cent of free, excess phospho- tungstic acid in solution insures a maximum precipitation of the alanine at 0°. At 20°, in the presence of 20 per cent phospho- tungstic acid solution, the solubility is 0.38 gram per 100 ce. Time required for the precipitation of d-alanine phosphotungstate at 0°. Portions of 0.05 gram of d-alanine were dissolved with 3 grams of phosphotungstie acid in 10 ce. of 10 per cent sulphuric acid, — The solutions were left at 0° for varying periods, at the end of — which they were decanted through dry filters, as in the solubility determinations described in the preceding sections, and the nitrogen remaining in solution was determined, . — : : | / r P. A. Levene and D. D. Van Slyke 115 TABLE III. TIME ALLOWED FOR PRECIPITATION ALANINE IN 100 cc. FILTRATE hours grams 3 0.21 6 : 0.20 15 0.17 22 0.16 40 0.14 While the greater part of the alanine is precipitated in three hours, over twenty are required for the complete attainment of solubility equilibrium. Solubility of dl-alanine in 10 per cent sulphuric acid containing varying concentrations of phosphotungstie acid. The results in the following table show that the phosphotung- — state of dl-alanine is more than twice as soluble at 0° that of d-alanine. The conditions of the solubility tests were t e same as those of the foregoing experiment. | TABLE IV. PTA apogp vem 100 cc, | Bxoms PTA wuesr7 ome ie grams grams 10 3 0.43 20 13 0.35 30 23 0.35 60 53 0.37 80 73 0.37 Solubility of d-valine phosphotungstate in varying concentrations of sulphuric acid at 0°. Portions of 0.4 gram of valine were dissolved with 6 grams of phosphotungstic acid each in 5 cc. of 2, 4, 6, 8, 10, and 12 per cent sulphuric acid respectively. The amount of phosphotungstic acid was found by a separate experiment to be a sufficient excess to depress the solubility of the valine to its minimum. The solu- tions were cooled to 0° and kept at that temperature for three days. The solubilities of the valine were then determined as in 116 Separation of d-Alanine and d-Valine the similar experiments with alanine. The solution with only 2 per cent of sulphuric acid showed no precipitate. The others showed crystalline precipitates varying in bulk with the concen- tration of the sulphuric acid. The percentages of sulphuric acid indicate grams per 100 ce. TABLE V. HeSO. VALINE IN 100 cc. OF FILTRATE per cent ; grams 4 we 4 4.95 6 2.78 8 1.87 10 1.21 12 0.88 At 20° the solubility in 10 per cent sulphuric acid in the presence of an excess of phosphotungstic acid is 3.4 grams per 100 ce. Solubility of valine and alanine in varying concentrations of acetone. As stated before, acetone was found a better agent than methyl or ethyl aleohol for throwing valine out of solution in the presence of the small proportions of alanine that escape precipitation with the main crop of alanine phosphotungstate. To ascertain the optimum proportion of acetone to add to the water solution of valine in order to cause it to crystallize most completely with- out carrying down alanine also, the solubilities of the two amino- acids in varying concentrations of acetone were determined at 20°. Fifteen cubic centimeters of the solvent were in each case shaken two hours with an excess of amino-acid, and 10 ee. of the filtered solution evaporated in a weighed dish. TABLE VI. ACETONE | satr a IN ee ie ant IN a taal cent | y grams grams 100 0,008 0.002 90 0,028 0.012 80 0.164 0.097 : 66.7 0,560 0.402 50 1,290 1.315 we ee P. A. Levene and D. D. Van Slyke 117 The following table shows that when 80 per cent acetone, in the ratio of from 2 to 7 volumes, is added to 1 volume of water, the solvent power of the water for alanine and valine is reduced to a point which remains nearly the same, whether 2, 3, 4, 5, 6, or 7 volumes of the 80 per cent acetone are added. The decrease in solubility caused by increasing the percentage of acetone is ap- proximately compensated by the increase in volume. TABLE VII. AMINO-ACID DISSOLVED IN MaceronE ACETONE IN MUON MIXTURE. q ag a ro 100 cc... | THE MIXTURE ; OF WATER Alanine Valine Alanine Valine cc. per cent grams grams | grams grams 200 53.3 1.08 1.16 3.24 3.48 300 60.0 | 0.71 0.85 2 84 3.40 400 64.0 052 | 0.67 2.60 3.35 500 66.7 0.40 0.56 | 3.40 3.36 600 68.6 0.35 0.48 | 2.45 3.36 700 70.0 01 | 04 | 2.48 3.44 The solubilities in the third column were graphically inter- polated from those given in the preceding table. Separation of a mixture of d-valine and d-alanine. The following separation serves as an example of the applica- tion of the method. One gram each of d-valine and d-alanine was dissolved in 35 ce. of hot 10 per cent sulphuric acid (prepared by diluting 10 grams of acid to 100 cc.) with 23 grams of purified phosphotungstie acid. The solution was allowed to stand till it had cooled to room tem- perature, and was then placed in a refrigerator at 0° for twenty- four hours. The crystals which had separated formed a solid layer about the walls and bottom of the flask. The supernatant liquid was decanted off, and the crystals were redissolved on the _ water bath with 35 cc. of fresh 10 per cent sulphuric acid. Eight grams of phosphotungstiec acid were then dissolved in the hot solu- _ tion, which was cooled and placed in the refrigerator for twenty- four hours as before. The mother liquors were again decanted off, and the crystals were quickly washed on a suction funnel with 118 Separation of d-Alanine and d-Valine several small portions of a solution containing 10 grams of sul- phurie acid and 20 grams of phosphotungstic per 100 cc., the washing solution being at a temperature of 0°. Alanine. The crystals were transferred as completely as pos- sible with a spatula from the funnel to a Jena beaker. A small residue adhering to the funnel and filter paper was washed into the beaker with hot water, and the flask in which the crystals had formed was also washed out with hot water, in order to obtain a few crystals of alanine phosphotungstate which the previous washing had not removed to the funnel. Enough water was added to the alanine phosphotungstate to bring the volume to 75-100 ec., and the beaker was covered and heated on the water bath until the crystals were dissolved to a slightly turbid solution. The latter was transferred to a 150 ec. measuring flask and diluted to the mark. Two cubic centimeters of the solution used for deter- mination of amino nitrogen in the micro-apparatus gave 3.37 cc. of nitrogen gas at 25°, 758 mm., indicating 0.1398 gram of nitrogen, or 0.889 gram of alanine in the entire solution. The remaining 148 cc. of solution were treated as described on pp. 108 and 109 to remove phosphotungstic and sulphuric acids. Thealanine regained weighed 0.83 gram, and gave the following figures on analysis. Substance, 0.1222 gram; ash, 0.0013 gram; substance, ash-free, 0. 1209 gram; CO:, 0.1777 gram; H,O, 0.0855 gram. Rotation in 20 per cent HCl: Substance, 0.1029 gram=0.1016 aslo solution, 1.9060 grams; concentration, 5.33 per cent; sp. gr., 1.1; rotation in 1 dm. tube, +0.80°. Calculated for Found: d-alanine ee ae OR 40.10 40.40 ae Sa COU Oe Oe 7.92 7.92 NE RR a +13.7° +0,2° +13.7° It is evident that the precipitate consisted of pure alanine phos- photungstate. The correction for the solubility of alanine‘as phos- photungstate in 70.ce. of solution under the conditions of precipi- tation and recrystallization is 0.70 * 0.15 = 0.105 gram of alanine. Adding this to the 0.889 gram precipitated gives 0:994 gram of alanine found to be present out of the 1 gram originally added, Valine. The filtrate and washings from the alanine phospho- tungstate were diluted to 150 ee. and 2 ce. of the solution taken -_—_---- P, A. Levene and D. D. Van Slyke 11g: for determination of amino nitrogen. The nitrogen obtained measured 3.28 cc. at 25°, 758 mm., indicating 0.1362 gram of nitrogen in the entire solution. This is equivalent to 1.001 gram valine besides the 0.105 gram of alanine which, according to the solubility of alanine phosphotungstate, should be present. The remaining 148 cc. of solution were freed from sulphuric and phos- photungstic acids with lead acetate and concentrated, first in vacuum, then in a Jena glass dish on the water bath, until valine began to crystallize at the surface. About 3 volumes of 80 per cent acetone were stirred into the hot solution, which was then transferred, with the aid of more 80 per cent acetone, to an Erlen- meyer flask. The flask was stoppered and the valine allowed to crystallize in the ice box. The crystals, washed with 80 per cent acetone, weighed 0.78 gram, and gave the following analytical figures. Analysis: Substance, 0.1198 gram (no ash); COs, 0.2258 gram; H,0, 0.1108 gram. : p Rotation in 20 per cent HCl: Substance, 0.0811 gram; solution, 1.291 grams}; concentration, 6.28 per cent; sp. gr., 1.1; rotation in 1 dm. tube, +1.78° +0.01°. ‘ . Calculated for Y Found: d-valine: Ci cnuall cae oe +a 51.38 51.24 cel ES 9.41 9.47 Fl A: "9 Ca a +25 .8° +25.9° The filtrate from the above crop of valine was concentrated to dryness, taken up with 25 ec. of 10 per cent sulphuric acid, and the alanine precipitated with 6.5 grams of phosphotungstic acid. The precipitate was dissolved in hot water and the solution di- luted to 50 ce. Two cubic centimeters gave 0.52 ce. of nitrogen, equivalent to 0.05 gram of alanine in the entire solution. This precipitation could have been made a little more complete if it had beé@n performed in the same manner as the first, using only 7 or 8 ce. of solution instead of 25, and recrysgallizing once. The filtrate from the alanine phosphotungstate was also brought to 50 cc. and 2 cc. taken for a determination of amino nitrogen, which yielded 1.89 cc. of gas at 24°, 766 mm., equivalent to 0.0266 gram of nitrogen in the entire solution. This indicates, besides the 0.04 gram of alanine soluble in the 25 ce. of solution from which it was precipitated, 0.17 gram of valine. When the solution had 120 Separation of d-Alanine and d-Valine been freed from mineral acids and the product crystallized from dilute acetone, 0.13 gram of analytically pure valine was obtained. Analysis: Substance, 0.1040 gram; ash, 0.9011 gram; substance, ash- free, 0.1029 gram; CO:, 0.1935 gram; H2O, 0.0896 gram. Calculated for : Found: CsHi102N: irae a ik oo oe oiosne <>, 058) Selene Riche. 51.28 51.24 i: CS 8 yo AIRES 9.74 9.47 The total amount of analytically pure valine regained was 0.91 gram, or, making allowance for the portions of solution removed for nitrogen determination, 0.93 gram. The amount present, as calculated from the nitrogen content of the filtrate from the alanine, was 1.001 grams. The loss of 0.07 gram in isolation is due partly to loss in crystallization, partly to adsorption by the heavy lead precipitates formed when the mineral acids are removed, these precipitates always adsorbing a few per cent of the amino- acid present. CONCLUSION. d-Alanine combines with phosphotungstic acid*in the ratio of : approximately 1:14 by weight, forming a crystalline salt. At 0°, — in a solution containing, per 100 cc., 20 grams or more of phospho- tungstic acid in excess of the amount combining with the alanine, — and 10 grams of sulphuric acid, the solubility of alanine is only — 0.15 gram per 100 ce. The solubility of d-valine under the same — conditions is 1.21 grams per 100 ce. By alternate crystallization — of valine as the free amino-acid and of alanine as the phospho- — tungstate, one can effect a practically quantitative separation of a mixture of the two amino-acids, J Sate oF re gE THE GASOMETRIC DETERMINATION OF ALIPHATIC AMINO NITROGEN IN MINUTE QUANTITIES. By DONALD D. VAN SLYKE. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, August 23, 1913.) In a previous number of this Journal! we have described an improved apparatus for the determination of amino nitrogen by the nitrous acid reaction. Its chief advantages over the form originally described? by us lay in its ability to be used an indefinite number of times without separating any of the parts, and in the fact that it permitted all the shaking to be done by a motor. By merely reducing its size this form of apparatus can be given an accuracy which brings the determination within the class of micro- methods. The gas burette of the micro-apparatus holds 10 ce. The upper part, measuring the first 2 cc., is of only 4 mm. diamete tage and is divided into sy ce. divisions. The remainder is wider, and is divided into twentieths. In order to keep the correction neces- sary for the reagents small, it is preferable that the amounts of the latter should be reduced in proportion to the volume of nitro- gen obtained for measurement. The deaminizing bulb is, there- fore, of only 11 to 12 ec. content, and the 10 cc. burette on the larger apparatus is replaced by one of 2 cc. capacity. Only 10 1 This Journal, xii, p. 275, 1912. 2 Toid., ix, p. 185, 1911. The appsratus is designed only for use with a motor. It can be obtained from Emil Greiner, 45 Cliff Street, New York, with motor for either direct or alternating street current, or from Robert Goetze, Leipzig. The substance in the nitrite which gives the small amount of gas obtained on blank determinations we have never been able to identify or remove. As a matter of fact, while some brands of commercial grades of nitrite are entirely unsuitable, others give results as good as those obtained with the most high priced ‘‘reagent”’ or ‘‘azur Analyse’”’ preparations. For the last _ two years we have used the ordinary grade supplied by the Powers-Weight- man-Rosengarten Company with uniformly good results. 121 122 Determination of Amino Nitrogen ee. of nitrite solution and 2.5 ec. of acetic acid are required for an analysis, and the correction for the reagents is 0.06 to 0.12 ce., according to the quality of the nitrite employed. The same size of modified Hempel pipette (see previous article) can be used for the small as for the large apparatus, and, because of the small amounts of nitric oxide absorbed, it lasts for an almost indefinite number of analyses without change of the permanganate solu- tion. With the micro-apparatus the error need not be more than 0.005 mgm. of nitrogen when 2 ee. or less of gas are measured, or 0.01 mgm. when more is obtained. Consequently one can analyze one-fifth the amount of substance required for the larger apparatus without reducing the percentage accuracy. The advantages of the micro-apparatus are: (1) It requires only 0.5 mgm. of amino nitrogen for an analysis accurate to within — 1 per cent. (2) It uses up relatively small amounts of reagents. & (3) Having shorter dimensions and being of equally thick glass, it is relatively stronger than the larger apparatus. In point of rapidity the little apparatus has, if anything, a slight advantage - ovet the large one. On a warm day we have made as many as _#en accurate analyses per hour with the former. The minimum ’ time required for the quantitative evolution of the nitrogen of a amino-adids in the thoroughly shaken apparatus is, at 15° to 20°, five to four minutes; at 20° to 25°, three minutes; at 25° to 30°, two and a half to two minutes. Because of its conveniences, we now use for physiological work the smaller apparatus almost — exclusively. , Practically the only alteration from the mode of operation, already detailed in the previous description of the larger appara- tus, is in the speeds at which the deaminizing bulb and the Hempel pipette are shaken. During the first stage of the analysis*® the deaminizing bulb should be shaken by the motor at very high rate of speed, about as fast as the eye can follow, or an unneces- sary amount of time is lost in freeing the apparatus from air. This stage is also much accelerated by warming the nitrite solution to 30° before it is used, in case a low room temperature has reduced the temperature of the solutions below 20°. In the third stage* when the nitrie oxide is being absorbed by the permanganate, the * This Journal, xii, p. 279, 1912. ‘ Thid., p. 280. Donald D. Van Slyke 123 Hempel pipette should be shaken not faster than twice per second. Absorption is approximately as fast as when more vigorous shak- ing is used, and the latter is likely to break off from the residual gas small bubbles, which stick under the nearly horizontal upper side of the pipette and escape being drawn back into the gas burette for measurement. Because of the small amount of nitrogen to be measured, it is especially necessary that in the first stage the removal of the air should be complete. This is assured by shaking the solution in the deaminizing bulb back each time, in this stage, until the bulb is two-thirds filled with nitric oxide. One point in setting up the apparatus appears to require espe- cial emphasis. The hook or wire loop from which the deaminizing ‘bulb is suspended* should be perfectly rigid and hold the capillary . outlet tube tightly. Otherwise the rapid shaking which is ad- vantageous becomes, instead of a smooth vibration, a rattle, disagreeable to the operator and dangerous to the tus. Binding the tube to the holder with a strip of rubber band is a satisfactory method of insuring a firmly held apparatus. The entire apparatus can be cleaned most conveniently by fill- ing the burettes and deaminizing bulb with dichromate-sulphuric acid mixture. When the apparatus is in daily use it is a good practice to let it stand regularly over night filled with the cleaning | mixture, _ Two points in which every apparatus should be tested, as soon > as it is set up, are the accuracy of the burettes and the tightness of the stopeocks. The two burettes are calibrated by weighing the water which they deliver; and the cocks are tested for their ability to remain air-tight when subjected to the suction or pres- sure of a column of water a meter high. For most work, the solutions for’analysis can be measured off with sufficient accuracy in the 2 cc. burette on the side of the deaminizing vessel. When especially accurate results are desired, however, one uses an Ostwald pipette, calibrated to deliver 1 or 2 ce. within 0.001 or 0.002 cc. respectively, and washes the burette twice, with six or seven drops of water distributed about the entire inner walls of the burette for each washing. 5 See photograph, this Journal, xii, p. 277. 124 Determination of Amino Nitrogen The following results were obtained on four successive analyses from solutions measured in this manner, and illustrate fairly the accuracy which one can attain with the method. For each anal- ysis 2 cc. of a 1 per cent solution of Kahlbaum’s leucine were used, the amount of leucine being therefore 20 mgms: No. N gas Soeen'? PRESSURE | N EVOLVED | CALCULATED ERROR ce. deg. C. mm. mgms. mgms. mgms. 1 3.75 | 20 762 2.140 2.138 +0.002 2 3.74 20 762 2.133 2.138 —0.005 3 3.74 | 20 762 2.133 2.138 —0.005 4 3.77 | 21 762 2.141 2.138 +0.003 The following results, obtained with solutions measured from the 2-cc. burette, indicate that, when the latter is clean and the delivery careful, it gives nearly as consistent results as a pipette. For each analysis 2 cc. of a #$ solution of alanine were taken. The time allowed for the reaction was two and a half minutes. No. N Gas apr cong PRESSURE | N EVOLVED CALCULATED . ERROR cc. deg.C. | mm. mgms. mgms, mgms. 1 2.56 28 760 1.398 1.40 —0,003 2 2.58 28 760 1.409 1.401 +0.008 _ 3 2.57 28 760 1.403 1.401 +0.002 4 2.57 28 760 1.403 1.401 +0.002 IMPROVED METHODS IN THE GASOMETRIC DETER- MINATION OF FREE AND CONJUGATED AMINO- ACID NITROGEN IN THE URINE. By DONALD D. VAN SLYKE. (From the Laboratories of the Rockefeller Institute for Medical Research New York.) (Received for publication, August -23, 1913.) Total amino-acid nitrogen. Our original method for the determination of the total amino- acid nitrogen! (free plus conjugated), although it gave accurate results and has been used successfully by ourselves and others, required a somewhat cumbersome manipulation before the urines were ready for the final amino determination. After being acidi- fied and heated in the autoclave to hydrolyze the urea, the ammonia was boiled off on a hot plate with lime, a process requiring careful watching for an hour or more, and rather offensive because of the odors eVolved. The calcium sulphate and hydrate were then filtered, the filtration and washing requiring another hour. The washings, of about 500 ce. volume, were then concentrated on the water bath, which required two or more hours additional. These manipulations have been greatly simplified by the ascer- tainment of the fact that one has merely to filter off the alkaline solution obtained after adding the lime, and concentrate the fil- trate on the water bath to dryness, in order to drive off every trace of ammonia. This process dispenses entirely with the troublesome boiling off of the ammonia. The washing of the "i _ precipitate of calcium salts can also be avoided to advantage by making the mixture up to a definite volume before filtering, and taking an aliquot portion of the filtrate for the rest of the deter- mination. All the operations are furthermore rendered more con- venient by the use of the micro-apparatus for determining the . 1 Levene and Van Slyke: This Journal, ‘xii, p. 301, 1912. 125 126 Determination of Amino Nitrogen in Urine amino nitrogen, which permits one to work with relatively small volumes of liquid, and yet have sufficient material for duplicates. The present method is the following: 25 cc. of urine are mixed with 1 ec. of concentrated sulphuric acid and heated in an auto- clave at 180° (oil bath temperature) for one and a half hours. The solution is then transferred to a 50-ce. flask and 2 grams of powdered calcium hydrate are added. The mixture is thoroughly shaken, made up to 50 ec., and filtered through a dry folded filter. Twenty cubic centimeters of the filtrate are measured into a Jena glass evaporating dish and concentrated to dryness on the water bath, the process of concentration requiring about a half hour. The residue is moistened with 1 ee. of 50 per cent acetic acid to bring the calcium hydrate and carbonate into solution, and is then washed into a 10 ce. flask and filled up to the mark. One can either use the entire solution for determination of the amino nitro- gen in the large amino apparatus, or use 2-cc. portions for the micro-apparatus. The length of time which the nitrous acid solution should be shaken in order to drive off all the amino nitrogen depends somewhat on the temperature. When the latter is 15—20° the time should be five to four minutes; for 20-25° it is three minutes; for 25-30°, two and a half to two minutes. It is preferable that the solution should be shaken vigorously with a motor and the time kept down tothese lim- its, for the sake not only of rapidity butof accuracy. The reason for i this is, that, even after removal of the ammonia and urea, urines contain small amounts of substances which belong to the class of slowly reacting amines, and are therefore not a-amino-acids, The correction for this nitrogen can be ascertained in the same manner as the urea correction in amino determination on the blood,? by continuing the reaction, after the gas from the amino- acids has all been driven off, for a length of time equal to that utilized in decomposing the amino-acids (two to five minutes, according to the temperature), and then measuring the nitrogen that has been evolved during this second reaction-period. The correction is so small and constant, however, amounting to 0.2-0.3 per cent of the urine nitrogen, that it will for most work be found unnecessary to take it into account. * Van Slyke and Meyer: ‘Phis Journal, xii, p, 402, 1912. Donald D. Van Slyke 127 The following results were obtained with normal human urines. For the final determination, 2 cc. of solution, equivalent to 2 cc. of urine, were used in the micro-apparatus. The temperature was 25°, the pressure 758 mm. for all measurements. TABLE I. Total amino nitrogen (free and conjugated). ] g PER CENT OF é ghEg | "So cremmm = | TRAN mt roy or NO. A 3 “ Baa) pee we =| U z orr se] 2 | BEET | vetch | Camino} Retaa | Cometea grams ce. mgms. mgms. | "Wears 1 1.211 Pie 3.06 | 31 28.4 | 2.57 | 2.35 1:18 | 1.04 | 31.9 | 28.1 f oa68 | 2.32 | 2 1.750 1.38 1.24 37.5 | 33.7 2.14 1.93 1.30 26 | 37.7 | 34.3 |) el 1.96 8 | 0.883 | 1.19 | 1.08 | 32.4 | 20.3 | 3.80MN 8.52 19 1 32.4 | 29.8 3.89 | 3.46 4°) 4747 | 1.8990" 1.15 | (36.0 | 312.4 areee 80 1.30.) 1.15 |) 35.3 | 31.4 | 2.02 | 1.80 5 | 4.809 | 1.10 | 0.97 | 31.0 | 27.0 | 2.37 | 2.09 ag 1.10 | 0.95 | 81.0 | 26.5 | 2.87 | 2.04 Free amino-acid nitrogen. At the time our first paper was published we had been unable to find an agent which would remove or destroy the urea without either hydrolyzing conjugated amino-acids (hippuric acid, pep- tone, etc.) or removing free ones. Treatment in an autoclave, as described in the first part of this paper, efficiently destroys the urea, but it also hydrolyzes the conjugated amino-acids. Mer- curie acetate with alkali precipitates urea completely, but it also precipitates almost all of the amino-acids. We were therefore forced to take advantage of the fact that urea reacts to the extent of only about 3 per cent with nitrous acid in the time that amino- acids react with 100 per cent of their nitrogen. After the amino- acid nitrogen has been driven off and measured one can ascertain 128 Determination of Amino Nitrogen in Urine the rate at which urea is evolving nitrogen in the same mixture, and thus make a correction for the small percentage of urea nitro- gen decomposed while the amino-acids were finishing the reaction. The method is satisfactory when, as in normal blood, the excess of urea is not too great, but in the urine the urea nitrogen is about 100 times the normal free amino-acid nitrogen. For this reason the method could- not be depended upon to give results more accurate than +0.5 per cent of the total nitrogen of the urine, and was therefore of value to determine amino-acids only when they were present in abnormally large amounts. Recently, however, Marshall* has found in the urease of the soy bean the specific reagent for the destruction of urea. He shows that the water extract of the beans (prepared by extracting the pulverized beans with 10 parts of water for an hour at room temperature, then warming the mixture to 35°, adding one-tenth volume of 4 HCl to coagulate proteins, and filtering) completely hydrolyzes urea in the space of a few hours at 35° to ammonium carbonate. We have been able to confirm his results, and_ find furthermore that the extract under the conditions used does not appreciably hydrolyze , acid, casein, or peptone, nor deam- inize amino-acids. One peculiarity which we have noticed is that the enzyme does not appear to follow the law of mass action. A given amount is required to decompose a urine under given conditions, and the — dilution of the reacting substances can be varied greatly without much affecting results. It is essential, therefore, that the amount of extract.taken should be sufficient to completely decompose all the urea present. This is most certainly assured by testing measured portions of the extract with urines of concentration at least as great as that of those to be analyzed, and ascertaining the proportion of enzyme necessary to give a maximum amount of ammonia. Our method is to take 3-cc. portions of urine in 100 ce. test tubes, add 1.0-, 1.5-, 2.0-, 2.5-, and 38.0-ce: portions of extract with a few drops of toluol to the respective tubes of the series, and place in a bath at 35° for three hours, or sixteen to twenty hours at room temperature. To each tube 2 cc. of sat- urated potassium carbonate solution are then added, and the * Marshall: This Journal, xiv, p. 288, 19138. oe 4 Donald D. Van Slyke 129 © ammonia is driven into 25 cc. of *, hydrochloric acid by ten min- utes’ aeration according to Folin’s recent method. As an example the amounts of 74 acid neutralized in one such test were 17.85, 19.50, 20.20, 20.05, and 20.20 cc. respectively. Under the con- ditions, 3 ce. of urine required 2 cc. of enzyme solution. The soy bean extract was tested for proteolytic activity in the following experiment. The extract added to 2 per cent urea solution in the proportions of 2 volumes of extract to 5 of urea solution, completely decomposed the latter in sixteen hours at room temperature, To test the action on a protein the following solution was prepared: 25 cc. H2,O; 10 cc. soy bean extract; 0.125 gram casein; 1.0 ce. 0.1 n NaOH (to neutralize the casein); 0.2 gram NaCl. In this and subsequent experiments toluene was used as pre- servative. ‘The solution was alkaline to alizarine, acid to litmus. Two-cubic-centimeter portions, taken at once and after the solution had stood sixteen hours at 25°, were analyzed in the micro-apparatus for amino nitrogen. The time of reaction was four minutes in each case. The results were: ARINB RG. .i5).--- <>. tn «spp 0.26 cc. Nz at 21°, 767 mm. After sixteen hours... Ae ne 0.29 cc. Ne at 19°, 767 mm. A control performed on a solution without casein, but other- wise like the above, gave 0.20 cc. nitrogen gas under the same con- ditions. The free amino nitrogen in the amount of casein present in the first solution (5.5 per cent of the total nitrogen, see later paper by Van Slyke and Birchard) would yield 0.09 ec. of ni- trogen gas if given time to react completely. It is evident from - the above that, under conditions that result in complete de- composition of urea, casein is not appreciably hydrolyzed by the ; urease. In fact a solution of sodium caseinate without. enzy showed under the same conditions as the result of autohydrolysis, more increase in amino nitrogen (0.05 cc.) than that noted above. In order to test the extract for the presence of an erepsin, the following solution was prepared: 25 ec. H:O; 10 ce. soy bean ex- ~ tract; 0.200 — Siegfried’s peptone from fibrin; 0.200 gram y ‘NaCl. ‘ This Journal, xi, p. 507, 1912. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 1. ia 130 Determination of Amino Nitrogen in Urine The solution was slightly alkaline to alizarine, acid to litmus. Two-cubic-centimeter portions were taken for amino nitrogen deter- minations. The results of the determinations are given in the following table. TABLE II. PERIOD OF ACTION | Ne GAS TEMPERATURE PRESSURE hours | ce. deg. C. mm. 0 0.66 22 768 s. | so 22 768 24 | 0.70 20 766 72 | 0.82 26 766 Within the period required to decompose urea (sixteen hours) the action of the extract on the peptone is barely discernible. To test the effect of the enzyme on amino-acids, the artificially digested meat termed “ereptone” and manufactured at Hoechst a. M. according to Abderhalden’s method was used as a substrate. The decomposition of the original proteins into amino-acids was almost complete in the preparation used, as heating for twenty- four hours at 100° with 20 per cent hydrochloric acid increased the amino nitrogen from 63 per cent only up to 67 per cent of the total nitrogen. The use of such a preparation, which probably contains all of the amino-acids found in the body, and in about the proportions in which they exist in the body as a whole, affords — a more practical test for the purpose of the experiment than would — the utilization of some of the individual amino-acids. The solu- tion contained: 25 ce. H,O; 10 ce. soy bean extract: 0.200 gram ereptone (containing 12.8 per cent N); 0.200 gram NaCl; 0.5 ce. 0.1 N NaOH to render the solution just alkaline to alizarine. Am- inp determination on 2-ce. portions gave: PEGG, caus . . a> « ea va so oe 1.60 ce, Ny at 19°, 767 mm, After sixteen hours....................1.59 ce. Ng at 19°, 771 mm. No deaminization whatever occurred. To test the extract for its ability to hydrolyze hippuric acid the following solution was prepared: Kahlbaum’s hippuric acid, 0.200 gram; 0.1 N NaOH (1 equivalent), 11.1 ce; HyO, 10 cc; soy bean extract, 10 ce. Donald D. Van Slyke 131 The solution was acid to litmus, alkaline to Congo and alizarine. Two-cubic-centimeter portions were taken foranalysis. Theresults were: PO) SNe ro a re 0.20 ce. Nz at 21°, 764 mm. After sixteen hours.................... 0.24 cc. Ne at 21°, 764 mm. The increase of 0.04 cc. indicates the hydrolysis of 2 per cent of the hippuric acid present. Whether this was due to the action of the extract or to spontaneous splitting of the hippuric acid was not determined. The effect is, in any case, negligible so far as its influence on urine analyses is concerned. All the above experiments were repeated, with similar results. They show that the soy bean extract, under the conditions used for complete decomposition of urea, does not hydrolyze casein, nor, to a significant extent, peptone or hippuric acid, nor does it deaminize amino-acids, Method for free amino-acid nitrogen. The proportion of extract necessary to completely hydrolyze urines of the maximum concentration is determined as described on p. 128. To 25 cc. of urine in a 50-cc. flask one adds the required amount of extract (usually about 15 ce. with the beans which we used) and lets the mixture stand for about one and a half times the interval which has been found sufficient to effect the maximum decomposition of urea, as observed by titration of the ammonia. These conditions assure decomposition of the last traces of urea. At the end of the digestion period 10 cc. of a 10 per cent suspen- sion of calcium hydrate are added, and the mixture is shaken and diluted up to the 50 cc. mark. It is then filtered through a dry _ folded filter, and 20 ce. of the alkaline filtrate are concentrated in - a Jena glass dish to dryness on the water bath, this process driv- ing off all the ammonia (hippuric acid is not appreciably affected _ by this treatment). The residue is moistened with 1 ce. of 50 _ per cent acetic acid, washed into a 10-cc. measuring flask, and _ diluted to the mark. One uses the entire solution for a deter- - mination in the larger amino apparatus, or 2 cc. for duplicates in _ the smaller. The reaction period with the nitrous acid should be _ kept as short as possible for the reasons given on p. 126, and the 132 Determination of Amino Nitrogen in Urine correction for the amines other than amino-acids can be made in the same manner. ; The results tabulated below were obtained with the same urines used for determination of total amino nitrogen. The amount of enzyme used was 15 cc.; the time allowed for it to act, five hours at 36°, three hours having been the period in which the proportion of extract used gavea maximum yield of ammoniain a previous test. TABLE IT. Free amino nitrogen. All determinations at 26°, 760 mm. Reaction period two and a half minutes. | | oe 2 N 100 cc. PER CENT OF | EE {| 7 sc Da no. | : <8 38 2 See | Oncor | Corrected | tynoore | 5§ ‘ 5 ¢ : 3 footed apes at Corrected i, ere ce. mgms. | mgms. | 1.211 0.36 0.28 Sa | 75 (me 0.62 0.34 0.26 92 | 7.2 0.76 0.58 2 1.750 0.48 0.35 13.0 | 9.8 0.74 0.56 0.48 | 0.35 13.0 | 9.8 0.74 0.56 ae | 3 0.833 1 | 0.25 7.7 | 6.80mneOLO3 0.31 0.25 7.7 6.8 0.93 4 1.747 0.55 0.33 | 15.0 8.9 0.85 0.54 0.33 | 15.0 8.9 0.83 5 1.300 | 0.28 | O18 | 8.0 | 5.0 0.61 0.28 0.18 | 8.0 5.0 0.61 The control solution, containing 15 cc. of extract with 25 cc. of water in place of the urine, gave 0.20 ce. of nitrogen gas during the first two and a half minutes of the reaction at 26°, and 0.05 ce. during the second two and a half minutes. These amounts, which were found without appreciable deviation in several dupli- cates, were subtracted from the volumes of gas read at each deter- mination. The corrections for the volume of gas evolved by amines other than amino-acids were made as described on p. 126, Dupli- cate amino determinations with 2-cc, portions were made in each case with the micro-apparatus. Donald D. Van Slyke EY 4938 It will be noted that the amount of nitrogen evolved by amines other than amino-acids is practically the same, amounting to 0.2-0.3 per cent of the total nitrogen of the urine, whether the urea was removed by hydrolysis with sulphuric acid in the auto- clave, or by the urease. When this correction is determined, we believe that the methods for both free and total amino nitrogen described above give with a close degree of approximation the actual amount of amino nitrogen present in the form of amino- acids. As the correction is relatively small and constant, it is probable that in most work, where comparative results chiefly are desired, it will be unnecessary to take it into account. Hippuric acid. _ Henriques and Sérensen have ascertained the conditions for the complete extraction of hippuric acid from urine with ethyl acetate, its subsequent hydrolysis to glycocoll and benzoic acid with hydrochloric acid, and its determination by titration of the amino nitrogen of the glycocoll by the formol method. The same methods can be applied, making the final determination by the gasometric method instead of the formol titration. When the gasometric method is applied, the results must be multiplied by the factor 0.93, as glycocoll, unlike the other amino-acids, gives off several per cent more gas than the volume corresponding > to its nitrogen content. Using this factor, however, one can obtain without trouble results accurate to within 1 per cent of the total glycocoll determined. The gasometric method has some advantages over the formol method,® and in this case should be particularly convenient because it simplifies the process of extrac- tion by permitting one to work with small volumes of urine. As work from Henriques and Sérensen hardly requires confirmation of its reliability, and the substitution of the gasometric for the formol method in the final determination is an obvious modifica- tion, a more detailed discussion here appears unnecessary. 5 Zeitschr. f. physiol. Chem., \xiii, p. 27, 1910; lxiv, p. 120, 1911. 6 This Journal, xii, p. 302, 1912. 134 Determination of Amino Nitrogen in Urine CONCLUSION. The previously published process of determining the total amino- acid nitrogen (free amino-acids+conjugated amino-acids in the form of hippuric acid, peptides, proteins, etc.) has been simplified so that the operation is much shortened and the more laborious parts, boiling off ammonia and washing bulky precipitates, are dispensed with. The free amino-acids alone can readily be deter- mined after decomposition of the urea with soy bean urease, hich hydrolyzes urea completely without either freeing conjugated amino-acids or deaminizing free ones. The applicability of the gasometric method for the determination of hippuric acid is indicated. RESEARCHES ON PURINES. XIII! ON 2,8-DIOXY-1,6-DIMETHYLPURINE AND 2,6-DIOXY-3,4-DIME- THYL-5-NITROPYRIMIDINE (a-DIMETHYLNITROURACIL).’ By CARL O. JOHNS ann EMIL J. BAUMANN. (From the Sheffield Laboratory of Yale University.) (Received for publication, August 26, 1913.) We find that an aqueous solution of the sodium salt of 2-oxy- 4-methyl-5-nitro-6-aminopyrimidine® (I) reacts readily with di- methylsulphate and gives an 80 per cent yield of the corresponding dimethyl derivative. It seemed probable that the compound thus formed was 2-oxy-3,4-dimethyl-5-nitro-6-aminopyrimidine (II), because, in our previous work on alkylations of pyrimidines which contained oxygen in position 2 and an amino or alkylamino group in position 6 we found that the alkyl group entered position 3 in the pyrimidine ring. The following experiments show conclu- sively that in the case now under consideration the alkyl group also entered position 3. The substance obtained by methylating 2-oxy-4-methy]-5-nitro- 6-aminopyrimidine was heated with 25 per cent sulphuric acid under pressure. This treatment removed the amino group and a good yield of a 2,6-dioxy-5-nitro-dimethyl-pyrimidine was obtained. Two such compounds can exist in which one of the methyl groups is attached to nitrogen in the urea grouping of the pyrimidine ring and the other methyl group attached to the carbon atom in posi- tion 4, namely, 2,6-dioxy-1,4-dimethyl-5-nitropyrimidine (VII) and 2,6-dioxy-3,4-dimethyl-5-nitropyrimidine (III). Lehman! ob- 1 Johns and Baumann: this Journal, xv, p. 515, 1913. The present inves- tigation was aided by a grant from the Bache fund. Behrend and Dietrich: Ann. d. Chem. (Liebig), eecix, p. 266, 1899; Behrend and Thurm: bid., ecexxiii, p. 163, 1902. 3 Johns: Amer. Chem. Journ., xli, p. 60, 1909. » 4 Johns: this Journal, xi, p. 75, 1912; xiv, p. 3, 1913. 5 Lehman: Ann. d. Chem. (Liebig), ccliii, p. 84, 1899. 135 136 Researches on Purines tained one of the above compounds by the action of methyl iodide on the potassium salt of nitromethyluracil.6 His compound melted at 149°C. and the structure assigned to it was 2,6-dioxy-1,4-di- methyl-5-nitropyrimidine’ (VII). Our 2,6-dioxy-5-nitro-dimethyl- pyrimidine melts at 191°C. Hence, it cannot be identical with the compound obtained by Lehman and if the correct structure has been assigned to his compound ours must be 2,6-dioxy-3,4-di- methyl-5-nitropyrimidine (III). Behrend and Kéhler® have shown that fuming nitric acid not only nitrates 4-methyluracil (X) but also oxidizes the methyl group forming 2,6-dioxy-4-carboxyl-5-nitropyrimidine (XT) and that this latter compound loses carbon dioxide with the consequent forma- tion of 2,6-dioxy-5-nitropyrimidine or nitrouracil (XII). We found that 2,6-dioxy-5-nitro-dimethylpyrimidine was also oxidized by fuming nitric acid and that the 2,6-dioxy-3-methyl- 4-carboxyl-5-nitropyrimidine (VI) lost carbon dioxide which re- sulted in the formation of 2,6-dioxy-3-methyl-5-nitropyrimidine (IX). The structure of this compound has been firmly established by the work of Behrend and his collaborators.® It melts at 255°C. and contains one molecule of water of crystallization and is there- fore readily distinguished from its isomer 2,6-dioxy-1-methyl-5- nitropyrimidine” (VIII) which melts at 263°C. and does not con- tain water of crystallization. The compound obtained by us contained water of crystallization and melted at 255°C. It was_ therefore 2,6-dioxy-3-methyl-5-nitropyrimidine. Hence in methyl ating 2-oxy-4-methyl-5-nitro-6-aminopyrimidine the methyl group — entered position 3 and the compound formed was 2-oxy-3,4-di- methyl-5-nitro-6-aminopyrimidine (II). The latter compound was reduced rapidly by the action of freshly precipitated ferrous hydroxide but the reaction was not smooth. After isolating about 40 per cent of the calculated weight of 2-oxy- 3,4-dimethyl-5,6-diaminopyrimidine (V) a tarry by-product re- mained. * Behrend: Ann. d. Chem, (Liebig), ccxl, p. 3, 1887. 1 Beilatein’s Handbuch, i, p. 1350 (third edition). * Behrend: Ann. d. Chem. (Liebig), cexxix, p, 32, 1885; Kohler: — ecoxxxy , p. 50, 1886. * Behrend and Thurm: Ann. d, Chem. (Liebig), eeexxiii, p. 163, 1902, 10 Thid. i _. C. O. Johns and E. J. Baumann 137 When 2-oxy-3,4-dimethyl-5,6-diaminopyrimidine was heated with urea we obtained an excellent yield of 2,8-dioxy-1,6-dimethyl- purine (IV). These researches will be continued. oC’ CNO;) ——> ..O0C CNO, —=> 70C >CNO, | || a). ba il HN—C:CH; CH;:N—C:CH; CH;N—C:CH; I Il Ill J : ee or HN—CO : OC C—NH «<— OC’ CNH, OC ONO, es | | | N=C—NH CH;-N—C:CH; CH,;:N—C: COOH : IV Vv VI-) ‘i CH,-N—CO Peas a | a by “ha oc aio: OC CNO, HN=—C.CH; ~ HN—CH CH;:N—CH ~ VII VIII Ix HN—CO a HN—CO hab fete) ‘ nt ae ae me " CNO, HN—C:CH; HN—C: COOH HN—CH x XI XII EXPERIMENTAL PART. 2-Oxy-3,4-dimethyl-5-nitro-6-aminopyrimidine. . oc CNO, fl CH;:N—C:CH; 138 Researches on Purines Ten grams of pulverized 2-oxy-4-methy]-5-nitro-6-aminopyrimi- dine" were dissolved in 100 ec. of hot water containing 2.8 grams of sodium hydroxide. After cooling this solution to room tempera- ture, 10 grams of dimethylsulphate were added and the mixture was shaken to keep the dimethylsulphate in suspension. In less than five minutes crystals began to form. The mixture was shaken two or three minutes longer and then allowed to stand until it gave an acid reaction, which usually required less than fifteen minutes. Heat was evolved during the reaction and, after cooling, the precipitate was filtered off and washed with a little cold water and alcohol. The yield was 8.7 grams or 80 per cent of theory. This substance contained but a trace of the original 2-oxy4-methyl-5-nitro-6-aminopyrimidine and was pure enough for subsequent experiments. The 2-oxy-3,4-dimethyl-5-nitro-6- aminopyrimidine dissolved readily on continued boiling in water and on cooling the solution slowly it crystallized in lustrous prisms that formed radiating clusters or sheaves. These crystals did not have a definite melting point but began to darken at about 170°C. and effervesced at 190° to 195°C. They were moderately soluble . in hot alcohol, slightly soluble in boiling benzene and insoluble in ether. They dissolved readily in dilute hydrochloric acid and glacial acetic acid. They formed yellow solutions in strong alkalies and were moderately soluble in ammonium hydroxide. The crystals which were obtained from aqueous solutions possessed a pearly luster. This was lost on drying over sulphuric acid in a desiccator for one or two days though analyses showed that one- half molecule of water of crystallization still remained. I. 2.0855 grams of substance dried over sulphuric acid for twenty-four hours lost 0.0993 gram at 120°-130°C. II. 2.8975 grams of substance dried over sulphuric acid for 48 hours lost 0.1375 gram at 120°-130°C, Calculated for Found: CeHsO3N 4.420: I II ji A a 4.66 4.76 4.78 Calculated for Found: CoHgOgNa: I Il BERG ov biicies «+o CME)» SERA o> oa 30.48 30.38 30 54 't Johns: loc, cit. C. O. Johns and E. J. Baumann 139 2,6-Dioxy-3 ,4-dimethyl-5-nitropyrimidine. HN—CO CH;:N—C-CHs; Three grams of 2-oxy-3,4-dimethyl-5-nitro-6-aminopyrimidine from the above experiment were dissolved in 20 ce. of 25 per cent sulphuric acid and the solution was heated in a sealed tube at 160°C. for two hours. ‘When the contents of the tube were coo ed a deposit of long, slender prisms was obtained. A second crop was isolated by neutralizing the filtrate with barium hydroxide, filter- ing, and concentrating this filtrate. The yield was 75 per cent of theory. The crude substance melted at 186° to 190°C. When recrystallized from alcohol the melting point was 191°C. It was easily soluble in hot water and on cooling the solution it erystallized rapidly in slender prisms. It also dissolved readily in hot alcohol, slightly in benzene but did not dissolve in ether. Dilute alkalies dissolved it easily. Calculated for Found: CeH704N3: I II a ae 22.70 22.61 22 .62 The oxidation of 2,6-dioxy-3,4-dimethyl-5-nitropyrimidine with nitric acid. The formation of 2,6-dioxy-3-methyl-5-nitropyrimidine.” HN—CO Oc CNO, | CH; -N—CH One gram of 2,6-dioxy-3,4-dimethyl-5-nitropyrimidine was dis- solved in 10 cc. of nitric acid of specific gravity 1.5 and 2 cc. of concentrated sulphuric acid were added. The solution was heated on the water bath for one and one-half hours. Oxidation took place with effervescence and the liberation of brown fumes. The solution was diluted with water and the acids were neutralized with ammonia. On evaporating to dryness and washing with cold 12 Behrend and Thurm: Ann. d. Chem. (Liebig), ccexxiii, p. 164, 1902. 140 . Researches on Purines water to remove salts, there remained a crystalline substance that weighed 0.2 gram. This melted at 254° to 255°C. and when re- crystallized from water it melted sharply at 255° to 256°C. The crystals contained water of crystallization and when mixed with a pure sample of 2,6-dioxy-3-methyl-5-nitropyrimidine the melting point remained the same. Hence, the methyl group attached to nitrogen was in position 3 in the pyrimidine ring. The substance was dried at 120° to 130°C. Calculated for Found: . CsH;04Ns: | ere ~ .. 24,66 24.44 2-Oxy-3,4-dimethyl-5,6-diaminopyrimidine. N==CNH: OC CNH: CH;: 1h *CH; Five grams of 2-oxy-3,4-dimethyl-5-nitro-6-aminopyrimidine were dissolved in a mixture of 50 cc. of concentrated ammonia and 75 cc. of water by warming gently. This solution was cooled to room temperature and a hot, concentrated aqueous solution of 53 grams of crystallized ferrous sulphate was added gradually. Reduc-. tion took place rapidly and was accompanied with the liberation — of heat. A solution of 63 grams of barium hydroxide was added to — precipitate the sulphate and the excess of baryta was removed by — adding ammonium carbonate. The mixture was well shaken and filtered after standing for an hour. The filtrate was evaporated to a small volume on the water bath and then cooled, A dark erystal- « line mass separated. This was dissolved in hot water and decolor- ized with blood coal. The diaminopyrimidine was thus obtained pure and colorless in the form of burrs that consisted of small plates. The yield was about 40 per cent of theory. The crystals possessed a pearly luster. They were very soluble in hot water and crystal- lized well on cooling the solution. They were moderately soluble in boiling aleoho! but did not dissolve in benzene orether. Dilute acids dissolved themeasily, They did not exhibit a definite melting point but began to decompose at about 230°C, When an aqueous C. O. Johns and E. J. Baumann 141 solution of this substance was added to a cold ammoniacal silver solution a white gelatinous precipitate resulted and, on heating, a silver mirror was produced. Calculated for Found: CeHi00Na: iad ugh uc ER Ss Mie il ea 36.36 36 .37 2,8-Dioxy-1,6-dimethylpurine. CH; . N—C A CH; oC C—NH . | | N=C—NH Two grams of 2-oxy-3,4-dimethyl-5,6-diaminopyrimidine and 2 grams of urea were pulverized together and the mixture was ~ heated for an hour at 170° to 180°C. inan oil bath. Themassmelted to a liquid and frothing occurred while there was a copious evolution . of ammonia. After some twenty-five minutes the reaction sub- sided and the mixture became a solid mass. After cooling, the . reaction-product was dissolved in dilute ammonia and the solution __was clarified with blood coal. After boiling off most of the am- - monia, the solution was acidified with acetic acid whereupon crystals of the purine began to separate as the solution cooled. After two hours the crystals were filtered off. Another crop was obtained by concentrating the filtrate. The yield was 2 grams or 85 per cent of the calculated weight. The portion used for analysis was re- i. * crystallized from water. This purine dissolves in about 60 parts of boiling water and is slightly soluble in hot alcohol but does not dis- _ solve in benzene or ether. It dissolves readily in hydrochloric acid or alkalies. It crystallizes from water in burrs composed of small prisms that contain one molecule of water of crystallization. It decomposed at 260° to 265°C. It did not form a difficultly soluble picrate or barium salt. Its water solution gave a gelatinous preci- pitate with mercuric chloride. This was soluble in hot water but reappeared on cooling the solution. With an ammoniacal silver solution a white precipitate was obtained. This did not darken when the contents of the test tube were boiled. Nitric acid oxi- 142 Researches on Purines dized the purine readily and on careful evaporation a yellow crust remained. This became rose colored when treated with alkalies. 0.8798 gram of substance lost 0.0825 gram at 120° to 130°C. Calculated for Found: C7Hs02N4.H20: | 0 OE 9.09 9.29 Calculated for Found: C7Hs02Nq: SP OA 81.11 31.32 eins = * ' POLYATOMIC ALCOHOLS AS SOURCES OF CARBON FOR LOWER FUNGI. By RAY E. NEIDIG. (From the Chemical Section of the Iowa Agricultural Experiment Station.) (Received for publication, August 30, 1913.) -The carbon nutrition of the lower fungi has been studied quite extensively. The diversity of simple organic substances which molds are able to utilize as sources of carbon is indeed surprising. Not only do the naturally-occurring sugars supply the carbon re- quirements of these fungi, but many other substances, among them laboratory products not known to occur in nature, appear to be more or less readily utilized by these organisms. The vigor of the culture, however, may yary considerably with the nature of the substrate. Many substances on which only a scant growth can be obtained have been reported in the literature as available sources of carbon. Itis of course impossible to express availability of a par- ticular substrate for a given organism on an adequate quantitative basis, yet some distinction should be made between a sparse growth and a vigorous culture. The cultures herein described were made for the purpose of deter- mining differences in availability in the series of polyatomic al- cohols. The substances selected represent a series differing pro- gressively by the group CHOH. They all occur in nature, the first three in the form of esters, the others in the free state. Iso- meric synthetic products will not be considered here. The series represents therefore substances containing one to six alcohol radi- s and a similar number of carbon atoms, and may be designated by the general formula CrHan420n. Following is the list: NUMBER OF CARBON | NUMBER OF CARBON ATOMS AND SUBSTANCE ATOMS AND SUBSTANCE ALCOHOL RADICAIS » | ALCOHOL RADICALS 1 Methyl alcohol | 5 Adonitol 2 Ethylene glycol | 6 Mannitol 3 Glycerol } 6 Dulcitol 4 Erythritol 6 Sorbitol Several of these substances have been reported by Emmerling (Centralbl. f. Bakt., x, II, p. 273) as available for Aspergillus niger, but quantitative differences were not considered. 143 144 Utilization of Alcohols by Fungi These substances were introduced into Czapek’s medium in place of the usual sugar. No other source of carbon was present. Inoc- ulation was made with the spores of pure cultures, and the tubes were allowed to remain in the dark at room temperature. The cultures were examined at the end of the first, second and third week. The following notation will be used to designate the appear- ance of the culture: oe ae he good normal culture rss ys on co se ce oe medium growth 3 iis as. kD. ee slight growth NB ieiae toss. 204 germination or submerged hyphae only riers Co. +50 3 no growth i SOURCE OF CARBON onganism «60 | ste 3 a ie 7 nel \.3| 8s | 3 sis | Se] ise] $8] | a | 3 g\3 : i jee |@ |e|a | 2/8] 4a/ a Aspergillus 410\0| O e++\+-+ G tHe tt niger 2}0\0| @ j4++4++ | G 4+4+4/}44+4/4+4++ 3)}0/0| G fe++/++ G jttepettit++ Aspergillus |1/0/0| G@if++) + | + | + }9G@ [+++ clavatus |2/0/0! G |j+++ + + + | G +++ 3/010) Se t+] oe + j++ |°@ [+++ Aspergillus 1;O01;0| @ |+++/++ + Reet t+ i++ fumigatus 2;O0/0) + |JFt+tit4etit¢+ [44+4+/44+4+/4+4+4+ |3)O;O) [bet etit+ [++4/4-++/4+++ Penicillium (1/0/0; G@ +4++/++ MN Pathe hs expansum)§=(§2/0/0| + [+++ F44+/+4+ [4+4+4/4+4+4+/4+++4+ 31/O}O] + lF+ttetit+ ++H4++4+i4++4+ Fusarium = | 10/0] + |+++| G +++/++4/++4+/4++4 oxysporium | 2/O0/O}++ [+++ G [t++i+++/++4+/+++ 13 /O/Ol++ |+++) G Heel ttt tlt tt Cladisporium 1 0O;O| G [+++] + + + G +++ herbarum = §(2/0/0) G |+++i4++ [++ + G j+++ 13 };O;O) G fe++ie++i+4+ i G +++ Penicillium 1;0;O} G f+++) + [++ [+4 eer roqueforti 2/0/O) + ltt ett tit ttiteeiet itt 8)O;O) + tee tte tt tee ttt ++ Penicillium 1;0;0} G \++ } + ee + + camemberti 2.0/0) G j+++)-+++) + eet itt+i+++ Ol + [+++-+4). + feet | t tht Ray E. Neidig | 145 It will be noted that the first two members of the series are not capable of producing normal cultures. Gl¥cerol is readily availa- ble and gives cultures equal in vigor to those grown on cane sugar. With increasing carbon, however, the availability does not increase, as might perhaps be expected. Adonitol, for example, does not compare favorably with glycerol or even erythritol, and two of the hexatomic alcohols fail to yield cultures equal to those on glycerol. It will be noted that the alcohols beginning with erythritol con- tain asymmetric carbon atoms. But considering the fact that glycerol is not asymmetric, no connection can be established be- tween availability and carbon asymmetry. On the other hand, there may be some relation between availability and the nature of . the intermediate oxidation products, since all the substances which are available, including glycerol, yield oxidation produets contain- ing one or more asymmetric carbon atoms. The writer takes pleasure in acknowledging his indebtedness to Dr. A. W. Dox, at whose suggestion this work was undertaken. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 1. anh. rank Ng THE COMPARATIVE COMPOSITION OF HUMAN MILK AND OF COW’S MILK. By EDWARD B. MEIGS ann HOWARD L. MARSH. (From the Robert Hare Chemical Laboratory of the University of Pennsyl- vania and the Wistar Institute of Anatomy and Biology.) (Received for publication, August 30, 1913.) INTRODUCTION. The following article is an account of work done by Arthur V. Meigs, Howard L. Marsh, William H. Welker and W. L. Croll on the chemical analysis of human milk and of cow’s milk. The work was carried out in the Robert Hare Chemical baa of the University of Pennsylvania and was to a large origene® vised by John Marshall, the Director of that laboratory. present account has been written by Edward B. Meigs, in Biave. ration with Howard L. Marsh, after the death of Arthur V. Meigs, on January 1, 1912. The subject of milk analysis was taken up by Arthur V. Meigs more than thirty years ago with the idea of discovering how cow’s milk should be modified in order to make a proper food for very young infants. On the basis of analyses made in 1881— 1884, Meigs devised a food which he afterward used in his prac- tice with great success. He published an account of his early investigations in book form in 1885,' and has since published a number of smaller articles on the subject.2 In 1908 he again began chemical work on the subject which he continued until his death in 1912. A short preliminary account of some of the results of this later work appeared in 1911; and since that time 1 Arthur V. Meigs: Milk Analysis and Infant Feeding, Philadelphia, i 1885. 2 Arthur V. Meigs: Transactions of the College of Physicians of Phila- delphia, Third Series, viii, p. 139, 1885; Ibid., xxiv, p. 136, 1902; Archives of Pediatrics, December, 1889; Feeding in Early Infancy, Philadelphia, 1896. 8 Arthur V. Meigs and Howard L. Marsh: The Medical Record, December 30, 1911. 147 148 Composition of Milk the authors of this article have been endeavoring to codrdinate the other results and to prepare them for publication. Meigs’ work on milk in 1881 consisted in an attempt to com- pare human milk and cow’s milk in their content of protein (or “casein’’),4 fat, lactose, water and salts. The chief outcome of the work was the conclusion that human milk contains less than half as much protein as cow’s milk and about 50 per cent more lactose. In other respects Meigs’ results were not in sharp dis- agreement with those of his predecessors, but his figures for pro- tein and Jactose were quite different from those which were usu- TABLE I. The average composition of human milk and cow’s milk according to Meigs’ early analyses, and those of certain well-known authors whose results were available in 1881. The various constituents are given as percentages of the whole milk. | WATER PROTEIN FAT LACTOSB ASH | | ee i Human Cow’s Human! Cow’s Human} Cow’s Human) Cow’s |Human| Cow’s : | milk | milk | milk | milk mille | milk’ milk | milk |" milk | milk Lehmann’. . 3.5 | 4.5 | 4 to6 Gorup- Besanezf. 88. 908'84.28 | 3, ee 2.666) 6.47 | _ 4.34 | 0.138) 0.63 Meigst...... 87 .16387 .780, 1. 46) 3.022) 4.283) 3.759 7. 4.949 0.101) 0.488 *Lehmann: Physiological Chemistry, Cavendish Society translation, London, 1851, vol. i, p. 383, vol. fl, p. 341. t Gorup-Besanez: Lehrbuch der physiologischen Chemie, Braunschweig, 1878, pp. 421 and 424: the figures for human milk are those quoted from Vernois and Becquerel, and are averages from i 89 analyses. t Meigs: Milk Analysis and Infant Feeding, Philadelphia, 1885, pp. 34 and 36. ally given at that time. A comparison of Meigs’ figures with some of the best known figures which were available in 1881 is given in Table I. Meigs was of course familiar with later analyses which had not at that time been quoted in text-books. In these analyses the protein of human milk was variously given at from 0.215 to ‘ The word ‘‘casein” was, at that time, very generally used to designate all the proteins of milk; it now means, as is well known, a particular pro- tein body which constitutes about 80 per cent of the total protein of cow’s milk and a somewhat less proportion of the protein of human milk. In the subsequent discussion the words ‘‘casein’’ and ‘protein’ will be used in their modern senses. . ae Edward B. Meigs and Howard L. Marsh _ 149 over 7 per cent, while the figures for lactose ranged from 1.921 to 8.805 per cent. In the miJk analyses which have been published since 1881, the figures given to represent the percentage of protein and lac- tose in cow’s milk have not varied widely from those published by Gorup-Besanez. But the percentages of these constituents in human milk have been quite variously given, as Table II shows; in the more modern analyses the results are quite near to those reached by Meigs in 1881. It is generally agreed among physiological chemists at present that the fat, lactose, ash and total nitrogen of milk can be quan- titatively determined satisfactorily. The difficulties of milk anal- TABLE II. The average composition of human milk according to analyses published since 1881. The constituents are given as percentages of the whole milk. WATER PROTEIN FAT | LACTOSE ASH Munk and Ufielmann*..,.) 80.2 | 21 | 34 | S.QMM——O2 iC 87.4 | 2.3 3.8 6.2 60.8 Heoubuerti 4 .......089 | 1.08 4.07 7.03 0.21 Camerer and Séldner§.... | 1.27 3.91 6.52 0.22 * Munk and Uffelmann: Erndhrung des gesunden und kranken Menschen: Wien and Liepzig, 1887, p. 269. : t Quoted by Camerer and Séldner: Zeitschr. f. Biol., xxxiii, 1896, p. 43. . t Heubner: Berl. klin. Wochenschr., 1894, Nos. 37 and 38. § Camerer and Séldner: Zeitschr. f. Biol., xxxvi, Table II, pp. 280 and 281, 1898. The figures are average figures for milk between the 20th and 40th day post partum. The figure for protein is obtained by multiplying the figure for total nitrogen given by the authors by the factor 6.25. ysis lie in the determination of the protein and of the unknown constituents. Meigs and his collaborators have, in the first place, determined the quantities of fat, lactose, ash and total nitrogen in a number of samples of human milk and cow’s milk; they then attacked the question of the protein and unknown constituents. It will be convenient to divide this article into two parts, in the first of which will be given the results on the easily determinable constituents of milk; and in the second, the work on the protein and unknown substances; this latter work was still in a very fragmentary state at the time of Meigs’ death. It will be well to begin with a description of the samples of milk used in the various studies. 150 Composition of Milk Each of the samples of human milk numbered 1, 2, 4, 6, 7, 8 and 10 was made up of milk from several women; each of the other samples was com- posed of the milk from one woman. The composite samples of human milk came from women in maternity hospitals, and all were taken between the fourth and ninth days after delivery. The other samples of human milk may be described as follows: No. 3. Colostrum—collected up to and including the fourth day after delivery; maternity hospital patient. No. 5. Milk collected near the end of lactation, eighteen months or more post partum; woman in moderate circumstances. The flow of milk had almost ceased so that only very little could be obtained. ; No. 9. Milk collected during the fifth month after delivery; woman in good circumstances. The cow’s milk came partly from thoroughbred Guernsey cows and partly from grade cows; samples 1, 9, 13 and 20 are the former, the others, the latter. Each of the samples of cows’ milk except Nos. 1, 9, 13 and 20 was made up of milk from several cows, and the periods of lactation have not been accurately determined. All the samples may, however, be taken to represent milk from the middle period of lactation, from two to six months after delivery. In preparing the work for publication, the data obtained from a few samples of milk have been left entirely out of consideration, either because they contained obvious analytical errors, or else because the samples were used for answering subsidiary questions which had no bearing on the com- position of milk. ss " PART I. STUDIES ON THE AMOUNTS OF ASH, LACTOSE, FAT AND NITROGEN IN HUMAN MILK AND IN COW’S MILK. : Methods of experimentation. The ash determinations were made by igniting the dry residues from the samples cautiously at low temperatures in platinum dishes. The nitrogen determinations were made by the Kjeldahl method. The lactose was determined by Fehling’s method as follows: 25 ec. of cow’s milk or 15 ce. of human milk were diluted to 400 ec. in a 500 ce. grad- uated flask. Fifteen cc, of ¥ NaOH and 10 ce. of CuSO, of the strength used in a Fehling’s solution were added and the volume of the liquid in the flask was diluted to 500 ce. with water. After this mixture had been well shaken, it was filtered through a dry filter into a dry flask. One hundred cc. of this solution were added to 50 ee. of hot Fehling's solution and the mixture was boiled for six minutes. It was then quickly filtered through an alundum crucible and washed with about 600 cc. of boiling water, The resulting cuprous oxide in the crucible was dissolved with nitric acid and the solution poured into the beaker which had been used in the reduction, The crucible was then thoroughly washed by passing hot water through it and this wash water was added to a Edward B. Meigs and Howard L. Marsh 151 the nitric acid solution which had been poured from the crucible. This nitric acid solution of the cuprous oxide was evaporated on a water bath until free from nitric acid. It was then dissolved with acetic acid and water, 8 grams of zinc acetate were added and the liquid was transferred to a small glass-stoppered bottle, 4 grams of potassium iodide were added and the solution titrated against jy sodium thiosulphate with starch as an indicator. The amount of lactose corresponding to the amount of copper found was ascertained by referring to a table given in Bulletin No. 107 (revised) of the U. 8. Department of Agriculture, Bureau of Chemistry, pp. 48 and 49. The figures represent lactose plus one molecule of water of crystallization. Fat was determined by a modification of the method recommended by Meigs in 1885.5 To 10cc. of milk in a 100 cc. glass-stoppered cylinder were added 20 cc. of distilled water and 20 cc. of ethyl ether and the mixture shaken for five minutes. Twenty cc. of 95 per cent ethyl alcohol were then added and the whole again shaken for five minutes. The cylinder was then allowed to stand until its contents separated into two distinct layers. The upper layer was removed by the specially designed pipette shown in figure 1, 5 cc. of ethyl ether were added in such a way as to wash down the sides of the cylinder, this was removed and added to the upper layer pre- viously removed, and the washing of the sides of the cylinder and of the top of the lower layer of the mixture was repeated five times im order to remove all the fat. The upper layer from the mixture plus washings was then evaporated to dryness on a water bath and the residue desiccated to constant weight over sulphuric acid at room temperature. The weight of the residue may be taken to be very nearly that of the fat contained in the milk. The residue in question was in a number of cases treated with dry ether, and it was found that a minute portion of it usually failed to dissolve. This insoluble portion was, however, never as much as 3 per cent of the weight of the original residue; it amounted on the average to about 1 per cent. It was shown by Fehling’s and by the phenyl-hydrazine tests to con- sist largely of lactose. * The fat as obtained by Meigs’ method from two portions of a sample of cow’s milk (No. 14) was analyzed for nitrogen: 0.0094 per cent and 0.0064 per cent respectively of nitrogen was found. It is probable that a eon- siderable portion of this comes from the ether-soluble lipoids. In one sample of human milk (No. 8) the ash, as determined in the fat obtained by Meigs’ method, was found to be 0.0081 per cent of the weight of the whole milk. In other experiments the fat as obtained by Meigs’ method in certain samples of milk was compared with the fat as obtained by the Soxhlet extraction of the dried solids. These results have already been published :* in twelve determinations on human milk the fat as determined by Meigs’ 5 Meigs: Milk Analysis and Infant Feeding, Philadelphia, 1885. 6 Hawk: Practical Physiological Chemistry, Philadelphia, 1912, p. 437. 152 Composition of Milk a r oo — Fic. 1. Preerre ArraNcep To Remove THe Upper Layer 1n Mera@s’ Mernop or Fat Extraction. The end of the tube A ig placed at the surface of the division between the two layers of the mixture of milk treated with the water, alcohol and ether mixture and the upper layer is then forced out through it by forcing air into the space above the mixture through tube B. Edward B. Meigs and Howard L. Marsh 1153 method averaged 0.017 per cent less than by the Soxhlet method; and in seven determinations on cow’s milk, 0.019 per cent less. It is by no means certain, however, that the figures for fat obtained by the Soxhlet method in these experiments are more nearly correct than those obtained by Meigs’ method; for the prolonged treatment of the milk residue with ether in the Soxhlet apparatus is apt to dissolve substances other than fat, and thus give a fictitious additional weight to the fat extract.? The result of this investigation has been that the product obtained by Meigs’ method of extracting the fat from milk contains 98 per cent or more of material soluble in dry ether. The results as obtained by Meigs’ method are practically the same as those obtained by the Soxhlet extrac- tion; and Meigs’ method is much quicker and subjects the milk to less treatment which is apt to produce changes in the protein and carbohydrate constituents of the fluid. Meigs’ method of determining the fat in milk was used as a matter of routine in this investigation, and the figures for fat in the tables repre- sent the dry weight of the material obtained from the “‘upper layer’’ and washings in the procedure which has just been described. In some cases, however, the fat was determined by the Soxhlet extraction as well as by Meigs’ method. The figures obtained by the Soxhlet extraction are given in a footnote to Table III. The results of the determinations of the ash, fat, lactose and total nitrogen of human milk and cow’s milk are shown in Table III. These data agree with those of Camerer and Séldner* and with most of the more recent work on milk analysis in showing the marked difference between human miik and cow’s milk in respect to their content of nitrogen and lactose. Human milk contains roughly 50 per cent more lactose than cow’s milk, and decidedly less than half as much nitrogen. The figures in Table III agree also, so far as they go, with the conclusion of Camerer and Séld- ner® that the lactose in human milk increases, while the nitrogen decreases with the progress of lactation. 7™Croll has published in detail his comparison of the results obtained by Meigs’ method of fat extraction and by the Soxhlet method in the _ Biochemical Bulletin for June, 1913. 8 Camerer and Sdéldner: Zettschr. f. Biol., xxxvi, Table II, pp. 280 and 281, 1898. 9 Tbid. Composition of Milk 154 "leg'¢ ‘oBvs0aw 'FPe"e ‘TI {61g"S ‘] 010M UuO}Z0eIZx0 JO poyzour qeIyxog O43 4q tu 8h6S' O6FGS' OSFES' OSFL'F GSL'F |THL'F O19" F SI19°F 09°F |T22°0 8920 FLL°0 Oi ee => ee SMoo 696F' 0"4np ouge6h’ OZIS'F \6I8'F \S08'F \269'F 689° F G69'F FEL'O 6E2°0 6220 °°" 6 ON) APSUieNy peiqy#noi0ygy G299' O:gaes' 0269S" 0:00" ¢ 0L0'¢ \0K6'F G6E°S \60F'S Z8E'S |169°0 FS9°0 8220 °° +1 ON) e[8uIs wo1y ¥ITUI Jo sojdureg 0g2o" 0161S" 692S' OZ8h'F PLb'F 06h F 0cZ'S Oz’ GIZ's 82°0 GeZ°O ZeL 0) °°" 61 ON) CL6F OOL6F OI86F OSIZ'F \€1Z'F [GIL F F6L'E (FOSS \S82'E [8z2'0 |TeL°0 SzL0 91 “ON 2908" 0809 0ZFOS' O€Z8" F €28'F [628 h \€8e'F |P8E"F le8e"F €69°0 [69'0 z69'0 |" OY ORT ae eae ** SMO9 apLls 1699" 0¢899"0 G OIPL F \LEL' F- (SPL F |IF6'S 96'S |266'S 16890 (099°0 6IZ'0 | TI “ON { Wloszy 4YTUT pexrur jo sajdureg GIGS" 0OESS' 0008S" 09T9'F |TO9"F E9"F 00Z'F |Fez'F 9LT' 0GL"O |IZL'0 0220 | """*"S “ON HOES" O|1ZES" OS8ze' 009" 169’ F 669° F Sh SIF ZST'F |189°0 929°0 9890 " *** LON} ‘yu 8.m0g : ooer-oecer’ r: +» looe:z bez: 2 leoe'o logs" locs'o leot-o oto loro fost cccc tenner ttenteees mors; -98[ JO pua 04} IvaU peza]]09 “C “ON AIT 609T OFLST’ v ‘4, |p60' 2 |290° 2 |T80°F |I80°F |I80°F OST'O S210 SST 0; umjivd 480d y3uoul W3¢ “6 “ON YIN PPSS OSZ8Z" ‘OPLS'9 |6FS°9 (009°9 OFI'S PET'S \ZFI'S ZSz'0 9970 BIZ OS OT ON ATW ZZIZ' OIZO9S’ OZFIT" ‘9 1h0S'9 OS'9 8z0°% \610°2 |280°% 9610 [1610 ZOZ OC S$ ON WINW | ....:...... anise 9661 OIS6T' OZ9GT' OGLE'9 |SZE°9 |6ZE°9 |898°Z |PZ8°% 98% ZSTO FSTO ISTO £°ON AIN ag0d dep 416 01 WIS 6922 O1G92Z' 09922 Ok’ 9 ehh 9 leFP'9 Gzz°% \LIZ’% [Pet's ISTO [TZT'O 161 0 9 “ON 4 ZH OFLZS’ 0607S OILL'9 |6G2°9 \E8L°9 [FOZ Z \/O1Z'Z 861% |Z8T'O 6810 98T'O | ON YIU) “ypu UDuUn fT esBi0Ay| IT I esvioAy| II I esvioAy) IT I Svi0ay II I NG@DOULIN 1VLOL asoLov1 iva HSV -96v.1980 9y} pup sishpoun yovoa ur pourwjqo sounbyf ay, sear.b 91q0) 9Y) pun ypu fo a)dwvs yove uo yno paitsv2 949m sashyouD oyooNdnd ~Y]Nu Hoym ay) fo 1ybram ay fo sabvjusosed sp uaarb ypu 8,moo pun ypu woUNY U2? UIb0.1j.U 7D}0} PuD 980720) ‘pos ‘ysp fo saNyuONd oy 7 < ‘I @TAVL Edward B. Meigs and Howard L. Marsh 1155 PART II. THE NATURE AND AMOUNT OF THE PROTEIN AND UNKNOWN SUBSTANCES IN HUMAN MILK AND IN COW’S MILK. Meigs hoped in his later work to satisfactorily settle the quan- tities of protein in human milk and in cow’s milk, and perhaps even to determine the nature of some of the more important unknown substances, which are supposed to be present in the two secretions. Both these hopes were still unfulfilled at the time of his death, but the work has thrown some further light on the problem; and the results, so far as they go, will be given here. Experiments which bear solely on the protein content of milk. The casein and globulin were precipitated from several samples of human milk and cow’s milk by magnesium sulphate and the albumin was precipitated from the magnesium sulphate filtrates by acetic acid and heat.!° The nitrogen content was determined in each precipitate separately, and also, in a separate portion, for the whole milk. The results are given in Table IV. In other samples of milk the colloids were completely precipi- ’ tated from the whole milk according to Marshall’s" aluminium hydroxide method. The nitrogen in the colloid precipitates, in the filtrates, and in separate portions of the whole milk was determined. The results are given in Table V. No very definite conclusions regarding the usual protein con- tent of human milk can be drawn from this part of the work, because the samples of human milk used were so few and from so early a period of lactation. So far as the results go, however, they confirm the view that human milk contains less than half as much protein as cow’s milk. The results are interesting also in showing that a large proportion of the nitrogen in early human milk exists in compounds which are not precipitable either by magnesium sulphate or by acetic acid and heat; and the results with Marshall’s reagent indicate that most of this “non-precipi- table” nitrogen exists in non-colloidal bodies. _ 10For details of these methods see Bulletin No. 107 (revised) of the U. S. Department of Agriculture, 1910, p. 118. 11 Marshall and Welker: Journ. Amer. Chem. Soc., xxxv, p. 820, 1913. Composition of Milk 156 *UaZ0I}{U [8}0} 94} WIOIy SezEydyoe01d UsWINGyS oy Uy puv UI[NGo[s puv ujesvo oY} Ul USZOI4JU JO sey4IGUeND oy} Jo uns = Buyjov1jqns Aq poule}qo o1¥ ueS0I9U ,,o[qGu}]dJoe1d-uoU,, oy} 103 Samy ony, 080° 0 940° 0890" 62Gb OLETH' 0 10eh" oeere O6tos Ose 0) 3 ON 6240" 0 ¢190° 0400" bese O9FGh ee Aes Seg ee 5 ea SM09 aptss 69Z0°0 1g¢0° sang G0 OZOFP OFSFF OLFEF OZ6IS' 06209 OSES 0," : "ON | Wily APIO poxrur jo sajdureg 08h" 0 9820 06290" 08690" OG TIF O}60TF 06 se: me “OGrIS" 1e1¢ pee Z ON 9120'0 C840 0 9690 0}69F0' O8z6F 0 “OoL6¥ Gz99" O;Seee' OZE9E'O) == Moo Aasusany) wo | “ON afdumg "YJ 8. Nog ZrS0'0 19Z0° 02080 oer: 1612 OSI" 0 (See Aaa a (4]!a1 Ayz¥a) Z “ON o]dureg 9280°0 1S0' 0,160" ee ee (uimi}80]09) ¢ “ON a]dumug “ypu upun Fy esvI0Ay oaueay IU I oawioay I I eviAy) 1 J I IVaH GNV f Vv mavuaias Rarsan | 2S XV GV OKO | avovm 2g GHEVEIED | — yanouziy 4¥S0 Padma, | MIRAGIY XI NMOOMEIN| \SuGy9 it NAOOUSIN LON NADOULIN Bg _ aqoym ay) fo 1yb1am ay) fo sabvjuaosad sp uaarb yj1wW s,noa pup yuu upuny ur urdu ,97)Q0):d1004d-uOu ayy puD ‘S27D} -ulwoud urnuingp ayy ut pun saynzidieid uyngo)b puv Uuraspo ay) U2 pauLDzUoI uabo.sj1u fo syunown ay; ‘uabosjr 7020) 2y 7 “AI WIAV.L 157 c¥eo'0 | oFeo'o | ogeo'o | spzeo | F920 | cezeo | sereo | e10g"0 | Iez¢"0° |*""¢ dl 80200 | 92z0°0 | o610'0 | geog'o | eg6r'0 | EIzg°0 | OGIS'O | Igz¢'0 | ezIg'0 |" 'F a gsco'0 | 69¢0'0 | 9060°0 | zzog-0 | go0g-0 | seogo | zerg-o | 6z0g'0 | s¢es-0 Le ae eee Pe zse0'O | S8g0°O | O8s0'O | SISO | 6E8F'O | T6ZF'O | SEIS'O | 6FIG'O | IZI¢'O |°"°Z "ON yyw s.nog 00¢0°0 | S0S0'O | 9640'O | SIZT'O | SzZ1'O | 66910 | 98IZ°O | T16IZ'O | IsIzo |" (yy Ajawo) Z ‘ON opdureg v€L0'0 | F200 | P2200 | 86Z2'0 | ESzZ'O | EFEZO | 29620 | S8OEO0 | S8820 | (um1}80]09) g “ON oduIeg esvI0Ay ll I OBBIOAY Il I eBVIOAY Il I NOILVLIdIaud §,TIVHSUVW @Lvildoaud NOUd ALVULINA NI NADOULIN 8,1TVHSUVYN NI NDOULIN SS “ypu apoym ay? fo sabvjyuaosad Sp Uuaarb y)1Uu $.M09 PUD YIU UDUNY UL WOLfa1ay4) $9zD.1} -pYf 94} UL pud poyjaw s.ypoysuvpy fig saynjidisa.d ay? ut paurmzuod uabo.jzvu fo syunowD ay} puD syuazU0d Uaho1j1Uu 10707 9Y,], ‘A GIGVL G n u 3°) = = me) u a Ca ° am y | fas) n = o = ea) ac) u iso) 2 mo) a 158 Composition of Milk Experiments bearing on the amount and nature of the unknown material contained in milk as well as upon the protein content. Camerer and Sdldner subtracted the sum of the quantities of lactose, fat, ash and citric acid” in their samples of milk from the amount of total solids, and called the remainder so obtained the quantity of the ‘‘protein plus unknown substances.”’ They found that the nitrogen in cow’s milk made up about 16 per cent of the quantity of ‘protein plus unknown substances,” while that in human milk made up only from 11 to 12 per cent.% They then precipitated the protein from milk by alcohol according to Munk’s method, and found that the dry ash-free protein of cow’s milk contained 14.54 per cent of nitrogen; while that from human milk contained 13.64 per cent.* From all these results they con- cluded that human milk contains a considerable amount of un- known substances which contain little or no nitrogen. Camerer and Séldner found, as have other investigators, that aleoho! does not precipitate the nitrogenous substances from milk — so completely as does tannic acid; further, the filtrate from the alcohol precipitation gives a positive reaction with the biuret test and with Millon’s reagent. They have studied the non-fatty material which goes into solution in alcohol when the proteins are precipitated from milk by Munk’s method and find that it is a soft, sticky material very slightly soluble in water but highly soluble in dilute alkali. Dilute hydrochloric acid gives a white flocculent precipitate from the alkaline solutions. The substance contained about 13 per cent of nitrogen, 0.2 per cent of phosphorus _ and 1 per cent of ash. From the high nitrogen and phosphorus _ contents Camerer and Séldner suspect that the material is altered casein," Meigs and his collaborators have gone over the ground covered by Camerer and Sdéldner on the protein and unknown substances in milk: Their most important results are given in Tables VI and vil. 12 Camerer and Séldner: Zeitschr. f, Biol., xxxvi, p. 278, 1898. * Camerer and Séldner: ibid., xxxiii, p. 549 and pp. 562-565, 1896. 4 Camerer and Séldner: ibid., xxxiii, p. 548, 1896; The alcohol precipi- tate from the cow's milk contained, on the average, 11.06 per cent of ash; that from the human milk, 6.29 per cent. '* Camerer and Séldner: ibid., xxxiii, p. 561, 1896. = nee es Edward B. Meigs and Howard L. Marsh 159 TABLE VI. The quantities of total solids and of ‘‘protein plus unknown substances’’ in human milk and cow’s milk given as percentages of the weight of the whole milk; and the percentages of nitrogen which may be supposed to be contained in the “protein plus unknown substances.’’ PERCENT- AGE OF TOTAL SOLIDS PROTEIN | NITROGEN UBSTAN UNKNOWN I | II |Average "Bpeinoae Human milk. : Milk No. 4..... 1.11811 .128 11.123) 1.911 | 11.73 Milk No. 6. .... {11.23811.28511.263| 2.864 | 11.7 bth to Oth day | Milk No. 7...... 1.00211 .06411 083] 1.684 | 12.35 ee «| Mite iter, 10.99710.99710.997| 2.219 | 11.82 Milk No. 10. ... |12.44612.41612.431} 2.415 | 11.78 Milk No. 9, 5th month post partum 12 .670/12 .66012.665} 1.274 12.63 Milk- No. 5, collected near end of | 7 A, | aan | 9 604. 9.569 9.586; 1.221 16.05 Cow’s milk. _ Fa (No. 7......... |13.82013.47113.395, 3.769 | 14.07 Sample at Giaad No. Ses 13 43713 48413 .460 3.744 14.73 milk from grade { NO: ls+---+- 18.1953 18813.191)..8.640 | 15.86 pace No. 15........ 13.34313 31313 328, 3.249 15.58 Pe” * Se No. 16........ |12.45912.50312.481) 3.066 16 .23 (NO. 10...7eue ; 11.688 11 .75411.721, 3.106 16.84 Samples of milk from [No. 1...... 15 26515 .23015.247, 3.976 | 14.14 single thoroughbred { No. 9...... 14.08014.10414.092, 3.674 | 13.50 Guernsey cows..... | No. 13..... 14.035 14.03314.034 3.725 | 15.97 * The average percentage of nitrogen in the ‘‘protein plus unknown substances” in samples 4, 6, 7, 8, 9 and 10 of human milk is 12 per cent. Sample 5 is omitted from this caleulation, as the milk came from very near the end of lactation and was evidently abnormal. The average percentage of nitrogen in the “protein plus unknown substances”’ in the samples of cow’s milk is 15.18 per cent. Table VI gives the amount of total solids and of “protein plus unknown substances”’ in a number of samples of human milk and of cow’s milk. The figures for the “protein plus unknown sub- stances” were obtained in the same way as those of Camerer and Séldner. The average nitrogen contents of the “protein plus unknown substances’’ as calculated from our figures is 12 per cent in the case of human milk and 15.18 per cent in the case of cow’s milk. Composition of Milk 160 f ‘syospaz :10uplog Puv JoIEWIED Ose CBG *7U00 rod 68 tins $,M00 JO pariove Ss Suyuyeurss oy} a ‘ * 76 0G 0} pouoyood Ss] 078} dooId 001j-Ysv oY} H]TUL UBUINY Jo sojduivs ZuyuyeurEs OY} UT “HUT s,Mo0o uy ayeyidyood S104 213 JO 314Zjam OY} JO uso 290d [] pus BayuyeyU0o-Ysw O42 Jo IY B}am oy Jo 3020 10d Y[fur uvuNY uy 0784] dyoo1d ojoy oY} JO YYZJom OY} JO yUdd 10d g oBvIEAG oY} UO—A]PAJOOdSeI HIJUI $,400 94} UT ERE"g Pue ‘gze"O “FIFO PUY AUT usuny ©Y} UF 990°0 PUY $80'0 ‘620°0 ‘80'0 94 OF Puno; sum 47 /HITUI s,M09 JO GT PUY gE] ‘6 Se[dUIeS Jo ses¥o O43 UT PUL “AIIUI UBUINY Jo 1 pue OT “6 ‘g Se[duIuS Jo saseD oy} Uy Ysieyy Aq pouyuliojop SVM so}ez]}dyooId [OYyoo]e oy} uy Yse oy, *sezez}djoo1d 901j-Ysv oy} JO 4yZjoM OY} Jueseide1 UNICO SPY} UT SeuNSy oul , 90° FT 96° €T 6I'FL | £9cb'O | ESzF'O | 92zF'0 0€0°€ L40'€ £10 € | 61 ON sa JOyoo]e jus Jad GE FI I FI €8° FI | Zg¢9°0 ISto'O | H9FS' 0 TI8' € 808° € FIS'€ | ST “ON ATW 98 Aq pozeqidisa1g 09 $1 69° ST co ST | 9I9F'0 | 9E9F'O | 96SF'0 €68° € GIP € Glee | °° 6 ON AA f aad Tv FI 09° FT | Sa rT | €16F'O | 6F6F'O | 8Z8h'0 OI C68 € ‘60h € 1° SON un patios. 16° §T £0' FI 16°81 | OLZ¢'0 | OFZF'O | OO8F'O Clr € cles Osh € | °° 2°ON ATA ‘yJIU $ NOg ‘or | 28" veto | wero | veet'0 | ¥26°0 | zg60 | I66'0 |°--2-ON seus NOPE Quresed IT STI LE° ST L8°OI | SLP ¥ Noll a 06 a, or the crystal is optically positive (+). On examining the crystal in the direction of the vertical axis, c, in con- vergent polarized light, the uniaxial interference figure is observed, thus confirming the tetragonal character. As a check on the angles, the angle over the pole was measured, angle 111 ATI1 = 32° 40’ as against 32° 4’ calculated. The profile view of the crystal lying on. the pole edge gave an edge angle over the pole of about 54°. These crystals are fairly soluble in ether, somewhat soluble in 95 per cent alcohol, and nearly insoluble in water, though treat- ment with water alters them, rendering them opaque and in time destroying their crystalline form. They dissolve very slowly in acetone. They have been submitted to Lassaigne’s test for nitro- gen, and thereby shown to be free from nitrogen; but the appli- 164 Composition of Milk cation of the plumbic acetate test! shows that they contain a considerable amount of sulphur. They represent a substance, which, so far as we know, has not been heretofore described as a component of milk. _ The immediately preceding paragraphs may be summed up by saying that the results of Meigs and his collaborators on the protein and unknown substances of milk are throughout in close agreement with those of Camerer and Sdéldner. Both sets of observers find that the total nitrogen in human milk makes up a decidedly smaller proportion of the “protein plus unknown sub- stances” than it does in the case of cow’s milk; both sets of ob- ‘servers find that the materials precipitated from the two kinds of milk by alcohol have about the nitrogen content which is usual for protein; finally, both sets of observers obtained similar mate- rials from the alcohol filtrates. This agreement is made all the more striking by the fact that neither Meigs nor any of his col- laborators was familiar with the work of Camerer and Sdéldner at the time when their experiments were carried out. They ob- tained their results quite independently. We are inclined, there- fore, to regard it as established, or at least very highly probable, that human milk contains a considerable amount of unknown material, which has either a low nitrogen content or else none at all of that element. What becomes of this material when milk is subjected to analy- sis? It has been shown above that unknown substances can be extracted from milk by means of alcohol, and we shall consider very briefly the question whether the material of the alcohol ex- tract is to any extent identical with the “xz material,” the exist- ence of which in human milk can be inferred from the work of Camerer and Séldner and of Meigs and his collaborators. Camerer and Séldner advance the hypothesis that the mate- rial of the alcohol extract is ‘altered casein.’”’ We think this improbable for the following reason. Cow’s milk contains about three times as much casein as human milk. If, therefore, alcohol alters casein so that a part of it becomes alcohol-soluble, more of this altered casein should appear in the alcohol filtrate from '* Hawk: Practical Physiological Chemistry, 4th edition, Philadelphia, 1912, pp. 108, 109. Edward B. Meigs and Howard L. Marsh 165 cow’s milk than in that from human milk. But just the oppo- site is the case. We have endeavored to get an approximate idea of the total amount of unknown material in the alcohol extract from cow’s milk and human milk. Two samples of human milk and three samples of cow’s milk were freed from fat by Meigs’ method, and the protein was precipitated from the fat-free residue by 86 per cent alcohol. The filtrates from the 86 per cent alcohol protein precipita- tions were then evaporated to dryness and the weights of the ash-free residues were determined. These residues contained all the lactose from the milk plus the unknown alcohol-soluble mate- rial. In other portions of the same samples of milk the lactose was determined by the Fehling method—see Table III. Sub- tracting the amounts of lactose from those of the ash-free residues in the five samples of milk gives the following results: Percentage of Percentage of ash-free alcohol Percentage of mame glooy ” material Hine tatth te 10:42%, 7.952 _ 6.574 = 1.378 No. 9..... 7.575 _ 7.080 = 0.495 No. 16..... 5.104 _ 4.713 = 0.391 Cow’s milk... ; No. 19..... 5 .068 4.482 = 0.586 Wo. 13. Fas 4.935 _ 4.748 = 0.187 The following is another method by which an approximate idea of the amount of unknown alcohol-soluble material in milk may be gained. In the samples of human milk Nos. 4, 5, 7, 8, 9 and 10 the weights of the material precipitated by alcohol have been determined (Table VII) and may be compared with the quan- tities of “protein plus unknown substances” found in Table VI. It will be seen that the quantities of the latter are always con- siderably larger than those of the former; the results obtained by subtracting the quantities of ash-free alcohol precipitates from those of the ‘‘protein plus unknown substances” are as follows: Milk No. 4, 0.853; No. 7, 0.610; No. 8, 0.916; No. 10, 0.981; No. 9, 0.485; No. 5, 0.331. The same data for the samples of cow’s milk, Nos. 7, 8, 9, 18 and 19 are given in Tables VI 166 Composition of Milk and VII; and the results of similar calculations are as follows: Milk No. 7, 0.357; No. 8, 0.334; No. 9, 0.281; No. 19, 0.076.19 The figures which have just been given to represent the amounts of the unknown alcohol-soluble material in human milk at dif- ferent periods of lactation correspond very fairly with the figures for the amounts of the x material at the corresponding periods of lactation as calculated from the data of Camerer and Séldner and from those of Meigs and his collaborators. It seems prob- able, therefore, that the alcohol-soluble material is to a consid- erable extent identical with the x material. That it is not wholly identical with the x material is evident from general consider- ations as well as from its high nitrogen content. Milk contains such bodies as urea, ammonia,?® and purine bases** which would not be precipitated by alcohol; and the nitrogen in these would account for a considerable part of the nitrogen in the alcohol- soluble material. The known alcohol-soluble substances in milk. Koch and Woods” find 0.036 to 0.049 per cent lecithin and 0.027 to 0.045 per cent kephalin in cow’s milk—0,.072 to 0.086 per cent total “lecithans,”** For human milk the figures of these investigators are lecithin, 0.041 per cent; kephalin, 0.037 per cent; total lecithans, 0.078 per cent. - Raudnitz™ gives a review of previous determinations of cho- lesterol® in milk. Tolmatscheff found 0.0252 to 0.0885 per cent of this substance in human milk; Schmidt-Miilheim was able to demonstrate its presence in cow’s milk. Marsh determined the amount of cholesterol in cow’s milk according to the method of Ritter® and found 0.021 per cent. * 1° In the case of sample No. 13 of cow’s milk there was some error in anal- ysis so that the weight of the material precipitated by alcohol appeared to be slightly greater than that of the ‘protein plus unknown substances.” 2° Camerer and Séldner: Zeitschr. f. Biol., xxxvi, p. 299, 1898. * Raudnitz: Ergeb. d. Physiol., 1903, ii, p. 257. * Koch and Woods: This Journal, i, p. 211, 1906. * By “lecithans’’ Koch and Woods mean the phosphorus-containing lipoids. * Raudnitz: Ergeb. d. Physiol., 1908, ii, p. 264. ** Raudnitz uses the older term, ‘‘cholesterin.”’ * Ritter: Zeitachr. f. physiol. Chem., xxxiv, p. 430, 1901-02. =< ie e! a a Edward B. Meigs and Howard L. Marsh 167 The substances which have been spoken of above—lecithin, kephalin, cholesterol, urea, ammonia and the purine bases—would account for less than one-fifth of the unknown material in the alcohol extract of human milk. It is realized that the data given above and bearing on the quantity of the material in the alcohol extract of milk are very incomplete. But they do indicate that human milk contains con- siderably more non-fatty, alcohol-soluble material than cow’s milk; and it has been thought worth while to present them in view of recent interesting work on the importance of unknown substances in milk as accessory factors in diet. Stepp,?’ for instance, has found that mice always die in a few weeks when given food mate- rials which have been fully extracted with alcohol and ether, and that the diet may be rendered again capable of sustaining life by the addition of the material from the alcohol-ether extract of skimmed milk. Stepp has been able to show that the mate- rial necessary to maintain life is not either fat, cholesterol, lecithin or salts. GENERAL CONCLUSIONS. Human milk differs from cow’s milk in three important ways. It contains considerably more lactose than cow’s milk, and more substances of unknown nature which contain little or no nitrogen; it contains very much less protein than cow’s milk. The com- position of milk varies more or less regularly with the progress of lactation so that average figures for its composition are not very satisfactory. The following, however, may be taken as the limits of normal variation of the constituents of the two kinds of milk from the beginning of the second month of lactation onward, the figures representing percentages of whole milk: UMS IMU... oR ROW MMMEIER oo, ka ee | Both kinds of milk contain substances which are important constituents of diet, which are soluble in alcohol and ether, which *7 Stepp: Zeitschr. f. Biol., |vii (N. F. xxxix), p. 135, 1911. 168 Composition of Milk contain little or no nitrogen, but of which the chemical nature is still unknown. These substances are most plentiful in early human milk and diminish in amount with the progress of lacta- tion. Early human milk contains about 1 per cent of these un- known substances; milk from the middle period of lactation about 0.5 per cent. Cow’s milk from the middle period of lactation contains about 0.3 per cent of the unknown substances. THE INFLUENCE OF THE ADMINISTRATION OF CREA- TINE AND CREATININE ON THE CREATINE : CONTENT OF MUSCLE. By VICTOR C. MYERS ann MORRIS S&S. FINE. (From the Laboratory of Pathological Chemistry, New York Post-Graduate Medical School and Hospital.) (Received for publication, September 2, 1913.) With the appearance of an accurate method of estimating creatinine, the story of the origin of this interesting urinary con- stituent has been found to be less simple than was formerly sup- posed. Without doubt, the information which has been brought to light with its aid has been very important, although in many _instances this has been contrary to the previously accepted views, and difficult of satisfactory explanation. That the creatinine of the urine has its origin in the creatine of the muscle seems obvious on a@ priori grounds, but the satisfactory demonstration of this has been beset with unforeseen difficulties. Folin,? with the aid of his colorimetric method, was the first to point out that the quantitative conversion of creatine to creatinine or creatinine to creatine 2n vitro was far more difficult than previous statements would lead one to believe. That the body was even less able to bring about a quantitative conversion of creatine to creatinine was shown by his feeding experiments on man. In fact he was unable to adduce any evidence of a conversion of the administered creatine to creatinine. Further experimental data bearing on the fate of administered creatine and creatinine have been presented by Klercker,? Wolf and Shaffer,* van Hoogenhuyze and Verploegh,® 1 For an interesting critical review of the general subject of creatine- creatinine metabolism, reference may be made to the paper of Riesser: Zeitschr. f. physiol. Chem., \xxxvi, p. 415, 1913. 2 Folin: Hammarsten’s Festschrift, III, 1906. ’Klercker: Beitr. 2. chem Physiol. u. Path., viii, p. 59, 1906; Biochem. Zeitschr., iii, p. 45, 1907. . 4 Wolf and Shaffer: this Journal, iv, p. 439, 1908. 5 van Hoogenhuyze and Verploegh: Zeitschr. f. physiol. Chem., lvii, p. 161, 1908. ; 169 170 Administration of Creatine and Creatinine Lefmann,* Plimmer, Dick and Lieb,’ Pekelharing and van Hoogen- huyze,® Foster and Fisher, Towles and Voegtlin,!° Folin and Denis," and Kraus.” It has been observed that when creatinine is admin- istered to man or animals, either per os or parenterally, about 80 per cent reappears in the urine during the succeeding twenty- four hours. After the administration of creatine, the excretion of creatinine has been found to be only slightly increased, if at all. When given in comparatively large amounts a considerable portion may reappear in the urine as such, though with small amounts it does not reappear. In contradistinction to creatinine, Folin has viewed the creatine under these conditions not as a waste product but as a food. Since the administered creatine cannot be accounted for by the creatine (and creatinine) of the urine, it seems logical to assume that this remaining portion is either metabolized or stored up in the muscle. This last possibility was considered by Mellanby™ in experiments on rabbits and chickens. He concludes, ‘‘Creatine and creatinine feeding has no effect upon the creatine content of. muscle after the muscle has reached a certain saturation point. There is some evidence of an increase in muscle creatine at an early stage of life on feeding with creatine and possibly with creat- inine.” In an endeavor to throw light upon the origin of creatine, Riesser,"* in his recent studies, observed a slight increase in the concentration of muscle creatine after the administration of betaine and choline. The possibility that the creatine, which does not reappear in the urine after its introduction into the body, is deposited in the muscle tissue, has not been answered, and it was primarily for this reason that the present experiments were undertaken. In his experiments Mellanby took no account of the portion of the ad- *Lefmann: Zeitschr. f. physiol. Chem., |vii, p. 476, 1908. 7 Plimmer, Dick and Lieb: Journ. of Physiol., xxxix, p. 98, 1909. * Pekelharing and van Hoogenhuyze: Zeitschr. f. physiol. Chem., \xix, p. 395, 1910. * Foster and Fisher: this Journal, ix, p. 359, 1911. ‘© Towles and Voegtlin: ibid., x, p. 479, 1912. ' Folin and Denisg tbid., xii, p. 148, 1912, * Kraus: Arch. of Int. Med., xi, p. 613, 1913. Mellanby: Journ. of Physiol., xxxvi, p. 470, 1908. Loe, cit, Victor C. Myers and Morris S. Fine 171 ministered creatine or creatinine that might beJost intheurine. Ob- viously, this is a factor of importance in the interpretation of the content of muscle creatine. Such analyses were included in the present experiments, in order that we might ascertain in so far as possible the fate of all the administered creatine and creatinine. METHODS. The general analytical procedures employed in the analysis of tissue and urine were those previously described. The creatine used in the first experiment (rabbit 47) was prepared from rabbit muscle. In the succeeding experiments, however, the creatine and creatinine were prepared from urine according to unpublished directions supplied us by Prof.S. R. Benedict. With these methods we have prepared about 100 grams of creatine and 10 grams of creatinine, both of the highest degree of purity. Our thanks are due to Professor Benedict for furnishing us with the details of these admirable methods. In all experiments the creatine and creatinine were administered subcutaneously in 1 per cent aqueous solution between the shoulder blades with a 10 cc. Luer glass syringe. In every case the strength of the solution employed was ascertained colorimetrically, although in several experiments this was checked by employing carefully dehydrated preparations, which could be accurately weighed. The time of administration was generally from 2-5 hours after the end of the previous twenty- four-hour period. The injections extended over a period of 5-13 consecutive days. It was believed that in this way the maximum retention of creatine in the muscle would be observed, and further that the average daily composition of the urine during this interval would furnish a very reliable comparison with that of the control period. There appeared to be a lag in the elimination of urine in one or two cases, notably in rabbit 71, although it is improbable that this could have had an important influence upon the average data of a period of several days. Nitrogen determinations were made on the urines of all rabbits with the exception of rabbits 71, 72,73 and 74. The daily variations were so slight as to fall within ‘the limits of error; and as they are without significance for our present purposes, they have not been included in this paper. 15 Myers and Fine: this Journal, xiv, p. 9, 1913. 172 Administration of Creatine and Creatinine EXPERIMENTAL PART AND DISCUSSION. Eleven injection experiments with creatine and creatinine are reported upon rabbits, eight with creatine and three with creati- nine. Two of the eight experiments with creatine (numbers 71 and 72) did not include estimations of the creatine of the muscle, while one experiment, number 57, was carried out on a rabbit fed upon a practically pure carbohydrate diet."* The important results of these experiments are summarized in Tables I and II, the experi- mental details being recorded in the Tables III—XII appended to this paper. The creatine employed in experiments 47, 49, 56, 59 and 57 was tested qualitatively for creatinine and failed to give an appreciable reaction. Following the period of injection, the creatinine excretion was found to be slightly increased, the equi- valent of a conversion of about 3 per cent of the administered creatine. On this account, three additional experiments, 71, 72, and 73, were conducted employing creatine, the purity of which had been quantitatively ascertained. Four colorimetric tests were made with this preparation during the course of the experi- ments, employing in each case 0.l-gram portions. The colori- metric readings indicated that the possible contamination of creatinine was not over 0.15 to 0.20 per cent, an amount insig- nificant in this connection. It is interesting to note that even with this very pure preparation of creatine there was an increased elimination of creatinine, if anything greater than that observed in the earlier trials. The data on the influence of the administration of creatine and creatinine on the creatine content of the muscle are summarized — in Table I. The results for the nitrogen and moisture content of the muscle are very uniform and all fall within normal limits. The figures for the creatine content are, however, uniformly above the usual normal figure of 0.52 per cent, if we except carbohydrate rabbit 57; and even here the creatine content is considerably higher than in other animals fed carbohydrate but no creatine.” The increase after the creatine injections in the first five experiments amounts to about 5 per cent.'* It is true that this increase is slight See Myers and Fine: this Journal, xv, p. 305, 1913. 7 Loc. cit, '* This is the percentage increase over the normal 0,522 per cent. The absolute amount of creatine retuined by the body is calculated by mul- Victor C. Myers and Morris S. Fine 173 TABLE I. The creatine content of rabbit muscle as influenced by the subcutaneous admin- istration of creatine and creatinine. CREATINE CONTENT LENGTH OF OF BODY COMPOSITION OF MUSCLE TIME AFTER sora | wetent | Calculated ‘nox aE seid ae Found Nitrogen | Moisture | Creatine cng dm Experiments with creatine. ’ kgms. grams grams per cent per cent per cent days 47 1.65 3.00 3.02 3.71 76.1 0.544 3 49 1.72 3.13 | 3.47 3.61 75.9 0.546 3 56 1.81 3.29 3.70 3.35 77.1 0.559 3 59 2.00 3.64 3.90 8.52 76.7 0.553 4 73 1.66 3.02 3.04 3.63 75.3 0.540 1 57¢ |1.98-1.22) 3.60 2.18 3.55 76.2 0.482 1 Experiments with creatinine. 58 1.68 3.06 3.49 3.55 76.3 | 0.540 3 62 1.70 3.02 |» 3.09 3.45 76.4 0.566 4 74 1.93 3.51 4.04 3.59 76.0 0.566 1 * Calculated from average data in Table VII of a previous paper, this Journal, xiv, p. 23, 1913. + See Myers and*Fine: ibid., xv, p. 305, 1913. and we would attach greater significance to the fact that a large amount of creatine was not stored up in the muscle than to the fact that a small amount was retained. This slight increase we believe, however, to be a little beyond the limits of error of the method when carefully carried out. Many colorimetric readings were made and compared with readings from control animals. The controls were found under the conditions of our experiments to give the usual readings of about 9.0 mm., whereas the samples from the animals to which the creatine and creatinine had been administered gave readings varying between 8.3 and 8.7 mm. As a further control, it was found that creatine added to muscle before extraction could be quantitatively recovered under the same conditions. The increase in the creatine content of the tiplying the increase per gram of muscle by the total muscle tissue of the body. The latter is estimated by dividing the total body creatine by the percentage found. 174 Administration of Creatine and Creatinine muscle is, if anything, more pronounced after the administration, of creatinine than of creatine. It would not be expected that creatinine would be retained by the muscle unchanged, and in the case of rabbit 74 a creatinine estimation” on the fresh muscle showed the presence of about 9 mgms. of creatinine per 100 grams of mus- cle, an amount not far from the normal. Rabbit 73, to which creatine had been given, contained 7 mgms. creatinine per 100 grams of muscle. It seems evident from the above trials that the creatine content of the muscle tissue can be raised slightly above the ordinary level by the administration of either creatine or creatinine. That creatinine exerts this influence is in harmony with the view that the reaction between these two substances is reversible. This is further borne out by the observations of other workers of an excretion of creatine following the administration of creatinine. The increase in the concentration of muscle crea- tine in rabbits, which Riesser®® observed after the injection of betaine and choline is quite similar to that observed in our own experiments; and it is further of interest that in his six control animals he obtained figures identical with those which we have observed, viz., 0.52 per cent.” Further evidence of a small retention of creatine is found in the comparison of the total content of creatine in tlfe body cal- culated from the body weight with that actually found to be present, the latter always being the greater. The error incident to this comparison is, of course, considerable, but, nevertheless, the fact that the figures based upon actual determinations are uniformly higher than those calculated lends support to the above view. It would seem, then, that by the administration of either creatine or creatinine the concentration of muscle creatine may be raised about 5 or 6 per cent above the values ordinarily obtained. That the extra creatine is quite firmly held is indicated by the fact that the concentration of the creatine in the muscle is prac- tically the same whether the rabbit is killed one day or four days after the last injection. The various factors concerned in the fate of creatine and '° The method employed will be described in a subsequent paper. 2 Loc, cit. * Myers and Fine: this Journal, xiv, p. 14, 1913, Victor C. Myers and Morris S. Fine 175 TABLE II. The fate of creatine and creatinine when administered subcutaneously in the rabbit. FATE OF ADMINISTERED Socewsell CREATINE eae OF PERIOD le ATT Py & cxmast, | giltie | omar | Tor”, (aces) 92 4(8¢ |35,| 3 CREATININE PER KGM. | § Es a 2 32 g INJECTED ee ae & ag = : 3 a| gs at? |at? | 3° en Experiments with creatine. kgms grams days mgms. | percent| percent| per cent | per cent 47 1.65 0.95 5 115 60 4 13 23 49 1.72 1.02 12 49 26 7 15 52 56 1.81 1.06 10 59 32 2 23 43 59 2.00 1.22 9 85 67 1 18 14 71 2.55 0.90 6 59 53 10 72 2.45 0.90 6 61 47 14 73 1.66 0.70 7 60 81 7 14 0 57 1.98-1.22) 1.09 13 52 58 2 Experiments with creatinine. 58 1.68 0.75 9 50 0 82 13 5 62 1.70 0.70 6 69 0 80 28 0 74 1.93 0.90 6 78 0 77 26 0 creatinine when introduced into the body are recorded in Table II. When creatine is administered subcutaneously to rabbits in amounts varying between 50 and 100 mgms. per kilogram of body weight per day, a considerable portion (25-80 per cent) reappears in the urine unchanged. In most of the trials about 3 to 4 per cent appeared in the urine as creatinine, although the amount ex- creted in this form may be considerably greater, as in rabbits 71 and 72. A possible explanation for these high results of 10 and 14 per cent may be found in the fact that these animals were very large and had ‘‘creatinine coefficients’? of 16 and 17 respec- tively. As we have previously pointed out,” the average creat- inine coefficient for the rabbit is 14, and these higher coefficients possibly indicate a greater efficiency on the part of these animals in the conversion of creatine to creatinine. 22 Myers and Fine: this Journal, xiv, p. 19, 1913. - 176 Administration of Creatine and Creatinine Van Hoogenhuyze and Verploegh were apparently the first to call attention to the slightly increased excretion of creatinine after the administration of creatine, an observation confirmed by Pekelharing and van Hoogenhuyze, Towles and Voegtlin, S. R. Benedict® and the present writers. Although creatine and creat- inine may not have the very simple metabolic relationship for- merly supposed, it is not quite obvious how Klercker and Lef- mann from their data draw the conclusion that exogenous creatine is not transformed to creatinine at least in small part. In a previous communication,™ attention was called to the sig- nificant fact that the percentage of administered creatine which was converted to creatinine was not widely different from the relationship existing between the daily creatinine of the urine and the total body creatine. It was further suggested that, since creatine is held in such a loose state of combination in the muscle, it is not illogical to believe that it experienced the same fate as the administered creatine. Under the conditions of our experiments, it appears that about 15 per cent of the injected creatine may be stored in the muscle. The portion of the creatine remaining unaccounted for seems to be chiefly dependent upon the amount of the creatine admin- istered, or, in other words, upon the opportunity given the body to oxidize it. The experimenta] data in the case of creatinine seem to indicate that the 20 per cent which is not excreted in the urine may be completely stored up in the muscle as creatine. CONCLUSIONS AND SUMMARY. The subcutaneous administration of creatine to rabbits appears to cause a small increase in the creatine content of the muscle, about 5 per cent in five experiments. This is quite insufficient, however, to account for the creatine which does not reappear in the urine. The administration of creatinine appears to exert a similar in- fluence upon the creatine content of the muscle. In three experi- ments the creatine content was found to be about 6 per cent above the normal, an amount sufficient to account for the creatinine » * Private communication. * Myers and Fine: this Journal, xv, p. 304, 1918. Victor C. Myers and Morris S. Fine 177 not eliminated by the kidneys. This apparent increase in the creatine content of the muscle was not due to a retention of the creatinine unchanged. Of the creatine administered in our experiments, 25-80 per cent— the quantity depending upon the amount injected—reappeared in the urine unchanged, while from 2-10 per cent was eliminated in the form of creatinine. We are inclined to attach considerable significance to this small conversion of creatine to creatinine, as throwing light upon the relationship of these two substances in metabolism. When creatinine was administered 77-82 (average 80) per cent reappeared in the urine. No elimination of creatine was detected. TABLE III. Rabbit 47—Injection of creatine. URINE CREATINE DAY BODY CREATINE RECOVERED ees Volume | Creatinine| Creatine oe Odum kgms, ce. mgms. mgms. mgms. per cent 1 1.60 280 57 0 oF 2 1.61 | 270 58 0 » 3 1.65 | 230 57 0 4 1.67 | 235 58 0 5 1.69 250 54 0 6 1.68 210 - 53 0 7 1.63 285 60 0 8 1.63 260 60 ey MIVOPORG. EE, .. wna sccaes sss 57 | 9 1.65 260 60 244 328 74 10 1.66 | 250 66 237 | (306 77 ll 1.68 285 67 35 131 27 12 1.65 225 70 26 109 24 13 1.67 230 61 30 77 39 ORM chin... ee ccomae» 65 114 190 60 14 1.66 235 68 0 15 1.69 260 - 61 0 Female albino, received 350 grams carrots daily. On days 9, 10, 11, the creatine was given in 3, 4 and 2 doses respectively, in the course of 4 to 6 hrs. Skinned and eviscerated carcass—870 grams. 178 Administration of Creatine and Creatinine TABLE IV, Rabbit 49—Injection of creatine. Ache BODY bd CREATINE Phi soins 1 pews Volume | Creatinine} Creatine mans wetbabescsicd kgms. ce. mgms. mgms. mgms. — per cent 1 ita 210 54 0 2 1.82 240 54 0 3 1.79 155 54 0 4 | 1.74 | 255 51 0 BROUESC..... Sickspagawewens 53 5 295 60 7 68 ] 6 1.70 295 54 23 68 | 7 1.68 255 59 5 68 18 8 1.70 295 65 10 68 9 1.70 | 295 56 14 68 : 10 1.70 | 225 60 16 77 } 11 215 | 60 11 85 ) 12 1.78 250 65 41 120 13 1.70 240 54 31 85 > 31 14 1.69 225 61 34 134 15 1.70 260 54 55 134 16 1.72 | 225 53 16 45 a ee 58 22 85 26 17 1.74 260 52 0 18 1.74 275 §2 0 19 1.75 275 50 0 Grey female, received 350 grams carrots daily. Skinned and eviscerated carcass—964 grams. | { ag | ue »|* \~ Victor C. Myers and Morris S. Fine 179 TABLE V. Rabbit 56—Injection of creatine. URINE CREATINE DAY RODE CRRATING RECOVERED Ni ea: Volume | Creatinine| Creatine oe OC URINE kgms. ce. mgms. mgms. mgms. per cent . 1 1.87 230 64 0 , 2 1.88 285 65 0 3 1.88 805 65 0 4 1.88 315 69 0 5 1.85 310 71 0 5 EVO. 2; . ss. scceueeeak 67 | 4 6 1.85 | 270 69 32 89 } ai +! 7 1.85 280 69 35 89 P i" 8 1.85 330 81 75 179 » 42 9 1.81 280 70 5 0 10 1.80 350 70 24 134 11 1.79 315 71 52 134 30 12 1.80 285 63 44 134 13 1.81 335 78 13 89 } 21 14 1.83 290 54 24 89 e 15 1.82 290 55 30 116 26 MWOFARG. oe... ccs aah tees cs 68 33 105 32 16 1.83 325 66 0 AN 1.80 285 61 0 18 1.80 805 70 0 Brown male, received 350 grams carrots daily. On the 8th day the creatine was given in two equal doses at an interval of 6 hours. Skinned and eviscerated carcass—1025 grams. 180 Administration of Creatine and Creatinine TABLE VI. Rabbit 59—Injection of creatine. | eae | popy CREATINE Phe sanwcstrah | eoeead Volume | Creatinine; Creatine pasties gies kgms. cc. mgms, mgms. mgms. per cent P14 207 295 80 0 2° | 2.07 |. 190 78 0 3 2.12 160 72 0 4 | 2.12 240 72 0 5 | 20 200 78 0 CS CEE 76 6 2.10 | 265 90 38 90 7 2.04 250 74 39 90 52 8 2.01 280 69 62 90 9 2.00 305 80 88 135 10 2.00 | 295 68 95 135 67 1 1.98 | 1853.70 04 135 12 2.06 | 280 | 81 133 181 13 2.00 350 80 146 181 75 14 1.98 330 77 128 18h ae Average ..:....gan/>. -Gmeees 77 91 135 a ee 15 | (1.98 | 886 71 0 F 16 1.95 815 74 0 a 7 | 1.95 | B15 80 0 i 18 1.95 280 67 0 Black female, first seven days fed 350 grams carrots and 400 grams on remaining days of experiment. Skinned and eviscerated carcass—1072 grams. TABLE VII. Injection of creatine. URINE CREATINE DAY BODE CREATING RECOVERED bifeae ind Volume | Creatinine | Creatine acai a Rabbit 71. kgms. ce. mgms. mgms. mgms. per cent 1 2.61 180 108 0 2 2.58 85 109 0 i 3 2.56 | 115 109 0 4 2.63 125 108 0 5 2.63 150 107 0 PPOPAMS sc cseseseecccyes,-| 108 0 6 2.62 185 133 50 100 7 2.57 100 102 30 100 34 8 2.57 80 108 23 100 9 2.55 175 123 135 200 10 2.51 200 131 125, 200 63 11 2.52 140 107 110 200 WOPONO: 05... . vcheebinea st 117 79 150 53 12 2.50 180 108 7 13 2.50 205 129 Rabbit 72 ta 1 2.43 | 230 111 0 ® 2 2.44 50 108 |' 0 3 2.48 9} 111 0 4 2.48 190 112 0 5 2.45 | 270 | 118 0 WVOPARO eaia ss icaese ss 68 112 0 6 2.45 240 128 32 100 7 2.43 275 141 20 100 30 8 2.40 220 131 38 100 9 2.43 250 122 116 200 10 2.40 290 133 103 200 56 ll 2.48 230 128 118 200 AVOTIMING cody ss. 06s css 130 71 150 47 12 2.40 255 122 0 13 2.40 270 119 0 Both animals were males. Rabbit 71 ate 180-350 grams carrots daily; F rabbit 72, 350 grams carrots daily. 4 . On days 9, 10, 11 in both experiments the creatine was given in two doses at intervals of 2-4 hours. 181 182 Administration of Creatine and Creatinine TABLE VIII. Rabbit 73—Injection of creatine. ' BODY URINE CREATINE CREATINE DAY WEIGHT wane re INJECTED a ee kgms ce. mgms. mgms. mgms. “per cent 1 1.66 125 61 0 oe os) 110 61 0 . |) eo 40 56 6 a: | 2 80 60 10 5 1.68 125 61 ce Serre 60 5 6 1.65 | 180 67 74 | 100 7 1.65 180 75 120 100 8 1.65 150 | 65 101 100 9 1.66 200 | 66 100 100 10 1.67 235 62 85 100 11 1.67 230 66 65 100 12 1.66 220 61 56 100 AVerawe .. ss. sckbes> omnes 08 66 86 100 81 Female albino, ate 140-350 grams carrots daily. The elimination of small amounts of creatine in the first period suggests M the possibility that growth had not been completed. Skinned and eviscerated carcass—847 grams. Victor C. M S yers and Morris S. Fine 183 TABLE IX. Rabbit 57—Carbohydrate feeding; Creatine injection. oer | a DAYS wright : z |. | g | SRBATINE |necovenz & 3 3 g E IN URINE o | i ree kgms. grams cc. gram | mgms.| mgms. mgms. per cent 1-7 1.92 | 22 80 0.44) 63 8 8-12 1.57 20 60 | 0.41 69 7 13 20 | 70 90 |) 14 20 70 90 15 20 70 90 16 1.39 | 20 | 60 |}0.23 68 | 40 90 |} 33 17 20 50 18 20 50 90 19 20 | 6 |) | 90 || 20 1.34 | 20 | 50 |0.30| 76 | 56 90 21 20 50 | 0.20) 81 65 90 22 20 | 50 |0.22| 56 | 39 90 23 1.38 20 50 | 0.40| 87 .| 127 90 20-23 : 75 72 90 60 (Average) : + 24 20 | 50 |0.29| 68 38 0 25 20; 50 |0.26| 65 | 20 0 26 20 | 50 | 0.22) 51 13 0 24-26 61 24 (Average) 27 1.24 | 20 | 50 |0.30| 66 | 67 90 28-29 1.24 20 50 | 0.16 42 51 45 (99) (Average) 27-29 50 56 60 60 (Average) White and black female. From 12th to 17th day diarrhoea was present although by frequently compressing the bladder, good urine samples were obtained. After the 17th day no diarrhoea was observed. On the 29th day, the animal still appeared to be in good condition, but it was thought best to kill it so as to avoid the acute changes immediately preceding death. Skinned and eviscerated carcass—687 grams. 184 Administration of Creatine and Creatinine TABLE X, Rabbit 58—Injection of creatinine. URINE CREATININE ae | waar | “TgzcTep | RECOVERED kgms. ce. mgms. mgms. mgms. per cent 1 1.79 355 68 0 2 1.76 310 66 0 3 250 67 0 4 1.76 335 69 0 5 By | 290 67 0 6 1.70 315 69 0 7 1.79 260 64 0 67 | 8 1.68 80 134 0 83 9 1.58 0 83 10 1.65 0 83 ll 2.68 0 83 12 1.67 a) 83 | 13 1.67 0 83 14 1.68 0 83 j 15 1.66 0 100 ; 16 1.68 0 75 SH: : Average....... ce 134 84 82 a 17 1.65 | 250 54 0 3 18 1.68 265 64 0 19 1.68 260 65 0 Brown male, ate 350 grams carrots daily. On the 8th day, drougill an oversight, the animal was not fed, which accounts for the low volume of urine on this day. ; Lg “a aS OOS EE Ore er 185 TABLE XI. Rabbit 62—Injection of creatinine. URINE CREATININE a BODY | CREATININE | acovennp hela 2 Volume | Creatinine| Creatine gs ee URINS kgms. ce. mgms. mgms. mgms. per cent 1 1.70 240 64 0 2 1.73 230 62 0 3 1.79 240 66 0 a 1.82 260 63 0 PS ee ee 64 5 1.85 | 320 | 146 0 83 3 6 1.80 | 340 130 0 83 84 7 1.66 250 137 0 83 a 8 1.66 255 157 0 124 9 1.66 260 | 184 0 165 75 10 1.66 305 | 192 0 165 ; Averag@Miy:..cieccies..0) 159 117 80 il 1.64 280 57 0 12 1.65 230 59 0 13 1.66 265 59 0 14 1.65 195 59 0 Black and white female, ate 350 grams carrots daily. Skinned and eviscerated carcass—847 grams. Eamets» TABLE XI. Rabbit 74—Injection of creatinine. | Bopr | CREATININE pains DAYS | waicet | a INJECTED it Cae kgms. ce. mgms. mgms. per cent 1.95 | 235 0 1.96 210 0 1.97 175 0 1.91 65 0 1.92 165 0 ae a 1.95 e 0 1.94 | | —~0 1.96 | e 0 1,91 5 0 1.91 v 0 1.90 | 295 0 thea eae be . 203 er ry Male rabbit; ate 350 grams carrots daily. Skinned and eviscerated carcass—1084 grams. 7&7 THE FATE OF PROTEIN DIGESTION PRODUCTS IN THE BODY. Il. DETERMINATION OF AMINO NITROGEN IN THE TISSUES. By DONALD D. VAN SLYKE. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, August 27, 1913.) The methods will first be described. In the latter part of the paper some particular points will be discussed in detail. I. Method for closest absolute results. The sample of tissue, — which may weigh from 5 to 30 grams, is immediately after exci-— sion weighed to within 0.01 gram, covered with boiling hot water, to which 1 cc. of 50 per cent acetic acid per liter is added, and heated on the water bath until the proteins are completely coagu- lated. In case the piece is a large one it is cut into several with a pair of scissors, in order that it may heat through more readily. Coagulation is complete in twenty or thirty minutes. The pieces are then lifted out of the water with forceps or crucible tongs, * minced fine with a food macerator, and returned to the same water. - The heating is continued for about ten minutes, with occasional stirring, and the supernatant liquid is then decanted through a filter of glass wool. The tissue pieces are covered with a fresh portion of boiling acidified water, using 5 or 10 ec. for each gram of tissue, and the extraction repeated for five or ten minutes, when the solution is decanted through the same filter. The use of 4 or 5 successive portions of hot water in this manner insures complete extraction of the amino-acids in the tissues. The extracts are transferred to a 1-liter, double-necked distilling flask, and concen- trated under diminished pressure to about 20 ce. The distil'ation may be run as rapidly as is possible without loss of solution by foaming. The concentrated solution is transferred with a minimum amount of water to an Erlenmeyer flask, and mixed with 9 to 10 volumes of 95 per cent ethyl! alcohol, or half that amount of absolute 187 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2 . 188 Determination of Amino Nitrogen in Tissues ethyl or methyl alcohol. The alcohol precipitates a small amount of protein which is not coagulable by heat. The solution is allowed to stand over night to complete precipitation. It is then filtered through a folded filter, washed with 80 per cent alcohol, and re- turned to the distillmg flask. A few drops of phenolphthalein solution are added, and enough 25 per cent sodium hydrate solu- tion, drop by drop,.to render the solution alkaline. It is then concentrated under diminished pressure to drive off both ammonia and alcohol. If the receiving flask is provided with a guard flask, as illustrated in this Journal, x, p. 21, 1911, containing % acid, the ammonia can be determined. In the tissues of the dog it amounts to about 10 mg. per 100 grams of fresh tissue. When the volume of the concentrating solution has been reduced to 10 or 20 ec. the distillation is interrupted, and enough 50 per cent acetic acid is added to acidify the solution. About 50 cc. of water are also added, in order to insure that the last traces of alcohol shall be driven off, and the concentration is continued. In case the amount of tissue analyzed exceeds 10 grams, the distillation may be continued in the same 1 liter flask, but if the sample is smaller the solution should now be transferred to a flask of 300 to 500 ce. The extracts are concentrated to a few cc., and are then transferred, with several small portions of water, to a measuring flask. The size of the latter depends on the weight of tissues taken for analysis. ‘ When the amount is under 10 grams we use a 10 ce. flask to hold the final solution; when over 10 grams have been taken, a 25 cc. flask is used. The solution is now ready for the amino determination. One can take a 10 ce. portion and use the larger amino apparatus,! but we have found it in general somewhat more convenient to take 2 cc. portions and employ the micro-apparatus.* The length of time which the reacting solution should be shaken in order to drive off all the amino-acid nitrogen depends somewhat on the tempera- ture. When the latter is 15-20° the time should be five to four minutes; for 20-25° it is three minutes; for 25-80°, two and a half to two minutes. It is preferable that the solution should be shaken vigorously with a motor and the time kept down to these limits, for the sake not only of rapidity, but of accuracy. The ! This Journal, xii, p. 275, 1912. * [hid., xvi, p. 121, 1913. Donald D. Van Slyke 189 reason for this is, that, even after removal of the ammonia, the extracts contain small amounts of urea and other substances which belong to the class of slowly reacting amines, and are therefore not a-amino-acids. The amount of nitrogen which these amines give off in the time required for amino-acids to react completely is small, but if the reaction were allowed to run for an indefinite length of time instead of being kept to a definite minimum the error might be both large and variable. The correction for the amount of nitrogen given off by these amines while the amino-acids are being decomposed is ascertained in the same manner as in blood analysis, * by continuing the reaction, after the gas from the a-amino-acids has all been driven off, for a length of time equal to that utilized in decomposing the amino-acids (two to five minutes according to the temperature), and then measuring separately the gas evolved during this second reaction period. The correction found is fairly constant at about 6 per cent of the total amino nitrogen obtained. II. Simpler method for accurate comparable results. As the correction for amines other than a-amino-acids is small and fairly constant, it can be left out without decreasing appreciably the constancy of results, or affecting the determination of differences in amino-acid content. The effect of the ammonia present is also small and practically constant. Consequently when, as in most physiological work, differences rather than strictly absolute results are desired, one can simplify the above outlined method by leav- ing out the determination of the correction and the removal of the ammonia. The results agree as well as when the ammonia is removed and other amines are corrected for, but are about 10 per cent higher. One must, under any conditions, however, accurately control the time of the reaction in the amino apparatus. In our own experimental work we have always removed the ammonia. After determining the correction for amines other than amino-acids in experiments with about 20 dogs, however, without finding that it varied appreciably under any conditions or added to the signifi- cance of the results, we ceased to utilize it. The accuracy of the determination is limited, not so much by sources of error in the method, as by the fact that one cannot obtain absolutely homogeneous samples of tissue. Duplicate amino 3 Van Slyke and Meyer: This Journal, xii, p. 402, 1912. 190 Determination of Amino Nitrogen in Tissues determinations on the same solution of tissue extract usually agree as closely as one can read the volume of nitrogen gas in the burette. The error in the final amino determination is practically negligible. We have also convinced ourselves that there is very little error con- nected with the processes of extracting the tissues and concentrat- ing the extracts in the manner outlined above. If one analyzes tissue which has been dried and pulverized, so that different samples of material have the same composition, results with dif- ferent portions do not vary more than 1 or at most 2 mg. of amino nitrogen per 100 grams of fresh tissue, the amino figure for the latter usually falling between 40 and 80 mg. per 100 grams. When, however, different portions of an organ, such as the lobes of the liver, are taken as duplicate samples or when symmetrically placed organs, such as the right and left gracilis muscles or the two kidneys, are used, one must expect variations up to 10 per . .cent of the amount of amino nitrogen determined. The two concentrations under diminished pressure involved in the analysis cannot be replaced by concentrations on the water bath. During the latter the extracts darken, and part of the amino nitrogen disappears. The effect on results is especially marked with liver extracts. If the latter are concentrated to dryness and then taken up in water a black solution results, which may show less than half the amino nitrogen originally in the extract. The chief problems of the analysis were the complete extraction of the amino-acids from the tissues, and the removal of the proteins from the extract, both operations being necessarily performed under conditions which result in neither loss of amino-acids nor — their formation by hydrolysis of the protein present. Folin and Denis in their methods for tissue analyses met these conditions by extracting the tissues with cold methyl alcohol. ‘This method was not suitable for our purposes, however, as alcohol extracts the lipoids, and they may settle out of the final water solution in such masses that they mechanically hinder an accurate amino determination. Extraction with hot water is much more rapid than with alcohol, and the extract contains a relatively small amount of'lipoids. It contains an appreciable amount of protein, however. For its removal we found the above described alcohol treatment the most satisfactory general method, although it makes two vacuum concentrations necessary, when one would be suffi- Donald D. Van Slyke IgI cient if the proteins could be satisfactorily thrown out of the water extract by one of the usual precipitants. We tried a number of these, and found that metaphosphoric acid in particular gave results with muscle extracts which were satisfactory and agreed with those by the alcohol method, but that in liver extracts, presumably because of the glycogen present, it failed to precipitate the protein satisfactorily. The only apparent objection to the hot water extraction is, that it might increase the amino nitrogen by hydrolysis of some of the proteins present. That such hydrolysis does not occur appears from the following experiment. Muscles fresh from the thigh of a dog were cut into pieces with scissors and dried over sulphuric acid at a pressure of 0.2mm. The dried muscles were pulverized and thoroughly mixed. Samples of 3 grams each were weighed out, placed in flasks with 100 cc. each of hot water, and allowed to digest at 100° for periods of five minutes, thirty minutes, and one, two# and three hours respectively. At the end of the period of digestion the water was decanted through a glass wool filter, and the residue in the flask washed with three 75 cc.-portions of water at 100°, each portion being allowed to remain five minutes on the muscle shreds. In case the action of hot water on the tissues produces a sufficiently rapid hydrolysis to become a factor in disturbing the accuracy of results it should make itself apparent by increased amino nitrogen content in the samples which were digested long- est. That this was not the case is shown in the table below. The water extracts were concentrated and precipitated with alcohol, and, after the alcohol had been driven off, the entire solution was used for determination of amino nitrogen in the larger apparatus. TABLE I. ° PERIOD OF DIGESTION WITH HOT WATER Cc. NITROGEN Gas AT 20°, i RRR Rs: a «Se 11.4 OE ESS CSRS pt SP a: oa Cs 11.4 Bis 0 TO ta rea Uae a, oe 11.5 OS eR tits. a.c ergo. J CME wo cgMIIG 5 6 3 ake 11.2 ANG GL Ey are) SS ee) a ) Sie The results also indicate the accuracy of the methods, from the initial extraction to the final determination, when the samples are taken from homogeneous material. The results of the following experiment exemplify some points of interest. 192 Determination of Amino Nitrogen in Tissues Fresh muscle of 123 grams’ weight was extracted with hot water and the water extract brought to 100 cc. The 100 cc. were divided into five portions of 20 cc. each. The non-coagulable protein was left in No. 1. From the others it was removed by the means indicated in the table. The solutions were freed from ammonia and eventually all brought to 25 cc., of which 2 ce. portions were used for duplicate determinations of amino nitrogen. Ten cc. portions were, furthermore, mixed with 10 ce. each of concentrated hydro- chloric acid and heated twenty-four hours to completely hydrolyze the pro- teins and intermediate products present. The hydrochloric acid was driven off as thoroughly as possible by concentrating in vacuum, and the ammonia boiled off in vacuum with calcium hydrate. The residual solutions were diluted to their original volume of 10 ee., and 2 cc. portions taken for amino determinations. The methyl and ethyl alcohols used were ‘‘absolute;”’ the zinc chloride solution contained 5 grams of the chloride dissolved in 100 cc. of 80 per cent ethyl alcohol. TABLE II. ~ AMINO N per 100 GRAMS No METHOD OF PRECIPITATING PROTEINS USCLS PEPTIDE : in 20 cc. WATER EXTRACT ———————— in . res BOUND N | er — I Free | sis of coment 1 | Not precipitated... .......... | 89 224 135 2 | 100 ce. methyl alcohol........ 68 | 100 32 3 | 100 ee. methyl alcohol+2 cc. ZnCl; solution............. 55 89 33 4 | 100 ee. ethyl alcohol.......... | 65 98 33 5 | 100 ce. ethyl aleohol+2 ce. ZnCl, solution............. 58 85 27 Alcohol alone added to the water extracts precipitates chiefly proteins and higher intermediate products. The free amino nitro- gen precipitated by methyl alcohol was 89—68=21 mg. The pep- tide bound nitrogen was 1385—32=103 mg. The ratio (free NH. : peptide bound NH) in the precipitate was, therefore, approxi- mately 1:5. The ratio in animal proteins is usually about 1; 12 (the animal proteins containing about 5 per cent of their nitrogen as free amino nitrogen, about 60 per cent more being set free by hydrolysis). The ratio 1:5 shows that the aleohol precipitate consisted, in part at least, of intermediate products, but that it could have contained little or no free amino-acid nitrogen. Similar ‘ Conditions for Complete Hydrolysis of Proteins, This Journal, xii, p. 205, 1912. ey aR Donald D. Van Slyke 193 results are found with ethyl alcohol as a precipitant. It removes proteins and intermediate products, but leaves the amino-acids. The action of zine chloride, which when added to the alcoholic mixture precipitates some substances that alcohol alone does not, is different. It apparently precipitates some free amino-acids, without much affecting the intermediate products. The decrease in free amino nitrogen caused by adding the action of zinc chloride to that of alcohol (compare No. 2 with No. 3 and No. 4 with No. 5) is greater than that of the peptide bound nitrogen. One can also obtain a precipitate with zinc chloride and alcohol in a slightly acid solution of the products of complete acid hydrolysis of casein, and the precipitate removes part of the amino nitrogen from the solution. It would seem, therefore, that zine chloride in alcoholic solution precipitates some amino-acid or acids. We have not fur- ther investigated the point, as it is of minor interest, but because - of the above results and others like them we have not utilized zine TABLE III. Bip So s tte ba he N ren 100 a Be ~ ISa- z as GM. TISSUE 3 |e o seees 3 al SAMPLE . es ve las = | 5 8 8 3 © | Ga) de teeta 2 | 2 | SE | Ba | p | SMEe Cage: 2 | bee | te] § Pa) si “eeege &§ | € | as | ba | : B |g] ae cm | Bae emeee | ) mg. | deg mm min Left triceps muscle..... 14.98} 5.1); 10 6.75 4 | 0.40 | 4 ; 6.35 21 758 | 60 64 Right triceps ) | | muscle... 20.37, 7.9| 11 | 8.70 | | 0.60 | at / 840 | 23 | 760 (+56 | 60 Liver, lobe 1.) 11.81 3.9, 9 | 6.55 | 4 | | 0.40 4 | | | 615 | 22 | 758 72 | 7% Liver, lobe 2.| 19.62) 7.2) 10 | 10.60 | 4 | | 0.50 | 4 | | 1010 | x | me; =| 3 | ie 194 Determination of Amino Nitrogen in Tissues chloride as a precipitant. It would not be likely to affect the signifi- cance of comparative results, however, like those of Folin and Denis, because the proportion of amino nitrogen precipitated is quite constant. The results in table III, taken from data in connection with Dog 17, serve as examples of the order of magnitude of the fig- ures obtained, the volumes of gas measured, etc. The determina- tions were performed according to the ‘“‘absolute’’ method, the am- monia being removed, and the correction for the amines, other than a-amino-acids, being determined. The following table, from analyses of another dog, gives some typical figures. TABLE IV. Bs , | FREE AMINO N IN AMINO N IN EXTRACT AFTER PRECIPITATION OF | WATER EXTRACT, PROTEINS WITH ALCOHOL | BEFORE REMOVAL TISUE | OF PROTEINS NOT NH: after com- | Peptide bound N | COAGULATED BY Free NH: plete hydrolysis (NH freed by ' HEAT with HCl hydrolysis) Gracilis muscle 50 38 61 23 Heart: . . ..sc3 50 37 61 : 24 Saver......cnn 75 ' 61 89 28 Spleen........ Ot | «+53 68 15 Kidney....... 73 65 77 12 Stomach...... 41 32 53 21 Duodenum 66 50 60 10 The results in the second and third columns show that the amino nitrogen found in the extract after treatment with alcohol (second column) comes chiefly from free amino-acids. In the case of even the simplest peptides, the dipeptides, the free amino nitro- gen is doubled by hydrolysis, and in the primary albumoses it is increased about eight times. Here the increase is only 20 to 50 per cent. A mixture of amino-acids and albumoses such that 93 per cent of the free amino nitrogen belongs to the amino-acids and only 7 per cent to the albumoses, would give an increase of 50 per cent in the amino nitrogen on hydrolysis of the albumoses. Such a relation apparently exists in the tissue extracts as prepared for analysis; for even after the alcohol treatment they show the pres- ence of traces of protein or higher intermediate products when Donald D. Van Slyke 195 tested for biuret with the precautions given by van Norman.° The relations between the free amino nitrogen and the peptide bound nitrogen given in the above table are typical. A much larger proportion of peptide bound nitrogen is never found. One can, therefore, depend upon figures for amino nitrogen in the tissues obtained by the above method as representing with a fairly close degree of approximation the simple amino-acids. SUMMARY. The amino-acids are extracted from the tissues with hot water. Uncoagulated proteins in the extract are precipitated by alcohol. Alcohol and the slight amount of ammonia present in the extract are removed by concentration in vacuum, and the amino nitrogen in the residue is determined by the nitrous acid method. The rapid- ity with which the amino nitrogen reacts with nitrous acid, and the relatively small increase which it shows as the result of hydro- lysis of the extract with hydrochloric acid, indicate ‘that the amino nitrogen determined by the method outlined represents approxi- mately the free a-amino-acids. Only a few per cent of the amino nitrogen appears due to proteins or their intermediate products, and to amines not of protein origin. The correction for the latter can, when desirable, be readily determined. . 5 Biochem. Journ., iv, p. 127, 1909. THE FATE OF PROTEIN DIGESTION PRODUCTS IN THE BODY. III. THE ABSORPTION OF AMINO-ACIDS FROM THE BLOOD BY THE TISSUES.' By DONALD D. VAN SLYKE anp GUSTAVE M. MEYER. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, August 27, 1913.) In our preliminary communication? we mentioned only the most recent articles on the fate of the products of protein digestion. In order that our work may appear in its proper relation to that which has previously been done in the field, it appears desirable to re- view the latter more fully before proceeding with the report of our own results. The contributions on the subject appear to be most readily grouped around the different explanations of the fate of ingested protein which have served as working hypotheses. These hypoth- eses may be formulated as follows: 1. The ingested proteins are absorbed and incorporated into the body without undergoing any marked chemical change. 2. The food proteins are first hydrolyzed in the alimentary tract; the.products of digestive hydrolysis are then absorbed into the blood and carried to the tissues. 3. The products are deaminized in the wall of the intestine before entering the circulation. 4. The products are synthesized into serum protein before en- tering the circulation. The serum proteins thus formed serve as nourishment for the tissues in general. We will take up these hypotheses in their order, reviewing the work which appears most important in supporting or antagoniz- ing each. 1The results in this paper were reported in abstract at the meeting of the Soe. Exper. Biol. and Med., Dec. 18, 1912. Proceedings, x, p. 38. 2 This Journal, xii, p. 399, 1912. 197 198 Absorption of Amino-Acids by Tissues 1. The simplest, and ayparently the earliest theory concerning the manner in which the proteins of the food reach the tissues and are incorporated into them, is that the proteins are absorbed, with little or no chemical change, directly into the circulation, from which they are taken by the tissues and incorporated into their substance. In support of this view the fact was demon- strated that unchanged proteins could be made to pass directly from the alimentary canal into the blood.* Egg albumin can even be absorbed in such amounts that it appears in the urine. More thorough investigation has shown decisively, however, that the absorption of unchanged proteins is an abnormal process. From a knowledge of the proteolytic powers of the digestive tract attained by the famous experiments of Spallanzani and Beaumont with the gastric juice, the discovery of trypsin by Kihne,‘ and the work of many later investigators, including Cohnheim,' the dis- coverer of erepsin, one is justified in stating that the more deeply the processes in the alimentary tract have been studied the more thor- ough has the breakdown of proteins in normal digestion been found. Cohnheim’s work on this point is particularly important. He fed meat to a dog with a duodenal fistula which permitted isolation of the products of digestion after they had passed the stomach and part of theintestine. The partially digested products obtained from the fistula were treated with erepsin for twenty-four hours, the ac- tion of this enzyme thus following, as in normal digestion, that of the gastric and pancreatic juices. The meat proteins were hydrolyzed so completely that all the arginine was free’ and the nitrogen — precipitable by phosphotungstic acid could not be farther de- — creased by boiling with sulphuric acid.’ No evidence of the presence of peptides or intermediate products could be found. It appears probable, therefore, that normal digestion proceeds in the intes- tinal lumen and wall until most, if not all, of the proteolytic produets are reduced to the stage of free amino-acids. Abder- -halden and his co-workers have isolated nearly all of the known * For references see article by Cohnheim: Zeitschr. f. physiol. Chem., xxxv, p. 397. * Virchow's Archiv, xxxix, p. 155. * Zeitschr. J. physiol. Chem., xxxiii, p. 451. * Thid., li, p. 415, 1907. ’ Thid., xlix, p. 64. D. D. Van Slyke and G. M. Meyer 199 amino-acids from intestinal contents, and have furthermore shown that protein which has been digested completely into amino-acids is as efficient as intact protein in maintaining the nitrogenous equilibrium and even the growth of dogs. Abderhalden has also found positive evidence that intact protein is not normally ab- sorbed into the circulation.® Injection of protein into the circu- lation results in the development of a proteolytic enzyme in the blood capable of hydrolyzing the injected protein. The fact that the normal blood is free from such enzymes shows that it does not absorb undigested proteins. Furthermore, the fact that it is djffi- cult or impossible to develop an anaphylactic state by protein feeding is proof against absorption of proteins as such, even in small amounts. Also, attempts to find evidence of food proteins in the blood by the precipitin reaction have given negative results.!° 2. The absorption of intact proteins being an untenable hypoth- esis, the simplest alternative explanation is that the “peptone,” or the mixture of digestive products, is absorbed directly into the blood and conveyed to the tissues. That this mode is not impossible is indicated by recent work of Buglia.’* He found that completely digested flesh in amounts equivalent to a day’s protein require- ment could be injected intravenously into dogs without injurious effect if several hours were taken for the injection, so that the rate of entrance of the products was similar to the rate of absorption in normal digestion. The injected products were mostly metabo- lized and excreted as urea. That during the actual protein digestion, however, the final hydrolytic products enter directly into the circulation without undergoing chemical change while passing the intestinal wall, re- mained uncertain because of the many failures to demonstrate these products in the normal blood. For several decades it has been inves- tigated for the presence of peptone with negative results." When 8 Synthese der Zellbausteine. 9 Schutzfermente. 1° Debré and Porak: Journ. phys. et path. gén., xiv, p. 1019. 4A clear presentation of this view was given in 1905 by Folin: A Theory of Protein Metabolism, Amer. Journ. of Physiol., xiii, p. 117. A thorough discussion of work up to 1912 on this subject is given by Cath- cart: Physiology of Protein Metabolism, chapter on protein regeneration. 2 Zeitschr. f. Biol., |viii, p. 162, 1912. 18 Abderhalden Rnd Oppenheimer: Zeitschr. P| physiol. Chem., xlii, p. 155, 1904; Howell: Amer. Journ. of Physiol., xvii, p. 273, 1906. 200 Absorption of Amino-Acids by Tissues the importance of amino-acids as the end products of digestion became appreciated these also were sought, but the most careful work failed to result in the isolation of a single amino-acid from normal blood, even during the height of digestion." Abderhalden, Gigon, and London were able, it is true, after in- jecting alanine into the stomach of a dog, to isolate that amino- acid from the blood and urine as the naphthylsulpho compound. Results likewise indicating the ability of protein digestion products to pass from the intestine into the circulation were obtained by Cathcart and Leathes.*® They found an increase in the “residual nitrogen’”’ (nitrogen left after subtraction of the urea and removal of the protein precipitable by heat and by tannic acid) following the absorption of peptone injected into a loop of the intestine of a dog. An increase in the non-protein nitrogen of the liver was also noted. Results on the blood similar to those of Cathcart and Leathes have recently been obtained with a refined technique by Folin and Denis"’ after injection of amino-acids into the intestines of cats. These authors have also determined an increase in the non-protein nitrogen of the muscles following the absorption of amino-acids from the intestine. These results, however, can hardly claim the same degree of finality as those of Folin and Denis on the question of deaminization and urea formation to be shortly dis- cussed. The interpretation of changes in non-protein nitrogen as changes in amino-acid nitrogen is arbitrary, as there was no evi- dence concerning the chemical nature of the nitrogen in which the changes were noted, aside from the proof that it was not in the form of urea or ammonia. The fact that amino-acids can enter the blood from the alimen- tary canal when the latter has been flooded with them, although clearly demonstrated by Abderhalden, Gigon, and London, did not carry with it the proof that absorption of unchanged amino-acids into the circulation oecurs during their gradual liberation in diges- tion, nor was it regarded as such proof by these authors. Egg albu- min can also be made to enter the blood and even appear in the urine when the alimentary canal is flooded with it under properly “ Abderhalden: Synthese der Zellbausteine. Zeitachr. f. physiol. Chem., liii, p. 118, 1907. 6 Journ. of Physiol., xxxiii, p. 468, 1906. " This Journal, xi, p. 87, 1912. D. D. Van Slyke and G. M. Meyer 201 chosen conditions. Proof of the presence of amino-acids in the blood under normal conditions none of the investigators in the field was able to accomplish, despite the application of all the meth- ods known to modern chemistry for the isolation of these sub- ‘stances. Investigation of the residual nitrogen of the blood in- dicated the possibility of the absorption of amino-acids or peptides. The failure to ascertain the chemical nature of the fraction of resi- dual nitrogen to which the changes noted were due, however, made the results inconclusive, as those based on residual nitrogen determinations noted in the preceding paragraph would also have been, had they not been confirmed by the more definite findings of Abderhalden, Gigon, and London. Hohlweg and Meyer'* found in the serum of fasting dogs an average residual nitrogen of 7 mg. per 100 cc. In digesting animals the average figure was 13 mg. As the amounts were so small, however, the individual fluctu- ations considerable, and in particular as there was no evidence of the chemical nature of the residual nitrogen, the results could not be regarded as decisive. Howell attacked the problem’ with the aid of naphthylsulfochloride, which had been introduced by Fischer and Bergell as a precipitant for amino-acids. He obtained a precipi- tate in the dialysate of the serum, and the bulk precipitated was observed to be larger in the serum of fed dogs than in that of fasting animals. The precipitate was an oil, however, which could neither be identified chemically nor measured quantitatively. For this reason, notwithstanding the interesting possibilities indicated by the results, it could not be stated with certainty what proportion, if any, of the precipitate was due to amino-acids. Cohnheim, work- ing with the alimentary canal of the octopus under conditions which to some extent simulated the natural, succeeded in separating crystalline amino-acids from the blood.?® The tract under normal conditions practically floats in the blood of the animal. Cohn- heim removed the tract with the digestive glands attached, filled it with peptone solttion, and floated it for twenty hours in blood through which oxygen was passed. The organs rémained alive during the experiment. At the end of the latter the residual nitrogen of the blood, which is ordinarily almost nil, was found 18 Hofmeister’s Beitrdge, xi, p. 381, 1908. 19 Amer. Journ. of Physiol., xvii, p. 273, 1906. 20 Zeitschr. f. physiol. Chem., xxxv, p. 396, 1902. 202 Absorption of Amino-Acids by Tissues to be greatly increased. No peptone could be detected in it, but ammonia was determined, and leucine, tyrosine, and lysine picrate were crystallized from it, though not enough of any was obtained for analysis. Cohnheim was conservative about generalizing these _ results to apply to normal digestion in the higher animals, and his reserve appeared to be justified by later results which he obtained | in repeating the above experiment witha vertebrate fish, Crenilabrus pavo. He obtained ammonia in the external blood, but no evidence of either mono- or diamino-acids, and concludes: ‘‘dass bereits beim Passieren der Darmwand die Eiweissspaltprodukte teilweise desamidiert werden und in Ammoniak und einen, zuniichst un- bekannten Rest zerfallen.’’ 3. In the view of the failure to obtain conclusive proof of the presence of amino-acids in the blood of the higher animals during digestion, the above results of Cohnheim suggested the possibility that the amino-acids are deaminized while passing the intestinal wall, the first stage of their catabolism occurring before they enter the circulation. This hypothesis derived support from earlier work of Nencki, Zaleski, Pavlov, Salaskin, and Horodynski, who found the ammonia content of the portal blood greater than that of the arterial during digestion. None of these results proved, however, that the greater part of the amino-acids suffers decom- position during absorption, and the deaminization hypothesis has. in all events been effectively retired by recent work of Folin and Denis.“ Using delicate quantitative methods which they had developed for the determination of ammonia and urea, they found that neither of these products appeared in increased amounts in the blood during absorption of glycocoll or alanine from a loop of the small intestine of the cat. They also showed that the ammonia of the portal blood is due largely to the products of putrefaction in the intestine. 4. The remaining explanation of the means by which the products of protein digestion reach the tissues without appearing in the blood is the antithesis of the recently demolished deaminization hypoth- esis. It assumes that the products in passing the intestinal wall, instead of being decomposed, are synthesized into protein again, and % Zeitschr. f. physiol. Chem., lix, p. 289, 1909. * For references, see Cohnheim: ibid., lix, p. 246. * This Journal, xi, p. 161, 1912. Beane D. D. Van Slyke and G. M. Meyer 203 that the result of the synthesis is one or more of the proteins of the serum. This explains at one stroke both the failure to find - amino-acids in the blood and the origin of its proteins. The explana- tion was founded on less positive evidence than the deaminiza- tion hypothesis, but its long life shows that it possessed the advan- tage of being difficult to disprove. The hypothesis was clearly formulated at least as early as 1870 by Funke:* “Daim Blute und im Chylus gar keine oder nur Spuren von Peptonen sich finden, so bleibt keine andere Annahme iibrig, als die, dass die Peptone unmit- telbar nach ihrer Aufsaugung, gleichviel ob dieselbe in’s Blut oder Chylus oder beide stattfinde, in gewohnliche Eiweisskérper, viel- leicht in die gleiche Modification, das Serumalbumin, zuriickver- wandelt werden.”’ Hoppe-Seyler localized the process of resynthesis even more definitely:* ‘Da nun in Magen und durch das Pankreas- sekret Acidalbumin und Pepton gebildet wird, so scheint auch das eine Funktion der Epithelzellen des Darmes zu sein, diese K6rper in Serumalbumin und fibrinbildende Stoffe iiberzufiihren.”’ Evidence which might be interpreted in favor of the resynthesis theory was brought by Hofmeister,”® who found that the stomach wall of a digesting dog contained peptone, which, however, dis- appeared rapidly when the stomach was kept for a half hour or more at 40°. Hofmeister pointed out that the disappearance could be due to either resynthesis or further digestion of the peptone. Glassner?’ decided the question in favor of resynthesis. He found -that the disappearance of non-coagulable nitrogen was due entirely to the albumoses (fraction not precipitated by heat, but precipi- tated by saturation with zine sulphate), the nitrogen in the filtrate from the albumoses, which contained the products of further diges- tion, remaining constant after removal of the stomachs from the animals. Embden and Knoop,?* however, who repeated the experi- ment, with the difference that they used intestine instead of stom- ach, found that the decrease in albumoses was accompanied by an increase in their filtrate, and was therefore to be attributed * Lehrbuch der Physiologie. Quoted by Popoff: Zeitsehr. f. Biol., xxv, p. 427, 1889. Popoff believed that the synthesis occurred in the lumen of the tract, before absorption. 26 Pfliiger’s Archiv, vii, p. 399. 6 Zeitschr. f. physiol. Chem., vi, p. 69, 1882. 27 Hofmeister’s Beitrdge, i, p. 329, 1902. 28 Ibid., iii, p. 120, 1903. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2. 204. Absorption of Amino-Acids by Tissues to further digestion of the albumoses rather than to their resyn- thesis into protein. This had been advanced by Cohnheim?® as the probable explanation of Hofmeister’s results, after Cohnheim had discovered the activity of the erepsin of the intestinal wall. Aside from the results of Glissner, which are opposite to those later obtained by Embden and Knoop, it appears that no positive evidence has ever been found for the hypothesis that the products of protein digestion are resynthesized in the walls of the alimentary canal into blood protein. The real evidence, the failure to identify the products of protein digestion in the blood, has been purely negative. This evidence, we believe we are justified in stating, has, even if one cannot admit | the conclusiveness of -the significant work of Howell, been decis- ively eliminated by the results published by us a year ago.*® Using the nitrous acid method for determination of amino groups under precautions which render it specific for a-amino-acids, we found that the latter are always present in the blood of dogs, the amount of amino-acid nitrogen being 3 to 5 mg. per 100 ce. of blood in animals after twenty-four hours fasting. After a meal of meat the figure rose to 10-11 mg. in the same animals. The results not only dispelled the negative evidence on which, because of lack of suffi- ciently sensitive methods, the resynthesis hypothesis had been built, but afforded positive proof that the products of protein digestion enter directly into the circulation. The amount of amino- acid nitrogen present at any one time in the blood is small, because - amino-acids which enter it leave it with great rapidity. We found that intravenously injected alanine disappeared from the circula- tion almost as fast as it entered. A similar disappearance of in- jected amino-acids had shortly before been already noted by Woelfel.*! Immediately after our paper one on the same subject by Abder- halden and Lampé appeared.” In their work the amino-acid nitro- gen of the blood was detected by the ninhydrin color reaction, the intensity of the color developed affording comparative results. 2% Zeitechr. f. physiol, Chem., xxxiii, p. 451, 1901. © This Journal, xii, p. 309, 1912. * Proc, Amer. Physiol. Soc., Abstracts, 1911, p, 38; published in Amer. Journ. of Physiol., xxix, p. xxxviii, 1912. * Zeitachr. f. physiol. Chem., \xxxi, p. 473, 1912. D. D. Van Slyke and G. M. Meyer 205 The results confirmed ours, but the authors still favor the resyn- thesis hypothesis, believing that the main portion of the digestion products is resynthesized into blood protein. during absorption. The passage of amino-acids unchanged into the circulation during digestion they explain on the basis of the difference in composi- tion between the proteins of the food and the blood respectively. The food proteins contain certain amino-acids in greater propor- tions than the blood proteins, and some of these amino-acids will necessarily be left over when the maximum amount of serum pro- tein has been synthesized from the food. It is, according to Abder- halden and Lampé, only these superfluous amino-acids that pass ‘unchanged into the circulation. The part of the absorbed products important for nutrition is that which enters the circulation as serum protein, and it is the serum protein, according to Abder- halden, that nourishes the tissues in general. These take up the protein from the serum, hydrolyze it again into amino-acids, and from the latter reconstruct their own proteins. In regard to the ascertained facts (absorption of amino-acids directly into the cir- culation during digestion) there is no disagreement between Abder- halden and ourselves. The above hypothesis, however, notwith- standing the valuable work which it has stimulated, is not, it seems to us, the most probable explanation of the facts thus far at our disposal. It assumes a number of processes (synthesis of absorbed amino-acids in the intestinal wall to serum protein, utili- zation of serum protein as pabulum by the body cells) as yet quite undemonstrated by established facts. Moreover, the demonstrated mechanism, by which the amino-acids liberated during digestion are absorbed directly into the circulation and transferred to the tissues, is sufficient to handle these products as rapidly as they are formed; and we know at present of no ground for assuming addi- tional and more complicated processes to provide the tissues with protein constituents. In brief, it has been found that an undeter- mined proportion, possibly all, of the amino-acids formed in diges- tion passes unchanged into the circulation; and it has not yet been shown that any of them, except such as may be altered by bacteria, are either conjugated or destroyed before entering the blood stream. Recently a paper appeared by Rona* which, furthermore, offered 338 Synthese der Zellbausteine. *4 Biochem. Zeitschr.. xlvi, p. 307, 1912. 206 Absorption of Amino-Acids by Tissues evidence that no large fraction of the amino-acids suffers chemical change while passing the intestinal wall. Surviving intestines of cats were suspended in Tyrode’s salt solution and filled with solu- tions of amino-acids or of digested peptone, the amino nitrogen content of these solutions having been determined by the gaso- metric method. After several hours, during which the intestines maintained their vitality and motility, from one-half to one-third of the amino nitrogen had diffused through the intestinal walls into the Tyrode’s solution. There was no decrease in the total . amount of amino nitrogen present, such as would have occurred if the passage of the intestinal wall had been accompanied by a syn- thesis of protein. Although Rona himself did not claim that his® results were conclusive, inasmuch as there was no circulation of blood through the intestines and conditions were therefore not entirely comparable to those in the living animal, absolutely neg- ative results would hardly have been expected if the intestines normally possess the ability to synthesize protein at the rate nec- essary to keep pace with absorption. From the above review it appears that the positive results of pre- vious work on the problem before us can be condensed into the following statement: Ingested proteins are hydrolyzed in the di- gestive tract setting free most, if not all, of their amino-acids. These are absorbed into the blood stream, from which they rapidly disappear as the blood circulates through the tissues. In the present paper we attempt to answer the question: What becomes of the amino-acids when they vanish from the circulation? Are they decomposed in the blood: are they at once synthesized into new protein; are they chemically incorporated into the com- plex molecules of the tissue proteins; or are they merely absorbed by the tissues in general, or by certain tissues in particular, without undergoing any immediate change? EXPERIMENTAL. Experiment 1. A male dog of 9 kg. weight, which had fasted four days, was etherized and kept with artificial respiration by the Meltzer and Auer insufflation method during the entire experi- ment. The bladder was washed out through a catheter, and a sample of 25 ee. of blood was drawn from the right femoral artery. The right gracilis muscle, a lobe of the liver, a short section of the D. D. Van Slyke and G. M. Meyer 207 small intestine, and the right kidney were removed and coagulated for determination of amino nitrogen. The lobe of liver was isolated by means of a large clamp at the base before excision; the other samples were dissected and tied off, so that all were taken practi- cally without loss of blood. One hundred and fifty, ce. of a solution of the amino-acids obtained by hydrolysis of casein were then injected into the right femoral artery. The solution was made by boiling casein forty-eight hours with seven times its weight of 33 per cent sulphuric acid. The sulphuric acid was removed with excess barium hydrate, and the ammonia removed by concentrating the alkaline solution in vacuum. The barium was then removed with sulphuric acid, the reaction being so balanced that the filtrate from the barium sulphate gave a barely perceptible reaction for sulphate. This condition assured complete removal of the barium. The solution was concentrated in vacuum, and the slightly acid reac- tion was changed by adding sodium carbonate until the solution gave a barely perceptible alkaline reaction with litmus. The final solution contained 27.08 mg. of amino nitrogen per ce. It was used instead of the solution of a single amino-acid because it un- doubtedly resembles more nearly the mixture of amino-acids ab- sorbed from the intestine during digestion. The 150 cc. injected contained 4.06 grams of amino nitrogen. The duration of the injection was thirty minutes. Half an hour after it had been finished another sample of 25 ec. of blood was drawn, the dog was killed by Bietng, and samples of the tissues again taken. During the period following the injection 125 ce. of urine were voided through the catheter or expressed from the bladder at the end of the experiment. The results of the analyses, which were made by the ‘‘absolute’’ method described in the preceding paper, are given in table I. The blood analyses were made as described in our first paper.® The urine excreted during the experiment contained 0.738 gram of nitrogen, of which 0.463 gram, or 11 per cent of the amount injected, was amino-acid nitrogen. Calculating the blood as 5 per cent of the weight of the animal, the increase of 41.5 mg. in the % This Journal, xii, p. 402, 1912. 8° For method of determining amino nitrogen in urine, see Van Slyke: ibid., xvi, p. 125, 1913. 208 Absorption of Amino-Acids by Tissues TABLE I. a MG. AMINO NITROGEN PER 100 GRAMS TISSUE TISSUE SAMPLE Before injection Bs. oO after ME Bie ics o's 9 Ll eee SESS, 34 Suveral.:. i o.ki ww eee 29 TEWOR G8: Se once. v's. és a. : 94 eee ee Se 93 Grasilis muscle 1:...........). 22 43 Gracilis muscle 2...........) 0g 67 Gracilis muscle 3..........+...eueeee 73 Midney 1.....-...:. 22.00... 45 | Madneéy 2......... <5 .sskh. dee 106 Entestine 1... ......:5:2. 6. si.) ae 48 Intestine 2.....:....¢.....5 . pe yf PANCTCAS............ 05+ tes 91 Spleen... ...... cg ieess esse seen 81 Blood'l.......2.0s..ces 255) Sa 3.9 Bleed 3. ... . . :iciecn lob 5ckte e. 45 .2 Blood 3... 0. ices eee esses ; ah 6 amino nitrogen indicates that 0.19 gram, or 5 per cent, of the in- jected amino nitrogen remained in the circulation at the end of the experiment. The intestinal juice measured, as nearly as could be estimated, 200 ce. It contained 45 mg. of amino nitrogen per 100 cc., or a total of 0.09 gram, 2 per cent of the amount injected. It is doubtful whether this small amount was due to excretion of part of the injected amino-acids into the intestine, or to traces of unabsorbed digestive products. Summarizing the results of the experiment: Of the amino nitro- gen injected, approximately 5 per cent remained in the circulation a half an hour after the injection. Eleven per cent had been ex-— creted in the urine. If the remaining 3.41 grams of amino nitrogen injected had been absorbed by the tissues evenly throughout the body, the average increase per 100 grams of tissue (taking the weight of tissues aside from the blood as 8.5 kg.) would have been 40 mg. The inereases found were: in the muscles 27 mg.; liver, 60 mg.; kidney, 60 mg.; intestine, 60 mg. Although strictly accurate cal- culations are, of course, impossible, the results indicate, as closely as one can estimate from such figures, that all the amino-acids which disappeared from the circulation were absorbed, without suffering immediate chemical change, by the tissues. D. D. Van Slyke and G. M. Meyer 209 It will be noted that the amino nitrogen content of the muscles did not rise so high as that of the internal organs. The figures exemplify a fact that we have noted in all our experiments, viz., that the amount of amino-acid nitrogen that the muscles can hold is limited with relative sharpness. By injection of the amino- acids from proteins hydrolyzed by either acids or enzymes we have never been able to force the amino nitrogen figure of the striped muscle above 80 mg. per 100 grams. If the figure is above 70 at the time of injection, little or no amino nitrogen is taken up. In the liver we have noted as high as 160 mg., and the other internal organs seem to possess a more elastic ability to absorb amino-acids than do the muscles. Experiment 2. The conditions were similar to those of the first experiment. The dog used for Experiment 2 weighed 7.4 kg., and the amount of amino nitrogen injected was 3.39 grams. In this case the second set of tissue samples was taken one hour after the in- jection had been finished. TABLE II. MG. AMINO NITROGEN PER 100 GRAMS TISSUE TISSUE SAMPLE oP Before injection One Diver fie cs: ... aes a 48 léver $0). te... ae. 44 Liver:3 igus: . sae < . . EES ss . 127 Liver 4. Sages». seeueee:« -- MEE, 3 124 Right gracuee.l. Se aa yea] jo yyZue7T] A, 9 Zz L 9 ¢ ¢ Fee) ae $001 jo joquin yy v a Gg | SOL} SPE] SFT) O08 | 06 |/ PIP od Sg eee eee yOOI JO Bua] |} ~~~ “azwuoqivo umIpog w i O8 | 8 16 | AIT AM £8 hodik { ee a ee ee jyeoy] jo yysuay ; ‘ “a 9 ea pa ¢ ¢ 9 C1) ae $}001 Jo JoqunN : = GZ Of | O11; OL | O81} OST) ¢9 0g Pomp {P" Oueenaas ood jo yyBuay > apuoTys uMIpog — = 08 68 G8 S83 G6 16 L9 GP rooeeseesssss yea JO 49Zuey 2 9 | 6 L 6 L L 9 CUE Ge Seen © eee $}001 JO JequUINN ; O ZI 001; OOT| 06 | $8 | 06 | OF og porp.¢ |r ee aap yoor jo yyBueyT |p ayeydins umrpog o 08 O03) 26 1) 00 aee 1 ee po [ym yea] Jo YBueT ° 5 9 LZ 6 8 | 9 9 9 Be eee Hs gimapos jo Jaquinn * Gal OST} OOT| OST | OOT| 46 | OF G Pep 4)" **g001 JO YSU] |> ©“ OpuoyYyo wiNID[eD a 08 cor| 06 | 06 | 08 | €2 | LF Co Pt a yea] JO YBue'y] WY) 9 | 9 8 9 : ee ee fh eee s}oo1 jo zoqumyy |) —-_ GZI | SIL| O11| S9T| 06 | OL | SF oy ih Pomp ¢)" "aes yoor Jo y3Zuery |} * “epuo0pyo umisoutepy v 08 | 88 | OOT| 6 | €8 | O08 | 99 1Z rere ssessess por JO UQBUeT |) 6 + 9 | G 9 6 ¢ g I | I Tee oa $}004 Jo aquiny a SBR OOK |: SE Ort] $8 | 08 | 99 | gO SE ee os jOO1 Jo Y}BuayT |} © -e7eYdins umisouseyy a 08 | 06 | 46 | GIT] &8 | 08 | 09 | OF 00 ere? +J¥2] JO WySue'7y (on ee ee —— —— - a = aos oo NOWLNOO oooo1 000% 001 aot OOL 9% as & - ples ! | Ss . agsa sxive 1) SNOLLVULNGONOO K. Miyake 239 five seedlings were grown in each culture at ordinary tempergture and the evaporated water was supplemented with distilled water from time to time to keep the solutions at their original concen- trations. After ten days, the difference of development was very striking, and then the determinations recorded in the table on p. 238 were made: The results show that each salt acted as a toxic agent or a stimu- lant upon the growth of rice seedlings, according to its concentra- tions. Magnesium sulphate and chloride, calcium chloride and sodium carbonate were injurious when the concentrations were greater than about ;%5, while sodium sulphate, chloride and bi- carbonate were toxic when the concentrations were greater than 7¥;. In every salt, when the concentration was such that the toxic action ceased, the stimulating effect began and attained its highest degree in the following order of concentration: magnesium sulphate ;%,5, magnesium chloride ,;45, calcium chloride 575, sodium sulphate 5, sodium chloride ;%,, sodium carbonate s“, to +245 and sodium bicarbonate 345. i Experiment II. On June 13, 1912, twenty-five rice seeds of almost uniform size and specific gravity (1.158-1.185) were sown in the beakers, about 55 cm. in diameter and 7 cm. deep, each containing 30 ce. of oe of each salt while distilled water served as control. The concentration of the salts is indicated in the table. The beakers were kept in a room of normal temperature, and evaporated water was supplemented with distilled water from time to time to keep the solutions always in their initial concentrations. After thirty- six days, the difference of their development was very striking. The measurements recorded on the following page were then made: It is assumed that the plant is adversely affected by the salts, if the length of root be half that of the control plaffts, even though the length of leaf be greater than that of the control leaf. In this case as in the previous experiment, the growth of the seedlings was injured or stimulated by each salt according to the concentration. In the concentration at which the toxic action ceases, the stimulating action began and attained its highest point at certain definite dilutions. The growth was injured by mag- 240 Influence of Salts upon Growth of Rice Plant 9 A L 9 9 c sa Zz peyeu ee $}00.1 jo Jaquin . SIT OOT| OOT| GOT) Got} SOT) Z sao to ae yoo1 Jo yyZuery |} * -ayvUOgaeaiq UMIpog 001 OIL | OOT| S2T| OZt| 96 | ge LI “Wi -- yeay Jo yySuey | 9 ip 1 l fé | 9 poieu i poyeu Sececereseseee s$j001 jo requinn i Soa CIT Get | OOT| 84 | OF | & does ‘cone eg, Oor Jo qyZuay |} ~~ - -ayeuoqivo umipog OOT |. OTT] OTT | OGL | OGL} $8 ¢ ATWO A Nanaia oer yea] Jo yySueT 9 L L 9 9 9 pone $j0O1 JO JaquINN : GIT OT | OZE| O€f| SZt| 96 pee on ee “g-gn 50 qyauey |p apoyo uMIpog O01 OIL} OTT | OIT| 2et} Sst} ge 7 | yea] JO YyBue'T 9 Vi L 9 p ¥ : peywu Rl i oe $}001 JO JoqUINN | CIT 08 cL G8 OL og eee ee ee ee 4yoor jo yyaue'y pL e+e e- azeydms umrpog 00T OOL| Oct | GhI| goT| OST) £ ZI So pyaT JO OTT | BS as oe TE el ae a | . GIT Ofl | O6T | 06 | G2 | OF sp008 9 i oor jo y}Bue'T i epHojys wMiayeD 001 G6 | ShI| OIT| SOT] $8 | 8 i Ra lies 389] JO yu] | 9 9 9 9 ¢ ¢ PACORAPE Pan Sart $1001 Jo soquiny GIL | hI | OZ} OOT| Gor} O8 hg hp A oat a at | tl le ar OOT | 401] GOT] OOT) GOT] &8 | 9% | ton yea] JO 4IBUIT |} » 5 P x 2 : pawn [ooo s}00z jo Joqun yy GIT | OOT| 08 | 08 | 09 | OT Be AN A ae }0OI Jo YIBueT |} ~-oyeydyns umtsousEpy 001 OZT | OIT| O&f | oor} S8 | O8 ae Gis tinea +J¥2] JO YABUIT | nowanog, | S008 | 2061 |" 208 | oe x x ° aasa Srivs ~@NOLLVHLNSONOD Oh pihpeeie K. Miyake 241 nesium sulphate in concentration greater than ;%, and _ highly stimulated by ;%5. Magnesium chloride was also toxic in con- centration greater than ,;3, and attained highest stimulating point in concentration of ;;;. The toxic concentration of cal- cium chloride, sodium sulphate, sodium chloride, sodium car- bonate and bicarbonate in each case was greater than ,%;, X,, iis, 140, * Tespectively and highest stimulation was reached in ‘dilution of ;3,5, 445, 140, s¥o and ;%, respectively. For convenience of comparison, the. concentration of toxicity and stimulation of the seven salts in the two experiments are brought together in the following table. = | 1 einen " DILUTION OF HIGHEST STIMULATION : ! SALTS USED _—— — Menon AB soe | Experi- | Experi- Experi- Experi- ment I . ment II ment I | ment IT if | greater | greater than | than | | Magnesium sulphate... .| vo0 (Fo) 100 ro00 (suo) | s00 Magnesium chloride. .. : mo Ge | x soou (r¢00) xenu ‘Calcium chloride ...... | zoo (reo) | YoU 10005 (x00) To0d Sodium sulphate ...... ros (30) | oS roo (en) Las Sodium chloride....... | My (Qa |. te res (ya) vos Sodium carbonate ...... 735 (25) rte 200 tO ry9u (yan tO 390) 300 Sodium bicarbonate..... 45 (4) | zoo (roo) 300 As seen’ in the table, both results almost coincide on the toxic and stimulating point. A slight fluctuation of these points is probably due to the fact that the plant growth varies to a cer- tain extent with the temperature and other factors, since these * experiments were not carried on at constant temperature and under identical conditions. Conclusions. From these two experiments we may safely conclude as follows: - 1. The alkali salts under examination act as agents both toxic and stimulating upon the growth of rice seedlings, according to their concentrations. 2. The toxic concentrations of magnesium sulphate and chlo- ride, caleium chloride, sodium chloride and carbonate are greater . 242 Influence of Salts upon Growth of Rice Plant than ,¥, while sodium sulphate and bicarbonate are greater than N Bor 3. The highest stimulation is observed in the dilution of ;%, for magnesium sulphate, ;%,, to ;%,, for magnesium chloride, re00 to ss for calcium chloride, 5, to .4; for sodium chloride, +x, to ;X, for sodium carbonate and bicarbonate. Part II. ON THE ANTAGONISM BETWEEN THE Toxic EFFEcTs or Two SALTS UPON THE GROWTH OF RICE SEEDLINGS. The results of experiments with a single salt solution have been reported in the preceding section, but they cannot be correlated with our knowledge of alkali soils, since, as Kearney and Cameron® pointed out, in nature we have always to do with a mixture of salts and never with single solutions. They found, as in Loeb’s striking experiments with marine animals, that by adding sodium salts to the solution of magnesium salts the critical concentra- tions of the latter could be raised considerably. In the case of Lupinus albus and Medicago sativa, the neutralizing effect became enormous when salts of calcium were added to the solutions of sulphate and chlorides of magnesium and sodium. The physiology of the decreasing toxicity of a salt due to the presence of a second salt in the solution, was specially discussed by Osterhout? from the view-point of Loeb’s conception of a “physiologically balanced salts solution.’’ It has been shown that marine plants as well as marine animals are very sensitive to pure salt solutions, but thrive well in solu- tions containing a mixture of salts, even though each component is pres- ent in an amount that is toxic in pure solution. A mixture of the more ‘important salts present in sea water, each at about the concentration at which it occurs in the sea, was found to be the best medium for the growth of marine algae. The same phenomenon has been observed in the case of land plants. Kearney and Harter* investigated the neutralizing effect of caleium sulphate upon the toxicity of magnesium and sodium salts withgeight ee ee — —— ‘ Bulletin N No. 71, Bureau of Plant Industry, U. 8. Dept. of AguleTtubal 1912. ’ This Journal, i, pp. 363-369, 1906; Bot. Gaz., xlii, pp. 127-134, 1906; Univ. of Calif. Pub, Bot., ii, p. 317, 1907; Jahrbr. f. wiss, Bot. xlvi, p. 121, 1908; Bot, Gaz., xlv, p. 117, 1908; Univ. of Calif. Pub. Bot., iii, pp. 331-337, 1908; Bot. Gaz., xiviii, pp. 98-104, 1909. * Bulletin No. 113, Bureau of Plant Industry, U. 8. Dept. of Agriculture, 1907. K. Miyake 243 different land plants and found that the presence of calcium sulphate tends greatly to diminish, not only the differences between different species as to their tolerance of magnesium and sodium salts, but also the differences between the latter in their toxicity to the same species. The neutralizing effect of calcium sulphate is generally much more marked with magnesium than with sodium salts. In 1907, Benecke® studied the poisonous action of various salts upon the growth of spirogyra. The result of his investigation was summarized as follows: Chloride, nitrate, sulphate and phosphate of sodium, potassium, magnesium and iron are more or less poisonous, and among these cations iron and magnesium are more poisonous than potassium, sodium is less poisonous than potassium; among the anions, chlorine is least poisonous. The toxicity of these anions and cations can be neutralized or decreased by the addition of calcium ions. Loew and Aso!® also studied the same subject in relation to spirogyra and observed that calcium salts can pre- vent the toxic effects of magnesium salts while potassium salts can retard but not entirely prevent the injurious action of the same. Takeuchi" has pointed out, at the end of his investigation on the behav- ior of algae to salts at certain concentrations, that the injurious action of magnesium salts can only completely be overcome by calcium salts, and - not by sodium or potassium salts. This has been observed not only with algae, but also with young plants of barley and maize which were deprived of their endosperm. Hansteen” has recently investigated the antagonism between cations upon the growth of wheat seedlings and shown that the pure solutions of potassium, sodium and magnesium salts are more or less injurious accord- ing to their concentrations. But in combination with calcium salts, their injurious effect on the growth of leaves, roots and root-hairs is greatly decreased. oe Toxic and antagonistic effects of salts as related to ammonia formation by Bacillus subtilis were also investigated by Lipman" and the following conclusions were reported: 1. Each of the four chlorides (CaCl,, MgClo, KCl, NaCl) is toxie for Bacillus subtilis, in the order given, the first being the most toxic and the fourth the least. This is different from the results with higher plants, where magnesium is the most toxic and calcium the least. 2. A marked antagonism exists between calcium and potassium, magnesium and sodium, potassium and sodium. 3. No antagonism exists between magnesium and calcium but the toxic effect of each is increased by combination with the other. This is just the opposite of what has hitherto been found for plants. ; 9 Ber. d. bot. Gesellsch., xxv, p. 322, 1907, 10 Bull. Coll. Agric., Tokyo Imp. Univ., vii, pp. 395-409, 1906-08. 4 Tbid., vii, p. 628, 1906-08. 12 Nyt. Mag. Naturvidensk., xlvii, pp. 181-192, 1909; ref. Exp. Sia. Rec., U.S. Dept. of Agriculture, xxiii, p. 28, 1910. 18 Bot. Gaz., xlviii, pp. 105-124, 1909. 244 Influence of Salts upon Growth of Rice Plant As above stated, it is clear that the toxicity of a single salt solution may be neutralized by the presence of a second salt, especially calcium salts. It was desirable to investigate the in- fluence of salts common in alkali soils upon the growth of rice plants. We have therefore selected chloride of sodium, magne- sium and calcium, and sulphate of sodium and magnesium as the salts to be tested and have examined the respective antago- nisms between these salts in combination. I. Experiment with NaCl and MgCh. The antagonism between sodium and magnesium chloride was established with young rice seedlings, about 10 mm. high, which were grown in distilled water from seeds of almost uniform size and specific gravity (1.185-1.200). Beakers of about 5.5 em. in diam- eter and 7 cm. deep, each containing the solutions noted in the tablet were used for the experiment, the seedlings being placed _ in the culture fluids on November 19 (1912). Five seedlings were grown in each culture in the greenhouse and the evaporated water was supplemented with distilled water from time to time to keep the solutions at the initial dilutions. After twelve days, the dif- ference in development in the respective cultures was very remark- able. The following determinations were made: rovrions oMED Lamgmn or|uaxorn oF] une mm mm, te NaCl, 30 cc../45.......0e--..ssees...| See 25 1 fy NaCl, 25 cc. +5 MgCly Bee.........., | 87 40 6 ty NaCl, 20 ce. + t5 MgCh, eC. .snmee... 53 30 4* Ty NaCl, 15 cc. +t MgOly, 15 0c..........) 50 32 3* +s NaCl, 10 ce. + to MgGls, 20 cc........... 45 28 2* ® NaCl, 5 cc. + to MgOls, 25 cc........... 45 20 i te Mais, 90 Coi....... cde. ...aees.. dnt. . 80 50 | 7 | K. Miyake 249 VII. Experiment with NaCl and MgSO,. The antagonistic action of Na*, Mg**, Cl’, and SO,” ions on each other was established with sodium chloride and magnesium sulphate in the same manner as in the case of Experiment I. The following result was obtained. soLutio%s usp ee | mm, . mm, eM abe) eee... ae a a tx NaCl, 25 ec. + fy MgSOu, 5ec.......... 55 | | 6* to NaCl, 20 ec. + 7h MgSOu, 10 cc.......... 50 | 33 | 4° qo NaCl, 15 cc. + te MgSO, lidcc..........| 45 | 80 | 3% NaCl, 10 cc. + 3 MgSO, 20cc....:.....) 48 | i ae | +; NaCl, 5 cc. + te MgSO,, 25 cc.......... 48 32. Ci 1 Were BO OO, mr ua iss........... OO i a | 1 mptted Water, Gtr. a. sts. ss.......... 50 | 80 | 6 bed Only one root was well developed. From the table, it is clear that the observed antagonistic action between these ions almost coincides with the results of Experi- ments ‘I and V or VI. VIII. Experiment with Na,SO, and MgCl. The same antagonism as in the preceding experiment was again examined with sodium sulphate and magnesium chloride as before. The result obtained almost coincides with that of the preceding experiment as will be seen in the following table. } : » SOLUTIONS USED LENGTH OF LENGTH OF| NUMBER LEAF (| ROOT | OF ROOTS mm. mm. to Ieee, 30 oc... ,cuRme.... we... cca. 45 35° | 1 to NasSO,, 25 cc. + 79 MgCl, 5ec:....... | 80 40 6* ty Na2SOu, 20 cc. + 7p MgCh. 10 cc. 80 38 6* to NaSOu, 15 cc. + te MgCh, l5ec......... x 65 45 4* to Na2SOu, 10 cc. + fo MgCl, 20ec........ | 50 30 | 1 yy Na.SO,, 5 ce. “bh * MgCl, 7) ei 60 25 1 . a ee oe | 55 40 | 1 Distwled water, 30 cc....055.....05....0.. | 80 50; ee 7 * Only one root was well developed. 250 Influence of Salts upon Growth of Rice Plant IX. - Experiment ASO, and CaCl. The antagonistic action of Na*, Ca**, Cl’ and SO,4” on each other was examined with sodium sulphate and calcium chloride in the same manner as in Experiment II. The result obtained was as follows: | LENGTH OF | SOLUTIONS USED waned neti oF | Balybesitic mm, mm. to Na,8O,, 30 cc..... 626.50. eee Bie) ABB ed 1 to Na.SOx,, 25 ce. + SS CaCh, eo ee 80 90 | 9 #; NasSO,, 20 cc. + 7h CaCle, 10 ce......... 80 |» 90 | 7 to NaSO,, 15 cc. + fo CaCle, 15 ee......... 70 © | 50 | 7 #, Na.SO,, 10 cc. + 7 CaCle, 20 ce......... Aa ee MY to} Na:SO,, 5 cc. + 75 CaCle, 25 cc......... 50 SBa- | 4 #, CaChs, 30 Ce... s.0cs2 0+... 45 oot Distilled water, 30 cc...........ccecceeeeess 80 Lea As will be seen in the above result, in a suitable mixture of Na’, Ca**, Cl’ and SO,” ions, their toxic effects completely dis- appear. It was also observed that the combined antagonistic actions of cations and anions have a more favorable effect than that of one of them. ,X. Experiment with MgSO, and CaCl. The antagonistic action of Mg**, Ca‘*, Cl’ and SO,” ions on each other was established with magnesium sulphate and calcium chloride as in Experiment IV. The following result which is similar to that of the preceding ie was obtained. | (LENGTH OF LENGTH OF NUMBER SOLUTIONS USED fs. LEAF nooT | Or ROOTS mm, / mm, te MESO, 30 cc....... Way. ee... cuecss soa 50. |- 8 2 1 is MgSO,, 25 cc. + fy CaCls, 50cc......... 135 5 | @ fs MgSO,, 20 ce. + fy CaCls, 10 ce.......... 135 50 6 fo MgSO,, 15 cc. + fe CaCls, 15 ce.,........ bib te fs MgSO,, 10 cc. + fy CaCl, 20 ce.......... 85 | 25 4 fs MgSO., 5c. + fy CaCls, 25 ce.......... 80 | 18 3 oo ey a ee - co) ae} r K. Miyake 251 Conclusions. The results obtained in all of these experiments, may be sum- marized as follows: 1. The salts under examination, used separately, are very poi- sonous in 3; concentration upon the growth of the rice plant, but when the two salts are mixed with each other in a suitable proportion, the toxic effect of each salt more or less completely disappears. This result is of great importance in alkali soil investigations. 2. The antagonistic action of salts is due to that of the ions formed by the dissociation of the salt. 3. In general, divalent cations are markedly antagonized by monovalent cations, but on the other hand, monovalent cations do not strongly antagonize divalent cations. 4. Among the divalent cations, calcium shows a more marked antagonism than magnesium. 5. The antagonism between Cl’ and SO,’’, though it is small in comparison with that between cations, is also present in no slight degree. Parr III. On toe AnTAGONIstTIc ACTION OF SODIUM AND PorasstuM SALts. In Part III we have specially undertaken to test the ahtagonism between sodium salts, potassium salts, and sodium and potassium salts. Chloride, sulphate and nitrate of soda and potash were selected as the salts to be tested, and the following, experiments were made: I. Experiment with NasSO, and K2SO,. Eight beakers of about 5.5 em. in diameter and 7 em. deep each containing 30 cc. of culture fluids, of the composition noted in the table, served for the experiment. One beaker containing distilled water served as control. On February 1, 1913, young rice seedlings which were grown in distilled water, were trans- planted into the beakers, each receiving five healthy individuals of uniform size (about 20 mm. long) and kept in a greenhouse. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2. 252 Influence of Salts upon Growth of Rice Plant The evaporated water was replaced with distilled water from time to time. A decided difference in plant growth was noticed from day to day. On February 18, measurements were made with the following result: SOLUTIONS USED eee or ee on noon | mm. | mm. 7 NasSO,, 30 cc... ....< 300s eS eSBs 208e 2* to NasSO. 25 cc. + ty KiSO4 5ee........., 67 | 43 | 6. 5 NasSO,, 20 cc. + fy K2SOu, 10 ce......... : Pi Ba 6* To Na,SO,, 15 cc. + to KsSO4, 15 00.......... 47; 37 | 6 fy NaSO,, 10 ce. + 5 K.SO,, Oo ae : 47} 35 6* 75 NaSO,, 5 cc. + ip K:S04 25ce......... 50 | 43 | 7 Ks80., 30 c0..:s.s. fis e e 40: EReer | ee Distilled water, 30:cc: =...) 4 auueeeeeenee cu. \ | 68 43 7 * Only one root was well developed. _ II. Experiment with NaCl and KCl. The antagonistic action of sodium and potassium on each other was again established with sodium and potassium chloride in the same manner as in the preceding experiment: SOLUTIONS USED ae | . s mm. mm, te N@Ol, 30 06.2.5... ..> 20MM cece ence 35 35 | Q* to NaCl, 25 cc. + te KCl, 5ec........... «| 60 45: ibis Gm to NaCl, 20 ce. + ty KCl, Wee.............] 55 45. ge," Of fo NaCl, 15 ec. + to KCl, IB cc............. 55 40 | 6 to NaCl, 10 cc. + fy KCl, 2ec.............} 42 45 | 7 te NaCl, 5 cc. + to KOU 25 cc............4 50 le fe Ol, 80 co.ct..... ees scat ss ce 40 25 | 6* Distilled water, 30 cc. same. >... .-se00--s sen 68 43 | 7 *Only one root was well developed. Ill. Experiment with NaNO; and KNOs. The antagonistic action of sodium and potassium on each other was once more tested with sodium and potassium nitrate and the following result was obtained: K. Miyake 253 SOLUTIONS USED LENGTH OF LENGTH OF NUMBER | LEAF ! ROOT | OF ROOTS | mm. | mm. ) Piaby og ar Wert, V8 ee a rr tx NaNOs, 25 ce. + tp KNO;, 5cc......... | Oe Ste 7 +5 NaNO;, 20 cc. + 7p KNOs;, 10 cc......... | 50 aa |. GF + NaNOs, 15 cc. + 7¢ KNOs, 15 cc......... eRe: | 7 #5 NaNO, 10 cc. + ty KNOs, 20ce.........) 48 | 35 | 7* # NaNO;, 5 ce. + $y KNOs, 25 cc......... 50 cas) EE 0 ne t . 40 a6. | 4" Beueeriled water, 00.CG,v.u...-.............0. | 068 4 | 7 *Only one root was well developed. P IV. Experiment with KSO, and KCl. The antagonism between SO,” and Cl’ ions was already observed in Experiment No. V of Part II with sodium sulphate and chloride. This was then once more examined with potassium sul- phate and chloride in the same manner as in Experiment I, and the result, which coincides with that of the experiment with so- dium salts, is given in the following table: . LENGTH OF |LENGTH OF| NUMBER SOLUTIONS USED LEAF ROOT OF ROOTS mm, mm, oa K.SO,, DR | a 40 35 » ) 6* +r K2S0,, 25 ec. + 4e KCl, 5ec............. #47 43 7* +5 K280,, 20 cc. + fp KCl, 10 cc............ 47 35 . 6* ‘+ K.80,, 15 ce. + 7p KCl, IBee............. 40 32 6* ty K280,, 10 ce. + 7 KCl, 20 ce............ 50 50 7* tw K280,, 5 cc. + to KCl, 25ee........... 47 35 7° eel 30 cc. ... ae... GE... ae 35 25 6* Distiied water, 30 oGce.....dee...... oes. 68 43 7 * Only one root was well developed. V. Experiment with NasSO, and NaNQ3. The antagonism between SO’ and NO,’ ions on each other was established with sodium sulphate and nitrate as in Experi- ment I. The following result was obtained: 254 Influence of Salts upon Growth of Rice Plant SOLUTIONS USED ‘nate Ox page Oa Ret peed del ' mm, mm, Te NaSO; 30 cc........ 2.54 ees 5 Pe SB 35 2* to Na2SO,, 25 ec. + i> NaNOs, 5ce........ be. 62 35 5* ts Na.SO,, 20 ec. + 7 NaNOs, 10 ec........ ey 2b Be 5* jo Na,SO,, 15 cc. + iy NaNOsy, 15 ce... ..... | 40 35. | > oF #5 Na.SO,, 10 ec. + 7 NaNOs, 20 ce........ | 50 35 5* to Na:SO., 5 cc. + iy NaNOs, 25 ce........ P45 30 pam qe NaNO,, 30 cc..........:- sc epee fey a8 35 | 9 Distilled water, 30 cc... ).4 semen 68 43 | 7 *Only one root was well developed. VI. Experiment with K2SO, and KNOs. An experiment similar to the preceding one was made with potassium salts and a similar result was obtained, as will be seen in the following table: SOLUTIONS USED ges ON | cee mm mm CO a ll | 40 35 6* +, K:8Ou, 25 cc. + fo KNOs, 5 ce.......... | 658 45 7* +5 K.80., 20 cc. + 76 KNOzg 10 cc.......... 53 38 6* fr K2S8O,, 15 cc. + to KNGgmpcc.......... 45 38 6* +5 K28O,, 10 cc. + t6 KNOs, 20cc.......... 50 42 6* 75K280., 5 cc. + to KNOs, 25 cc.......... | ao 35 6* A, ENO, 80 06.05... :. gL... ss. ota 40 35 4* Distilled water, 30 cc... .Gg5y........-......| 68 43 7 *Only one root was well developed. VII. Experiment with NaNO; and NaCl. The antagonistic aetion of NOs’ and Cl’ ions on each other was examined with the solution of sodium nitrate and chloride in the same manner as in the case of Experiment I. The transplanting of young rice seedlings, about 25 mm. high, took place on Feb- ruary 20 (1913) and the plants were measured on March 5 with the following result: K. Miyake 255 LENGTH OF |LENGTH OF| NUMBER | ores Pes LEAF ROOT OF ROOTS eo eMN ok. ke en cape * NaNO;, 25 ce. + io NaCl, 5 cc.......... +. NaNO;,, 20 ce. + fy NaCl, 10 cc.......... mm 43 1 63 | 59 * NaNO, 15 cc. + iy NaCl, 15 cc.......... 55 CE 50 57 46 78 Sas +, NaNO, 10 ce. + to NaCl, 20 cc.......... mis -7 §* 4s NaNO;, 5 cc. + to NaCl, 25cc.......... NS eS SCION WAUGF, GU GOli.ss..-...-----. 0.0008 BRS *Only one root was well developed. VIII. Experiment with KNO3 and KCl. An experiment similar to the preceding one was made with potassium salts. The result obtained was as follows: SOLUTIONS USED See Rags a : | mm, mm, mi SO NMMIIINIOD.... PEER ces he sess... | 45 25 3* +; KNO;, 25 ec. + to KCl, 5cc............ | 60 35 6* tr KNO;, 20 cc. + iy KCl. W0cc............ 50 35 6* +; KNOs, 15 ec. + to KCl, 15 cce............ | 48 30 6* *. KNO,, 10 cc. + fo KCl, Wee............| 50 40 7 4, KNO;, 5 cc. + to KCl, 25ce............ | 72 | & Ch, Geena oaks UUM. ac... | 50 =| Distilled water, 30 cc.......He............. | = 58 7 *Only one root was well developed. IX. Experiment with K2SO, and NaCl. The antagonism between K*, Na’, SO,’ and Cl’ ions on each other was established with potassium sulphate and sodium chlo- ride according to a method similar to that of Experiment VII. The following result was obtained: 256 Influence of Salts upon Growth of Rice Plant | LENGTH OF LENGTH OF NUMBER aterm Mo LEAF | ROOT OF ROOTS ‘ j mm, | mm. Te KeS0,; 30 cc.:.... 0... ee | 47.) 2b ae to K2SO,, 25 cc. + to NaCl, 5ce....... 68° | (35 are +; KsS0,, 20 cc. + 74 NaCl, 10ce...........) 59 | 35 .| 6% To K:SO,, 15 cc. + to NaCl, 15 e¢...........): 52 | 30 (x ts K2SO,, 10 ec. + 35 NaCl, 20 €6...2.0 2222s) 60 33° oe to K2SO,, 5 cc. + io NaCl, 25 e6...........) 65 | 35 + 1s we NaCl, 30 cc.......52:.. See bit MB | 25° 4 3* betes 68 ae Distilled water, 30 cc... . eee oe = Only one ‘Toot was well ‘developed. X. Experiment with NasSO, and KCl. A similar experiment to the preceding one was made with sod- ium sulphate and potassium chloride. The following result which coincides with that of the previous experiment, was obtained: SOLUTIONS USED see eee | eee mm, mm ¥e NaSOg 30 cc........:) Rss... 45 | 20 Q* to NaSO,, 25 cc. + fo KCl, B5ee........... 72 | ° 25 7* to NasSO,, 20 cc. + to KCl, W0ce........... 65 | 2% | 7 * 35 NasSO,, 15 cc. + fo KCl, IScc........... 60 18 6* tx NaSO,, 10 cc. + 7) KCl, Wee............ 644 | «| 25 7* $s NasSO., 5 cc. + to KCl, 25cc........... 71 28 5* a KCL «= 30 ec... . . sae - ss cc nas ss. 50 | 2 7* Distilled water, 30 cc...Jjumme.............. 7 / 58 7 . Only one root was well developed. XI. Experiment with NasSO, and KNOs. The antagonistic action of Na‘, K*, SO,” and NO,’ ions on each other was examined with sodium sulphate and potassium nitrate in the same manner as in Experiment VII. The result obtained was as follows: > ° K. Miyake 257 soLom0Ks vse eee om mm. | mm. | iy NaeSO,, CO LA a 45 | 20 2* +; Na2SO,, 25 cc. + 7 KNO;, 5 cc......... 65 | aoe} 6° io Na2SO,, 20 ce. + ip KNOs, 10 cc......... maa |. 7* +, Na.SO,, 15 cc. + te KNOs, 15 ce......... 60 {= 20 3* +, NasSO,, 10 cc. + ty KNOs, 20 ce......... 77 | 30 7* +t NaS8O,, 5 cc. + ty KNOs, 25 cc......... 60 | 35 4* ee Core ee... sae res, | Bt DNUHAOG WAULGT, GU OGs0) tees 6.--.. 5. eee 78 | 58 | 7 * Only one root was well developed. ; XH. Experiment with K,8O, and NaNO. The same antagonism as in the preceding experiment was again observed with potassium sulphate and sodium nitrate. A result similar to that of Experiment XI was obtained as will be seen in the following table: SOLUTIONS USED bie or asc | ates | mm. mm, SOR Se ee 47 35 7 : +, K2SO,, 25 cc. + to NaNOs, 5 ce......... | a 30 7 ) *. K280,, 20 cc. + ty NaNOy, 10 ce.........| 60 20 5* , K.80,, 15 cc. + fy NaNOg, ld ce.........| 60 25 5* qr K2SO,, 10 ec. + te NaNOs, 20ce........., 60 35 6* to K280., 5 cc. + ty NaNOs, 25cc.........; 72 40 5* we ENO:,00 GOab ites... . MGs. .iekae--- | 43 20 1 Distilled water, 30 c¢..........00....e00 005. | 7% | bs 7 *Only one root was well developed. ¢ XIII. Haperiment with KNO; and NaCl. The antagonism between K*, Na*, NOs’ and Cl’ ions in com- bination with each other was examined with potassium nitrate and sodium chloride in a manner similar to that of Experiment VII. The following result was obtained: 258 Influence of Salts upon Growth of Rice Plant SOLUTIONS USED cate Re a "a dealt mm, | mm. Te IKNOs, 30 cc... .. 5, 0 oe. 45 25 3* ts KNOs, 25 cc. + 7) NaCl, 5ce...:....... 65 30 5* to KNOs, 20 cc. + ty NaCl, 10 ce........... 62 30 5* ts KNO;,, 15 cc. + 75 NaCl, 15 ce........... 60 25 4* ts KNO;, 10 cc. +35 NaCl, 20 ce........... 63 25 5* js KNO;, 5 cc. + 75 NaCl, 25 ce........... 75 35 7 fd we NaCl, 30 cc............ see. 46 25 1 Distilled water, 30 cc. 2.) ) 33a eee 78 58 vf 7 Only one root was well developed. XIV. Experiment with NaNO; and KCl. The antagonistic phenomenon observed in Experiment XIII was again tested with sodium nitrate and potassium chloride. The following result, which almost coincides with that of the preceding experiment, was obtained: SoLUTIONS USED eee | coe mm. "mm, we NaNQs, 0 co........-. REMIND cvs... — 20 1 tx NaNOs, 25 ec. + 4 KCl, 5 ee | 75 30 7 to NaNO,, 20 cc. + 75 KCl, 10 ce........... 65 35 6* to NaNO,, 15 cc. + > KCl, 15 ce...°....... | 35 6* tc NaNO,, 10 cc. + 75 KCl, Wee........... | "@4 25 4* to NaNO;, 5 cc. + tp KCl, 25ce........... | 70 25 7* H KOI” 30'cec..... AE... cus... | 50 25 7* Distilled water, 30 cc........... 00.0 cece eee 78 58 . 7 ~®Only one root was well developed. Conclusions. From the above results we may conelude as follows: 1. Sodium and potassium salts are antagonized by each other. The curve of antagonism between these salts shows two maxima and the location of these maxima is almost constant, occurring at the point of the proportion of 5:25. This coincides with the result which was observed by Osterhout" on wheat seedlings. 2. The antagonism between these salts is due to cations as well as anions. Bot. Gaz., xiviii, pp. 98-104, 1909, K. Miyake 259 3. The antagonism between anions is small in comparison with that between cations. Part IV. On THE ANTAGONISM BETWEEN POTASSIUM AND MAGNESIUM oR Catcrum IOoNs. The antagonism between potassium and magnesium or calcium ions is especially interesting as shown in the experiment of Oster- hout*® on wheat seedlings and of Loeb’? on Fundulus. In these experiments, it was shown that the toxicity of potassium ion is antagonized by magnesium or calcium ions, though calcium shows a more marked antagonism than magnesium. We have under- taken to investigate this relation in the case of growing rice seed- lings, and accordingly we have made the following experiments: I. Eaperiment with KCl and MgCl. The antagonism between potassium and magnesium ions was established with the young seedlings of rice, about 25 mm. high, which were grown in distilled water from seeds of almost uniform size and specific gravity (1.185-1.200). Beakers of about 5.5 cm. in diameter and 7 cm. deep, each containing 30 cc. of the culture fluids, were used for the experiment. The seedlings were placed in the solutions on March 7, 1913. Five seedlings were grown in each culture in the greenhouse and the evaporated water was supplemented with distilled water from time to time so as to keep the solutions at their initial dilutions. On March 24, the difference in development in the respective cultures was very striking, and the following determinations were made: SOLUTIONS USED a ae mm, mm, 2 KOEURO cc... css... aa a 3 40 17 3 qn KCl, 25 cc. + ty MgCle, 5 ce............ 60 16 ) 4 tr KCl, 20 cc. + fy MgCle, 10 ce............| 58 iS oF 6 40 KCI, 15 cc. + to MgCls, 15 cc............| 55 Si to KCl, 10 cc. + 75 MgCle, 20 cc............ 55 Se 5 to KCl, 5 ce. + 75 MgCh, 25 cc............]) 56 pl pm 4 Res O0lbG...;...... 00... . a 43 B3 le 3 Distilled water, 30 cc.............-.0e-0- eae 63 a7 S11 16 Bot. Gaz., xlv, p. 117, 1908; xlviii, pp. 98-104, 1909. 17 Amer. Journ. of Physiol., iii, p. 327, 1900. 260 Influence of Salts upon Growth of Rice Plant In pure magnesium chloride solution the seedlings had grown only 18 mm. in eighteen days; in potassium chloride solution, only 15 mm.; while in distilled water the length of leaf had at- tained to 63 mm. Therefore, it is evident that potassium chlo- ride and magnesium chloride have a poisonous action upon the growth of rice seedlings. This poisonous effect largely disappears when we mix the two salts (MgCle + KCl) in proper proportions. In the mixture # KCl, 25 ec. + 4% MgCh, 5 ce., the growth of the seedlings was most vigorous and their height had reached to60 mm. Therefore, it is evident that in the mixture of magnesium and potassium chloride in favorable proportion, the seedlings grow about twice as much as in pure solutions. It will be noticed that decreasing the proportion of potassium or increasing the amount of magnesium beyond the optimum proportion causes unfavorable conditions for the growth of the seedlings. Accordingly, it is inferred that a small amount of magnesium retards the toxic effect of potassium, and on the other hand, potassium retards the injurious action of magnesium in large amount. II. Experiment with KCl and CaCl. The antagonistic action of potassium and calcium ions on each other was examined with potassium and calcium chloride in the same manner as in the first experiment. The result obtained was as follows: rove Ti0Ns UmED irene ov emma] aoxaa - mm. mm, feeCl, 80 coly...... canes... clea, .. cg | 40 a fo KCI, 25 cc. + 1; Cals, 5 cc............ | 68 51 | 10 te KCl, 20 cc. + 7h Cals, 10 cc............ | 65 2 C&S 8 fe KCI, 15 cc. + {5 Cals, 15 ce............| 65 2 | oe fs KCI, 10 ce. + 7 CaCls, 20 cc......./.... 65 2 | 8 fs KCl, 5 oc. + {> CaCls, 25 ce............ 64 20 | 8 Oe) a a ae oe 35 lar ie .3 Distilled water, 30 66.........cc00s steers. | 68 1 47 —_ ‘ K. Miyake 261 The result obtained was similar to that of the previous experi- ment, but it is clear that calcium has a more marked antagonistic action than magnesium and decidedly prevents the: toxicity of the potassium ion. Conclusion. Potassium and magnesium or calcium salts are poisonous to the rice plant when used separately but when mixed together in suitable proportion the poisonous effect more or less completely disappears. The results coincide with those of Osterhout and form an important factor in the question of soil fertility. - Parr V. Can Barium Anp Srrontium REPLACE THE ANTAGONISTIC ACTION OF CALCIUM? It has been pointed out that the injurious action of certain metallic ions upon the growth of rice seedlings may be perfectly neutralized by the presence of calcium ions. It was of interest to experiment with barium and strontium, which are similar to calcium in chemical properties, to determine whether they exert an action similar to that of calcium. In order to investigate this problem we have used sodium and magnesium chloride as the toxic salts for the following experiments. I. Experiment with MgCl. Twenty beakers of about 5.5 cm. diameter and 7 em. deep, served for the experiment. While one beaker which contained 30 ee. of distilled water served as check, the other nineteen beakers received the solutions noted in the table. Five seedlings, about 25 mm. high, which were grown in distilled water from seeds of almost uniform size and specifie gravity (1.185-1.200), were trans- planted in each of the respective beakers on February 25, 1913, and kept in the greenhouse. The evaporated water was supple- mented with distilled water from time to time so as to keep the culture solutions at their initial concentration. On March 14, the difference in development in the respective cultures was‘ very striking, and measurements were then made with the following result: "262 Influence of Salts upon Growth of Rice Plant SOLUTIONS USED iar oF — Sf) Seas $ | mm. mm, Distilled water, 30 cc.;.... ieaeeeeeres ss 68 60 8 a MeCl,, 30 cc....... «<5. See | 53 12 1 5 MgCle, 25 cc. + ie CaCle, 5 e¢............ | 68 <| = 80 9 * MgCl: 20 cc + 35 CaCls, 10 ec........... | 65 40 10 $5 MgCls, 15 cc. + fo CaCls, 15 ec.......... 62 35 8 +; MgCl, 10 cc. + ty CaCle, 20 ce........-. | 60 i} 2 8 *. MgCls, 5 cc. + 3 CaCls, 25 ¢e........... | $2 a 6 J CaCls, 30 cc....... 55s. .ae ee es | 44 ee 8 #5 MgCl, 25 cc+ ip BaCl, 5¢¢.............| 40 |) 12 1 * MgCle, 20 cc. + 3) BaCls, 10ce.......... | 33 10 1 5 MgCh, 15 cc. + 7) BaCle, 15 ce.......... ae Soa +; MgCh, 10 cc. + 7 BaCh, 20 ¢e.......... Eo ae 7 ie +; MgCl:, 5 ce. + to BaCle, 25 ce........... 28 5 1 46 BaCls, 30 Ce. : i... - ee sss 24 8 1 +; MgCls, 25 ec. + iy SrCle, 5 e....:....... eat) 12 3 Fs; MgCle, 20 ec. + 7 SrCle, Wee...........| 45 8 3 5 MgCl, 15 ec. + 7 SrCle, 15 ce........... fan 5 2 +5 MgCle, 10 ec. + to SrCle, 2ec...........) 35 7 1 #, MgCl, 5 cc. + to SrCle, 2ee............ 30 9 1 ro SrCls,'30 ce.......... J ge 22 8 | 1 The result shows that the presence of calcium in proper pro- portion can exert only a beneficial action, while in the case of barium, on the contrary, a depression resulted. Although stron- tium in suitable proportion retarded the toxic action of magne- sium, it is far inferior to calcium. Il. Experiment with NaCl. Twenty beakers, each containing 30 ec. of culture fluids, served for the experiment. The culture solutions were applied in the same proportion as in Experiment I using NaCl instead of MgCh. Five seedlings, about 20 mm. high, were transplanted on March 7, 1913, and kept in the greenhouse. The evaporated water was supplemented with distilled water from time to time. The plants had developed very well with remarkable differences in growth. The plants were measured on March 24 with the fol- lowing result which coincides with that of the preceding experiment. K. Miyake 263 SOLUTIONS USED meee OF a | Fleer mm, mm, Dyigtatiecd WAeer mses. ss. ok oon eee 65 50 9 ep 0 ak ll rr 44 © 13 | 1 +; NaCl, 25 ce. + fy CaCl, 5cc............ 70 40 | -9 iy NaCl, 20 ce. + 75 CaCl, 10 cc........... 70 40 | 9 qo NaCl, 15 ce. + fy CaCl, 15 cc........... 60 25 8 to NaCl, 10 cc. + 7) CaCl, 20 cc........... 56 20 8 qo NaCl, 5 cc. + 7p CaCh, 25 cc............ eee 20 6 TR Tr | 44 2 | 6 dy NaCl, 25 cc. + 3 BaCl, 5cc............ | 40 oT 3 to NaCl, 20 cc. + 7h BaCh, 10 cc........... ) ma}. 6S ty NaCl, 15 ce. + 7) BaClh, 15 cc........... | 30 | 3 qo NaCl, 10 ce. + 75 BaCh, 20 cc........... | 30 ma) 3 to NaCl, 5 cc. + fy BaClh, 25ce............ | oa m |. 3 a | 98 mo) 63 to NaCl, 25 cc. + fo SrCle, 5cc.............) 50 22 6 qo NaCls, 20 cc. + 7p SrCle, 10 ce........... | az 20 | 4 to NaCl, 15 cc. + 7p SrCh, l5cec............| 45 16-9. 5 to NaCl, 10 cc. + to SrCle, 20cc............; 40 10+" aes to NaCl, 5 cc. + iy SrCle, 25cc.............| 36 16 Le MET OMIGD........ 00a deck se... ..... | 1b | 3 Conclusion. The injurious effect of certain metallic ions upon the growth of rice seedlings may be perfectly counteracted only by the presence of calcium ions. Strontium ions can exert an influence only slightly retarding the toxicity of the metallic ions. Barium ion not only has no beneficial action, but a depressing effect is ob- served. Consequently, it is concluded that barium and strontium cannot replace the antagonistic action of calcium. ard THE DETERMINATION OF OXYBUTYRIC ACID.' By PHILIP A. SHAFFER ann W. McKIM MARRIOTT. (From the Laboratory of Biological Chemistry, Washington University, St. Louis, Mo.) (Received for publication, September 4, 1913.) In 1908 one of us described a method? for the determination of oxybutyric acid in urine, based upon its oxidation by chromic acid with the formation of acetone, the latter being determined by stand- ard iodine and thiosulphate solutions in the usual way. The procedure was combined with the determination of acetone, pre- formed and from diacetic acid, so that the technique as described accomplished the determination of the three “acetone bodies” in the same sample of urine and with practically the same reagents. The advantages of this method for the determination of oxy- butyric acid over those based upon the optical activity of the extracted acid are believed to be the shorter time required; the fact that even very small as well as large amounts may be determined with the same degree of accuracy; and the combined determination of the related ‘‘acetone bodies” in the same sample of urine in one procedure. At the time this method was described it was believed that prac- tically theoretical results were obtained when the specified condi- tions were adhered to. This belief was based on a series of deter- minations on a solution of the inactive acid which had been purified _ by recrystallization of the sodium salt, the titrated acidity being i taken as the criterion of the concentration. These results varied 1 A part of the work described in this paper was done during the summers of 1911 and 1912 in the Biochemical Laboratory of the Harvard Medical _ School. My thanks are due Professor Folin for placing the facilities of his laboratory at my disposal. I am indebted also to Professor Christian and to Dr. Joslin of Harvard, to Dr. I. Greenwald of the Chemical Laboratory of the Montefiore Home in New York, and to Doctor Jesse Myer of Washing: ton University, for a supply of diabetic urine.—P. A. S. * Shaffer: This Journal, v, p, 211, 1908. 265 266 Determination of Oxybutyric Acid from 98.8 to 103.3 per cent of the theoretical values, and were accepted as showing that the oxidation and formation of acetone is practically a quantitative reaction. None of the knowh salts of oxybutyric acid proved suitable to use as a standard, and the levo acid was not used for the purpose because of an inability to get it to crystallize and of a lack of confidence at the time in its pub- lished specific rotation. Gorslin and Cooke*® have suggested slight modifications in the method but did not question its accuracy. Mondschein* also has published figures indicating that he obtained satisfactory results. It now appears, however, that the claim of theoretical results for the method is not altogether correct. Embden and Schmitz,’ in criticizing the method, state that the results are probably somewhat too low since they obtained only a fraction of the optical value when the raw extracts prepared from urines for the polarization method were oxidized by bichro- mate. With urines containing much of the acid 80 to 90 per cent or more of the optical values were obtained on oxidizing the ex- tracts, while the extracts from urines containing little of the acid gave on oxidation only a small part of the optical value. Embden and Schmitz accordingly believe that the ether extracts of such urines contain levo-rotary substances other than the oxybutyric acid and that the optical values are therefore too high, while the results of the oxidation method are somewhat too low. They sug- gest that the true results lie between the values obtained by the polariscopic and oxidation methods. These findings together with our intention of using the method for other work involving the determination of oxybutyrie acid and acetone in blood and tissues, for which there existed no suitable procedure, led us to undertake a review of the various optical points of the oxidation method and of the method based upon the activity of the ether extracts. A series of determinations by the oxidation method on solutions of B-oxybutyric acid as extracted from the urine by ether (Black’s method) gave results which confirmed the statement of Embden * Gorslin and Cooke: this Journal, x, p. 291, 1911-12. * Mondschein: Biochem. Zeitschr., xlii, p. 95, 1912. *Embden and Schmitz: Handbuch der biechemischen Arbeitsmethoden, 1910, Bd. iii, p. 934, 7 4 i { 4 1 4 ) 4 rae ge’. bE Philip A. Shaffer and W. McKim Marriott 267 and Schmitz and show that the optical values of such extracts are greater than those obtained by oxidation. SOLUTION OPTICAL VALUB Daan ha Enraged grams i\n 100 ce. 1. From diabetic urine............ | 28.0 - 26 .4 94.5 (a= —14.84°, 1=2;2 25.75 92.0 [a] P= —24.12°) | 26.1 93.3 25 .95 92.7 25 .67 91.8 25 .83 92.5 25 .33 90.5 Te oly ae 25 .86 92.4 2. From diabetic urine........... 3.98 3.40 85.5 (a=1.92°, 1=2.) 3.42 86.0 3. From non-diabetic urine....... 1.576 1.265 80.3 (a= —0.76°, 1=2.) 4. From non-diabetic urine....... 1.72 1.44 84.0 (a= —0.83°, 1=2.) | | 5. From non-diabetic urine....... 3.50 2.97 | 85.0 (a= —1.69°, 1=2.) 3.07. | 87.6 3.05 | 87.2 6. From diabetic urine..........., 2.06 2.05 | 986 7. From diabetic urine........... | 2.10 2.05 | 97.8 8. From diabetic urine........... 2.04 1.95 | 95.5 9. From diabetic urine........... 2.06 1.0354 93.3 10. From normal urine®............ 0.58 0.21 48.0 According to these results the values obtained on oxidation of the urine extracts vary from about 48 per cent of the optical value in the case of the extract from normal urine (the greater part of which was not oxybutyric acid) to about 98 per cent, usually _ between 85 and 95 per cent. These findings led to a somewhat 6 The same solution after precipitation by basic lead acetate and am- monia gave by the oxidation method values equivalent to 0.058 per cent B-oxybutyric acid, or only 10 per cent of the original optical value; 90 per cent of the levo-rotary substance extracted from this urine was therefore not oxybutyric acid. Oxybutyric acid is not precipitated when its solu- tions are treated with basié lead acetate and ammonia. Occasionally it - may appear that-there is a loss of 1 to 2 per cent, but such amounts are within the limits of accuracy of which the method is ordinarily capable, and are negligible. The following example shows the difference sometimes observed: 0.5885 gram of pure calcium zine oxybutyrate was dissolved and made up THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2. 268 Determination of Oxybutyric Acid lengthy study of the factors which might explain the discrepancies. In order first to test whether the oxidation method as carried out actually yields low results, it was necessary to carry out a series of determinations, on solutions of synthetic and /-oxybutyrie acid of known purity. Since the methods of preparation and purification differ in some particulars from those heretofore used, the details are given below. Purification of B-oxybutyric acid. The impure /-acid as extracted from urine, especially after fer- mentation of the sugar, invariably is contaminated with other organic acids. For the removal of these Magnus-Levy’ has sug- gested the neutralization of the extract with calcium carbonate and the addition of an equal volume of alcohol to the solution of calcium salts, under which conditions he claims the salts of most of the other acids separate, the calcium oxybutyrate remaining in solution. This is not in accordance with our experience; as a rule one gets no precipitation. Calcium lactate and the salts of what- ever other acids that may be present are quite soluble in 50 per cent or even much stronger alcohol. If, however, instead-of calcium the acids are converted into the zine salts and aleohol added, the lac- tate and perhaps also other salts are fairly completely precipitated on standing, while the zine oxybutyrate, quite contrary to state- ments in the literature, is very soluble in both alcohol and water, and remains in solution. We have used this procedure. After filtering off the precipitated zine salts the alcohol is boiled off, the to 250 cc. 100-cc. portions of this solution were (1) treated with basic lead acetate and ammonia, diluted to 250 cc. and filtered; and (2) diluted to 250 ce. and filtered. Determinations were made by the oxidation method in 75-ce. portions of each filtrate, equivalent to 30 cc. of the original solution which contained 0.0706 gram salt or 0.0567 gram oxybutyric acid, Mem. Oxybutyric ol acetone — acid con found of theory (wt, 29.1 52.2 5 92.0 WithouGilead.. cites... cuuscs.saewess ua 29.2 §2.4. 92.4 29.0 52.0 91.8 28.4 . 51.0 90.0 After lead precipitation. ............0...005 23.3 50.8 80.5 23.6 61.3 90,5 7 Magnus-Levy: Brgeb. d. inn. Med. u. Kinderheilk., i, p, 414, 1908. Philip A. Shaffer and W. McKim Marriott 269 syrup cooled, strongly acidified with sulphurie acid (50 per cent) and plaster or anhydrous sodium sulphate added and the mixture allowed to harden. If the syrup is much colored, pure bone black may be added before the plaster. The coarsely powdered material is then extracted with ether in a Soxhlet apparatus. After remov- ing the ether from the extract the residue is dissolved in 10 parts or less of water and if necessary shaken cold with a little pure bone black and filtered. The inactive acid is conveniently made according to Wisli- cenus® by the reduction of aceto-acetic ester by sodium amalgam. The free acid is best isolated by extraction with ether after evapo- rating, acidifying and dehydrating the solution of the salt with plaster or sodium sulphate. For the further purification of the acid, the salts hitherto used, _ with the exception of the sodium salt which is very deliquescent, have in our hands not proved suitable. Repeated and varied efforts to get the l-acid to crystallize have so far not been success- ful. A new double salt of calcium and zine was, however, dis- covered which has been of considerable service. This salt, which is quite stable, crystallizes in long needles or needle-like plates, and while soluble in about 10 parts of water (dl-salt in 7 parts) is much less soluble than the other salts of the acid with which we have worked. The salt is prepared as follows: Calcium-zine oxybutyrate. Equal parts of the free acid are neu- tralized by warming with zine carbonate and calcium carbonate respectively, the solutions filtered and poured together. Both the calcium and the zine salts are very soluble, but when mixed, the double salt crystallizes almost at once if the concentration is greater than 10 per cent (14 per cent in the ease of the dl-salt). The greater part of that remaining in solution is precipitated beautifully crystalline after a few hours on adding an equal volume of hot aleohol. It may be repeatedly erystallized by precipitation by alcohol, though the final crystallization should be from water by evaporation, because the alcohol causes a slight hydrolysis and the precipitation also of a little zine hydroxide which remains undissolved when these preparations are again dissolved in water. After recrystallizing several times from water, the preparations are practically pure. The free acid may be recovered by acidify- 8 Wislicenus: Ann. d. Chem., cxlix, p. 205, 1869. ed Py et 270 Determination of Oxybutyric Acid ing the solutions the salt, selting with sodium sulphate’ pan extraction with ether. Determinations of the specific rotation of the recrystallized l-calcium zinc 8-oxybutyrate in 3 per cent to 9 per cent solutions gave the average value: [a]>=—16.26° PREPARATION I (a). Five times recrystallized from alcohol and once from water. a 0.8742 gram in 10 ce. 1=2.2,«= —3.12° [a]; = —16.25° PREPARATION I (b). Aftef again recrystallizing from water. 2.9646 gram in 50 ce. 1=2.2, a= —2.10° [a]; = —16.15° Preparation II (a). Recrystallized once from alcohol and twice from water. 3.9630 grams in 50 ec. 1=2.2, a= —2.84° [a]; = —16.28° PreparaATION II (b). After again recrystallizing from water. 3.7584 grams in 50 ce, 1=2.2, a= —2.69 [a] = —16.26° Analysis of this salt for ash (CaQ0+ZnO)and calcium corresponds to the formula CaZn(C,H;Os),. Ash: Cautiously ignited to constant weight with small portions of pure ammonium nitrate. if I. .1.2279 grams substance =0.3243 gram ash= 26.41 per cent. Il. 0.5194 gram substance =0.1386 gram ash=26.68 per cent. III. 0.7163 gram substance=0.1898 gram ash= 26,50 per cent. Average found =26.53 per cent. Theory= 26.55 per cent. _ Calcium determination by precipitation as oxalate after removal of zine by hydrogen sulphide gave the following results: I, 1.3014 grams salt =0,3661 gram calcium oxalate (Ca(COO),.+H,0) =7.77 per cent calcium. Il. 1.3550 grams salt=0.3837 gram calcium oxalate =7.76 per cent calcium. II], 1.2817 grams salt=0.3648 gram calcium oxalate «7.79 per cent calcium. Theory 7.74 per cent. ee oem —— * Powdered anhydrous sodium sulphate frequently contains small amounts of material soluble in ether which appear in the extracts. The substance may be removed by several reerystalligations of sodium sulphate from water. ee Philip A. Shaffer and W. McKim Marriott 271 The salt melts with decomposition and not sharply at about 240°C. a Determinations by the oxidation method on solutions of the above preparations of /-double salt and on solutions of the free acid obtained therefrom gave the following results. For the determinations the contents of the distilling flask containing the. oxybutyric acid was diluted to about 600 cc., 30 ec. of sulphuric acid (sp. gr. 1.59) added, and a total of about 0.5 gram of K,Cr.O; in very dilute solution dropped in during the distillation which was continued about three and one-half hours. The acetone in the distillates was titrated with iodine and thiosulphate solutions. I (a). 25 cc. of a solution containing 0.0874 gram salt =0.0703 gram acid were taken for each determination. Grams acetone found Grams oxybutyric acid Per cent of theory 0.0358 0.0642 i 91.4 0.0364 0.0653 93.0 0.0356 " 0.0639 90.9 0.0356 0.0639 90.9 ’ I (b). 50 cc. of a solution containing 0.0886 gram double salt =0.0712 gram acid were used for each determination, which was carried out as above. Grams acetone found Grams oxybutyric acid Per cent of theory 0.0372 0.0668 93.8 0.0373 0.0670 94.1 0.0371 0.0665 93 .4 I (c). Preparation I (b) again recrystallized from water. 0.0885 gram salt =0.0711 gram acid was taken for each determination and carried out as above. Grams acetone found Grams oxybutyric acid Per cent of theory 0.0373 0.0669 94.2 0.0370 0.0664 93.4 II (a). 0.0887 gram salt =0.0713 gram acid taken for each determination. Grams acetong found Grams oxybutyric acid ~ Per cent of theory * 0.036 0.0658 92.5 0 sae 0.0658 92.5 0.0367 § ~ 0.0658 92.5 II (b). 0.0884 gram salt=0.0711 gram acid taken for each determination. Grams acetone found Grams oxybutyrie acid Per cent of theory 0.0367 0.0658 92.6 0.0367 0.0658 92.6 0.0368 0.0660. 92.8 0.0366 0.0656 92.3 II (c). A Siiition of the l-acid isolated from preparation II (b). 272 Determination of Oxybutyric Acid OpricaL vatvuE: ([a], =—24.12°) =4.45 per cent. Acetone found in 50 ce. of a dilution equivalent to 1 cc. of original solu- tion containing 0.0445 gram acid (optical value): Grams acetone found Grams oxybutyric atid Per cent of theory (optical value) 0.0232 0.0416 93 .5 0.0230 0.0412 92.7 0.0228 0.0409 92.0 III. J-Double salt three times recrystallized from water. 25 ec. con- taining 0.0730 gram salt =0.0586 gram acid was used for each determination carried out as above. Grams acetone found Grams oxybutyric acid Per cent of theory (weight of salt) 0.0299 0.0536 91.5 0.0305 0.0547 93.3 0.0297 0.0533 91.0 0.0299 0.0536 91.5 IV. Synthetic dl-calcium zine salt, recrystallized three times from water. 50 ec. solution containing ee gram salt=0.0715 gram acid taken for each determination. Grams acetone found Grams oxybutyrie acid Per cent of theory 0.0371 0.0665 93 .0 0.0370 0.0663 92.8 0.0375 0.0672 94.0 The results by the oxidation method upon the supposedly pure preparations point to the conclusion that the method as carried out yields values from 5 to 10 per cent less than the theory. We have made many efforts to find conditions which would give theoretical results, but so far without complete success. Other oxidizing agents have not proved suitable; raising the temperature by inert salts has proved futile; and the substitution of other acids for sulphuric, or changes in the concentration of acid, do not give higher results. The low results are probably to be explained by a portion of the aceto-acetic acid undergoing the well-known acid decomposition, with the formation of acetic acid instead of ace- tone. On boiling solutions of pure oxybutyric acid or its salts in 5 per cent sulphuric acid the distillates contain little or no acid, but if a large excess of bichromate is added, considerable amounts of acid other than carbonie (about 50 cc. 7) in some experiments from 1 gram of caleium zine salt) pass over, and under these conditions acetic acid is readily recognized in the distillate. It is therefore probable that small amounts of acetic acid are thus formed even Philip A. Shaffer and W. McKim Marriott 273 when the bichromate is added very slowly, and that even under the best conditions this decomposition amounts to about 5 per cent of the oxybutyric acid present. It is not likely that there is a decomposition of acetone once it is formed, for experiments show that acetone is unaffected under the conditions and passes into the distillate very rapidly. Nor does the trouble lie in the titration of the acetone (see page 283). Somewhat higher results can be obtained by adding the bichro- mate very slowly and continuing the distillation for a correspond- _ ingly long-period; and conversely, as was pointed out in the first paper, if the bichromate is added too rapidly, very low results are found, and as noted above considerable amounts of acetic acid are formed. The following determinations were made as described on page 271, except that instead of 0.4 to 0.6 gram K.Cr.O;, - only 90 mgm. were added in very dilute solution during the first _ two and three-quarter hours of distillation and another 99 mgm. during a subsequent hour, the distillation being continued with slow boiling for four hours. The amount of bichromate added was still more than twice that theoretically required to oxidize the oxybutyrie acid to aceto-acetic acid. The results are expressed as oxybutyric acid: ) AVERAGE OF THREE Ai CALCULATED DETERMINATIONS x r wr tees | MA QTALET CARNE | Glin KeOntae KeCr207 grams per cent grams | per cent et eT 0.0711 0.0667 | 93.7 0.068 ) 95.7 es ee 0.0778 | 0.0657; 92.3 ) 0.0696 97.5 Thee: ... eee 0.0715 | 0.0667') 93.5 0.0694 97.0 TViewae....... 5c 0.0711 | 0.0676) 95.2 se 0.0445 0.0412 | 92.7 | 0.0425 | 95.5 It is possible that still slower addition of bichromate and still. longer distillation would yield nearly 100 per cent, but the accuracy of the method would not thereby be increased because when carried out as originally described, the results are practically constant, though it now appears that they represent only 90 per cent to 95 per cent-of the true values. It seems preferable to retain the origi- nal directions and to add a correction of about 10 per cent to the result. 274 Determination of Oxybutyric Acid The oxidation method as applied to urine. At the time of the first attempts to apply the oxidation of oxy- butyric acid to its determination in urine it was found that several substances which normally or occasionally are present, interfere with the results by yielding products which use up iodine when the distillates are titrated. The substances considered were glucu- ronic acids, sugar, lactic, butyrie and formic acids, perhaps leu- cine,!® phenols and some unidentified substances. The possibility of any material interference by these substances was effectively obviated by the introduction of three modifications; glucuronic acids and sugar are removed by preliminary precipitation with basic lead acetate and ammonia; phenol, butyric and formic acid if present are removed during the first distillation of the acidified filtrate, which also removes the acetone (preformed and from aceto-acetic acid), the latter being titrated after redistillation from alkali. Lactic acid if present is in part converted into acetic aldehyde on oxidation with bichromate, and to obviate possible interference from this source the distillate containing it and the acetone derived from the oxybutyric acid is redistilled from alkali and hydrogen peroxide, which completely holds back as acetate the small amounts of aldehyde which may have been formed. The effect of the unidentified substances is almost wholly removed by the precipitation with lead and the final redistillation with hydrogen peroxide. The chief criticism of the method advanced by Embden and Schmitz" is that sugar, which of course is usually present in urines in which it is desired to determine oxybutyric acid, gives rise on oxidation with bichromate to substances which use iodine when the distillates are titrated and that, contrary to the finding of one of us, this interfering product is not removed on redistilling with alkali and hydrogen peroxide, and that therefore the method can- not be applied directly to urines containing sugar. We have con- firmed this statement to the extent that glucose, when oxidized with sulphuric acid and bichromate, does give (as was pointed out 1° Later experiments have shown that leucine does not yield acetone when boiled with sulphuric acid and bichromate under the conditions of the method, ; " Loc. cit, Philip A. Shaffer and W. McKim Marriott 275 in the first paper)” small amounts of a volatile substance which reacts with hypoiodate, and that it is frequently not wholly re- moved by alkali and hydrogen peroxide, which was first believed to be the case. The following experiment illustrates this behavior: 2 grams glucose in 600 cc. water plus 10 cc. concentrated H.SO, were dis- tilled, dropping in 3 per cent K,Cr.0;. The distillate was redistilled after adding 30 ce. of 3 per cent HO, and 5 cc. of 10 per cent NaOH. rms GOONS... ..............00 see 5.2 BUUDUGEEE Sessa... ......... 006 cee 4.3 . Paver @teGsO.G.............. 00 0.6 POREG ees... ee 2.8 - These results as well as the criticism by Embden and Schmitz are, however, quite immateral so far as the method is concerned because it has never been suggested that the method be performed on the urine direct, but only after precipitation by basic lead acetate and ammonia, which wholly removes the effect of the sugar as the following experiments show. A series of determinations were carried out on a normal urine with and without the addition of 3 grams of glucose per 100 cc. urine. In each instance to 50 ce. urine were added 50 to 7§ cc. basic lead acetate solution and 15 ce. concentrated ammonium hydroxide. The mixture was diluted to 500 ce. and 200 cc. of the filtrate, equivalent to 20 cc. urine, were diluted and distilled first with 30 cc. 1 to 1 sulphuric acid (distillate A which was discarded), and then with the gradual addition of potassium bichromate. Some of the distillates were titrated direct and others were redistilled after adding 5 cc. of 10 per cent NaOH and 20 ee. of 3 per cent HO». ce. ty iodine used. NORMAL URINE NOT URINE PLUS 3 PER CENT SUGAR | URINE PLUS 3 PER CENT SUGAR REDISTILLED NOT REDISTILLED REDISTILLED 0.4 0.5 0.6 0.6 0.6 0.8 0.6 0.6 0.7. 0.6 0.7 | These figures are equivalent to about 30 mgm. of acetone from oxybutyric acid per liter of urine. It is probable that this repre- 12 The presence of glucose is clearly without effect upon the results when the method is properly carried out. This Journal, v, p. 218, 1908. 276 Determination of Oxybutyric Acid sents actual oxybutyric acid present in normal urine, but whether or not this is so, the amount is negligible so far as it affects the re- sults obtained from urines containing significant amounts of oxy- butyric acid. When a known amount of oxybutyric acid is added to normal urine, the results by the oxidation method correspond to about 90 per cent of the amount added, as illustrated by the following experiment. 2.540 grams pure [-calcium zine salt were dissolved in 250 cc. normal urine (equivalent to 8.16 grams of the free acid per liter of urine). Found grams 10 per cent added : per liter to result Urine without oxybutyric acid............... 0.029 Urine plus oxybutyrie acid.................. 7.54 8.29 7.50 8.25 7.40 8.14 «7.30 8.08 AVCL ARCs oss oiv'sie soles + =: UII aoe oes os ous au wating 8.18 Amount oxybutyrie acid added ................. Pris: 8.16 There are no essential changes in the procedure from the tech- nique as originally described, but for convenience the descrip- tion may be repeated here with the addition of some further details. * The oxidation method, combined with the preliminary distillation for the removal and determination of acetone and diacetic acid, is carried out as follows: From 25 ce. to 100 ce. or more of urine (usually 50 ce.) are meas- ured with a pipette into a 500 ce. volumetric flask containing 200 ce. to 300 ec. of water. Basic lead acetate solution (U. 8. P.) is added in amount equal to the urine used" and the liquid well mixed. Strong ammonia water, about half the volume of the lead acetate, is next poured in, the flask diluted to the mark with water, shaken, and after a few minutes’ standing, the liquid is filtered, preferably through a folded filter, 200 cc. of the filtrate is meas- ured into a round bottom flask (800 ce. or liter Kjeldahls are convenient) diluted with water to about 600 cc., 15 ec. of the concentrated sulphuric acid and tale or boiling stone added, and the mixture distilled until about 200 ce. of the distillate have collected (Distillate A). The distilling flask must be fitted with a dropping tube and water run in from time to time to prevent the volume in the flask from becoming less than 400 to 500 ce, a te - — a — aeiemitie —- — "If the urine contains but little or no sugar only half the amount or less of lead acetate should be used. Philip A. Shaffer and W. McKim Marriott 277 Distillate A, which contains the acetone preformed and from aceto-acetic acid, and which should be collected in a second Kjeldahl flask, is redistilled (for about twenty minutes) after adding 10 cc. of 10 per cent sodium hydrox- ide.14 The distillate so obtained (A:) is titrated with standard iodine and thiosulphate solutions. The residue of urine plus sulphuric acid from which Distillate A was obtained is again distilled’® dropping in either water, when necessary to keep the volume between 400 and 600 cc., or a dilute solution of potassium bichromate. From 0.5 gram to 1 gram of bichromate will usually be suffi- cient, and not more than 1 gram should be added unless the liquid turns green indicating a great reduction to chromium sulphate; very rarely 2 or 3 grams of bichromate may be necessary, especially if the sugar has not been completely removed. A 10 per cent solution of potassium bichromate is kept on hand and 10 cc. of this, diluted to 100 cc. are measured out for each determination. 20 cc. of the dilute solution (0.2 gram K.Cr.0;) are first added slowly through the dropping tube and then 10-cc. portions every fifteen or twenty minutes until the whole has been added. Should the liquid become markedly green the bichromate must be added at correspondingly shorter intervals and - in amount sufficient to maintain aslight red-yellow color of the chromic acid, which may be detected even in the presence of the green. The distillation is continued with moderate boiling for from two to three hours. The distil- late (B), which should be collected in a liter flask to avoid transference, is again distilled for about twenty minutes after adding 10 cc. of 10 per cent .sodium hydroxide and 25 ce. of 3 per cent hydrogen peroxide. The flask must be heated cautiously until the peroxide has decomposed. This dis- tillate (Bz) is titrated with the standard iodine and thiosulphate. 1 ec. of 75 iodine=0.968 mgm. acetone =1.736 mgm. oxybutyric acid, or 1.035N 10 oxybutyrie acid. 1 ce. of iodine (=13.13 mgm. I.)=1 mgm. acetone=1.793 mgm. Comparison with results by the extraction method. The fact that the oxidation method gives results for oxybutyric acid which are uniformly from 5 to 10 per cent too low, explains in part the differences between the results by this method and the values calculated from the levo rotation of the ether extracts, 4 Tn many instances, when a high degree of accuracy is not required, this redistillation may be omitted and ‘‘distillate A’’ titrated direct; the results so obtained are slightly higher than those after redistillation from alkali. 15 The distillation is actually not interrupted; after “A’”’ has collected, a new receiving flask is adjusted and bichromate solution slowly added i } through the dropping tube. The receiving tube of the condenser must dip ' below the surface of the water in the receiving flask. ‘ ~ 278 ~ Determination of Oxybutyric Acid but it appears that this does not explain the differences in all cases. Were this the only factor the results by the oxidation method would regularly be from 90 to 95 per cent of the optical values of the acid extracted by ether, whether the oxidation method were carried out on the urine as usual, or on the solution of the extracted acid. Occasionally this is the case as illustrated by the following figures: Diabetic Urine. Grams oxybutyric acid per liter. OXIDATION METHOD EXTRACTION METHOD Pa (Black"*) On urine 3 On extracts 4.65 4.75 4.61 5.25 4.73 §.12 5.10 4.80 4.87 5.15 | 4.77 4.81 But freuenilae the results from oxidation of the extracts are much below the expected 90 to 95 per cent of their optical value; — and after treating the extracts with basic lead acetate and ammonia the oxidation results are still lower. As examples the following may be cited; the results are expressed as grams of oxybutyric acid per liter: : Oxidation method on urine......... 3 iat Black’s method (3 hours’ extraction)...... .7.88 Oxidation of extractiggmue..............; 6.32=80 per cent Black’s method (4 hours’ extraction)....... 8.60 Oxidation of extradtammes......-.-.... 555 7.21=84 per cent Black’s method (4 hours’ extraction)....... 8.76 ~“J Oxidation Of extragmemes...sues....-s0as .58=86 per cent .10=81 per cent The results from solutions 2, 3, 4, 5 and especially solution 10. from normal urine on page 267, show the same point—differences = i ° 5 NI 6 Blac kc: this Journal y, p. 209, 1908, For the extraction method 50 ce, to 200 ec. of urine were taken and Black’s directions followed except that regular Soxhlet extractors were used. The extraction was continued for from six to ten hours, usually in two periods. After removing the ether by cautious warming on the water bath,the residue was dissolved to 25 ce.,, the solution was shaken cold with a little purified bone black, filtered and polarized, Parts of these solutions were then subjected to determination by the oxidation method, with or without a preliminary precipitation with basic lead acetate and ammonia, . + \ % h i } i. Philip A. Shaffer and W. McKim Marriott 279 between the results by oxidation and by polarization which are greater than the 5 or 10 per cent already accounted for. As first suggested by Embden and Schmitz!’ there would appear to be present small amounts of a levo-rotary ether-soluble substance which tend to give somewhat too high results by the polariscopic methods. The identity of the substance we do not know. It may be precipitated by basic lead acetate and ammonia, and this preliminary treatment might well be adopted by those who use the extraction method. There are two other points which we have encountered which tend to low results by Black’s technique of the extraction method. Although Black’s plan of dehydrating the evaporated urine with plaster and extraction of the dry material is far more rapid and convenient than the Magnus-Levy liquid extraction, the com- plete extraction frequently requires rather longer than the three or four hours recommended by Black. As a rule we have found in the extract from the second four-hour period 5 or 10 per cent of the amount obtained during the first four hours. A more serious objection is the occasional apparent decomposi- tion of a part of the oxybutyric acid durjng extraction. After ten hours and longer extraction we have repeatedly found only about 90 per cent of the acid which we had added to urine or other solutions. Thus from the urine mentioned on page 276, to 250 ce. of which were added 2.540 grams of pure /-calcium zine oxybutyr- ate (=8.16 grams oxybutyrie acid per liter), duplicate determina- tions on 100-ce. portions by the Black method (10 hours’ extraction) gave 7.675 grams and 7.675 grams per liter. (a= —1.63°, 1=2.2, [als = —24.12°) Subtracting the blank equivalent to the extract from the urine alone, we have 7.43 grams or only 91 per cent of the amount added. The results of the oxidation method (page 276) represent the ex- pected 90 to 92 per cent of the amount added. The reason for this loss on extraction is not altogether clear, but is probably due to an oxidation of some of the oxybutyric acid. We have frequently found in the ether extracts a substance which distils off without the addition of any oxidizing agent and which readily reacts in the cold with hypoiodate to form iodo- 17 Loc. cit. 280 Determination of Oxybutyric Acid form.'$ ‘The substance apparently is not derived from the ether or the plaster, and the preformed acetone and diacetic acid are of course driven off during the preliminary evaporation of the urine. Although the evidence is not conelusive, it is probable that the substance is acetone produced from an oxidation of a little of the oxybutyric acid during the dehydration with plaster or during the extraction. It appears that Black’s application of the extraction method and polarization of the extract usually gives practically correct results, but that the results are somewhat uncertain because they are influenced by the opposing errors of a levo-rotary substance, not oxybutyric acid, tending to give too high values, and on the other hand an occasional incomplete extraction and decomposi- tion of some of the oxybutyrie acid, tending to make the results too low. The extraction method is very serviceable, although from our experience we prefer for most purposes the oxidation method, because the latter is quicker, requires less manipulation and appa- ratus, less urine, and especially for small amounts of oxybutyric acid is, with the correction, more accurate. It is of interest that the parallel determination by the two methods, one of which determines both d and / forms of the acid, has given no evidence for the occurrence in diabetic urine of d-oxy- butyric acid. The asymmetric formation of the levo oxybutyric acid, so far as indicated by available evidence, appears to be perfect. SUMMARY. 1. The method for thé determination of oxybutyrie acid by oxidation to acetone with chromic acid is found to give uniformly about 90 per cent of theoretical values. The results obtained by the method must therefore be corrected by the addition of 10 per cent of the amount found. 2. A procedure for the isolation and purification of oxybutyric acid in the form of a new double salt of calcium and zine is de- scribed. 3. Results by the oxidation method are compared with results obtained by Black’s technique of the ether extraction method. '* The iodoform-forming substance is usually lost during the removal of the ether on the water bath. THE DETERMINATION OF ACETONE. By W. M. MARRIOTT. (From the Laboratory of Biological Chemistry, Washington University, St. Louis, Mo.) (Received for publication, September 4, 1913.) In the determination of 8-oxybutyric acid through oxidation to acetone by the Shaffer bichromate method! with subsequent esti- mation of the acetone by Messinger’s iodimetric titration,’ diffi- culty was experienced in obtaining the theoretical amount of acetone from known amounts of pure 6-oxybutyric acid. It was suspected that either the Messinger titration was inaccurate or that the acetone was incompletely recovered in the distillates. Messinger in his original paper,’ found that in dilute aqueous solutions of acetone, slightly low results were obtained. Colli- schonn! found that the Messinger method gave low results in very dilute acetone solutions. Geelmuyden® using the Messinger method obtained satisfactory results on purified acetone in aqueous solu- tions, but found that, on distillation of.such solutions, loss of acetone of from 5 to 10 per cent was unavoidable, even when ice- cooled receivers were used and the solutions were distilled almost to dryness. Denigés® claims that the first quarter of the distillate from aqueous acetone solutions contains only about 90 per cent of the acetone present. In view of the results cited above, it seemed desirable to inves- tigate the accuracy of the Messinger titration and also to deter- mine whether a dilute aqueous acetone solution could be distilled without loss. In addition I have tested the accuracy of the recently described acetone estimation of Scott-Wilson.? 1 Shaffer: this Journal, v, p. 211, 1908. q 2 Messinger: Ber. d. deutsch. chem. Gesellsch., xxi, p. 3336, 1888. 3 Loc. cit. . 4Collischonn: Zeitschr. f. anal. Chem., xxix, p. 562. 5 Geelmuyden: Jbid., xxxv, p. 503. 6 Denigés: Ann. Pharm. de Bordeaux, 1910. 7 Scott-Wilson: Journ. of Physiol., xlii, p. 444. 281 282 Determination of Acetone The Messinger method. A sample of acetone prepared from the bisulphite compound (Eimer and Amend) was further purified by distillation over potassium permanganate, and redistillation over fused calcium chloride. The product, which was anhydrous and free from alde- hyde, was subjected to fractional distillation and two portions collected between 56° and 57°C. Dilute aqueous solutions containing known amounts of this purified acetone were prepared as follows. Thin glass bulbs of 2 or 3 ce. capacity, and provided with a capillary side tube, were blown. These were weighed and then filled with acetone,’ sealed, and again weighed. Each bulb was introduced into a 2-liter glass stoppered volumetric flask nearly filled with water. The bulbs were broken under the surface of the water by a sharp blow from a glass rod; distilled water was then added to the mark and the contents thoroughly mixed. Bulb I Bulb IT Bulb and acetone............. 3.1928 Bulb and acetone........... 2.5939 aD CMP ys... ss. Sakis Oe aaulb empty. .i.daene sc en sch 0.9194 2.4474 1.6745 Corr. for air displacement. . .0.0024 = Corr. for air displacement. . .0 .0017 Mcdtonémees: . ...soctck cee eens) | Acetone... | aR. 2. in syns extan 1.6762 In measuring out even such dilute solutions as the ones thus | prepared, care was necessary in order to obviate loss of acetone. The solution was forced up into the pipette by air pressure from an atomizer bulb, the neck of the flask being closed by a double- holed rubber stopper. In delivering the solution the pipette was always under the surface of water in the receiving vessel. In this manner 25-cc. portions of the solutions prepared as above, were measured into 700-cc. Florence flasks containing each about 500 ce. of distilled water. A measured amount of standardized iodine solution was then run in, 10 ec. of 60 per cent sodium hydrate added, the flasks stoppered, shaken a little, and allowed to stand for five or ten minutes, after which 15 cc. of concentrated hydro- chloric acid were added and the liberated iodine titrated with * The bulbs were warmed, and then the tips of side tubes dipped into acetone, so that on cooling the bulbs acetone rushed in. W. M. Marriott 283 standard sodium thiosulphate® in the usual manner. The following results were obtained. Sotution I. cc. Toememeuon a0G@e0 60 'CC........ 2. .s..60 eee bees (49 .8) SS Oa a ieee log! Sa 19.0 SR hy i Oo Oa 30.8 Then since 1 cc. of 74 iodine is used up by 0.968 mgm. of acetone 30.8 X 0.968 X 102.8 per cent =30.64 mgm. os beam by cogiag. ee. « 30.62 atl f Avaone: Sororron IT. oe. Toaine' solution added 50 cc:............... .. pees reteen es. (49.8) IDO ALG, areata vos cs... ... 2... oooh weeny ae « 28 .6 SOMPGOO UDC ser noo skc---...-. vs ees aeeens «ena 21.2 4 Then 21.2X0.968X 102.8 per cent = 21.09 ee Dana ? Present by weighing............ 20.95 mgm. The Messinger method, then, is accurate even in quite dilute solutions. Distillation of acetone from dilute solutions. It is frequently necessary to distil acetone solutions before mak- ing the final determinations. I have found, contrary to the results of Geelmuyden and Denigés,’® that if proper precautions are taken, acetone may be completely distilled off from even a dilute aqueous solution and entirely recovered in the distillate. Ten minutes’ distillation is sufficient to accomplish this result, as is shown in the following experiment. An acetone solution was used, 500 cc. of which when titrated by the Mes- singer method were found to contain 33.7 mgm. of acetone. This amount.of solution was distilled from an 800-cc. Kjeldahl flask, using a block tin con- denser connected with a glass delivery tube, the end of which dipped under _ the surface of about 50 cc. of water contained in the receiving flask. No ice % cooling was used or found to be necessary. Distillations were continued for the length of time indicated, and acetone in the distillates immediately determined by the Messinger method. ® The thiosulphate was standardized against pure potassium bi-iodate and also against bichromate and found to be 102.8 per cent of 74. 50 cc. of the iodine solution were equivalent to 49.8 ec. of thiosulphate, on blank titrations. 10 Loc. cit. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2. CAI A ee = 7 284 Determination of Acetone Time distilled Acetone in in minutes distillate mgm. 5 31.1 10 33 .6 10 33.5 10 33.7 15 33 .7 20 33 .6 25 33 .7 30 33.7 The anomalous results of Geelmuyden and of Denigés may pos- sibly be explained by their failure to always have the end of the delivery tube dip under the surface of the liquid in the receiving flask. The mercury cyanide method of Scott-Wilson. Although the Messinger method gives correct results and is the most satisfactory for general use when considerable amounts of acetone are to be determined, yet it is not of sufficient delicacy to determine such small amounts of acetone as occur, for example, in a few cubic centimeters of blood. A more delicate method is that described by Scott-Wilson.! This depends upon the precipitation of acetone as a keto-mercury- cyanide compound with subsequent determination of the mercury by titration with a standard sulphocyanate solution under pres- cribed conditions. + In carrying out the method, as described, several difficulties were encountered, and correct results were not obtained. With certain modifications of the procedure, however, I have found the method to be capable of considerable accuracy with no small] quantities of acetone.” The best results are obtained in the following manner: Dilute solutions of’ pure acetone are run into an excess of the recently filtered reage omt'4 contained i in small Erlenmeyer flasks, allowed to stand " Loe. cit. %* The method is applicable only for quantities of acetone less than five milligrams. “ The reagent is made up as follows: Mercuric cyanide, 10 grams; Sodium hydroxide, 180 grams; Water, 1200 ce, The solution is agitated in a flask and 400 ce, of a 0.7268 per cent solution of silver nitrate slowly run in, At least 30 ec, of the reagent must be taken for each milligram of acetone present. W. M. Marriott 285 twenty minutes and then filtered through an asbestos mat" in a separable bottom Gooch crucible. By first filtering an aqueous suspension of talcum powder so as to partly close the pores of the filter, less difficulty is experi- enced in obtaining clear filtrates. In some cases the first portions of the filtrate are turbid and have to be refiltered. The precipitate is washed with cold water until the washings are free from silver. With the aid of a pointed hooked glass rod the precipitate, mat, and cru- cible bottom are transferred to a 50-cc. beaker, any adhering particles of the precipitate being washed into the beaker with about 10 cc. of ‘“‘acid mix- ture,’’* 1 ec. of ¥ potassium permanganate is added, the beaker covered with a watch glass, and the liquid boiled until colorless. More permanganate is then added a few drops at a time, until a persistent brown color is obtained which does not disappear on boiling for a couple of minutes. The brown color is then discharged by the addition of a few drops of strong yellow nitric acid. The greater the amount of acetone present the more permanganate is required, and it is essential to the accuracy of the method that an excess be added as indicated above, otherwise the results are low. ¥ The beaker is cooled under the tap, 2 cc. of saturated ferric alum added, and a standard solution of potassium sulphocyanate (approximately 0.1 per cent) run in from a burette until a very faint pinkish brown color is obtained throughout the solution. The end point, which consists in the faintest trace of color, can be detected only when the titration is performed on a pure white surface. A control beaker with one drop excess of sul- phocyanate should be at hand for comparison. A whole cubic centimeter of sulphocyanate may be run in after the end point is reached without very greatly darkening the shade. In the calculation of results Scott-Wilson has assumed that the keto-mercury-cyanide compound has the formula HgCOC.(HgCN), and that consequently 1 mgm. of mercury should be equivalent to 0.058 mgm. of acetone. He determined the value of the sulpho- cyanate solution in terms of mercury by titrating it against a known mercury solution. In applying this method to the estima- tion of pure acetone solutions he obtained results about 3 per cent too low. The error he attributed to loss of acetone by evaporation, __ or to impurities in the acetone. My results which follow, have led ~~ me to believe that the error is instead in the method of calculation. _ The dilute acetone solution made up from Bulb I, and used as previously described for the Messinger titration was also used in _ this case. Twenty-five cubic centimeters of this solution were made up to 1 liter with distilled water, and 50 ce. of this latter 4 Filter paper cannot be used as the strong alkali quickly attacks it. © Nitric acid, 40 parts; Sulphuric acid, 5 parts; Water, 55 parts. 286 Determination of Acetone solution, containing 1.53 mgm. acetone, used for each determina- tion. The acetone solutions were each run into 50 cc. of acetone reagent and the estimations carried out as described above. On titration the following results were obtained, and calcula- tions based on the value of the sulphocyanate solution made as indicated. KSCN® ACETONE FOUND | ACETONE PRESENT | ‘‘ACETONE FACTOR” ce. mgm. mgm. Beir ie D061... van 1.44 1.53 0.0646 eo. O OG)... . catenin a 1.47 1.53 0.0635 me. 0.061... ca 146+ | 1.53 0.0637 geo xX O06] eseee ea 1.44 1.53 0.0648 ' A second acetone solution was made up from the same stock solution from Bulb I, by diluting 25 ec. of this to 250 ce. with water. Of this latter solution 10 ce., containing 1.225 mgm. acetone, were used for each determination. ° i KSCN ACETONE FOUND | ACETONE PRESENT | ‘‘ACETONE FACTOR” _— eae ~~ =Cl| tits gm 19 2 RESOOL «ois wives > 8 cea 1 eg 1.225 | 0.06388 18.6 Xe 061. 5... Sit « vial 1.13 1.225 0.0658 The results are uniformly low. The figures under the heading “acetone factor’ represent the value by which each cubic centi- meter of potassium sulphocyanate solution used should be mul- tiplied in order to give correct results for the amount of acetone actually in the solution. The average of these values, which is 0.0644, is then to be taken as the true value of the sulphocyanate solution in terms of acetone. From the foregoing it is evident that the sulphocyanate solution cannot be standardized by its mercury © equivalent, but that solutions of pure acetone of known strength, as determined by weighing or Messinger titration, can be used to advantage. The diserepancy in the results obtained by using the The KSCN solution was standardized against a solution of mercurie nitrate, that had been analyzed for mercury by sulphide precipitation. 1 ee. of KSCN was found to be equivalent to 1.05 mgm. mercury, which from Scott-Wilson's formula, would correspond to 0.061 mgm. of acetone. W. M. Marriott 287 mercury equivalent as a basis of calculation may be explained by the possibly incorrect formula for the keto-mereury-cyanide com- pound or by the reaction not being a complete one. Having thus standardized the sulphocyanate solution, a series of determinations on a different acetone solution was made. The solution used contained 0.172 mgm. of acetone per cubic centi- meter, as determined by Messinger titration. Varying amounts of the solution and of acetone reagent were used in order to test the accuracy under different conditions. ee ce KSCN co en cc. ; ce. cc. mgm. | mgm, 1 50 2.65 X 0.0644 0.170 0.172 1 50 2.85 * 0.0644 0.183 | 0.172 5 50 13.40 x 0.0644 0.863 0.860 oc. | 50 13.50 * 0.0644 0.869 0.860 10 | 100 27.20 0.0644 1.75 1.72 meee 100 26.80 x 0.0644 i. | aa 20: | 100 54.00 x 0.0644 3.48 3.44 20 100 54.40 x 0.0644 3.50 3.44 The method, then, gives accurate results with varying amounts of acetone and the accuracy is not affected by considerable amounts of acetone reagent in excess of the quantity required. As previously mentioned it is necessary to use at least 30 ec. of the acetone re- agent for each milligram of acetone present, or expected to be present. The acetone reagent is not affected by alcohol, but a precipitate forms with very small amounts of aldehydes, chlorides, hydrogen sulphide, or ammonia. In making determinations, therefore, the absence of these substances must be assured. If the acetone solution is extremely dilute so that several hun- dred cubic centimeters are required to make a determination, the results have been found to be somewhat low. In such cases it is necessary to distil the acetone into a smaller volume of water, or better, directly into the acetone reagent. Boiling for ten minutes is sufficient to bring over all of the acetone, and the distillate need not amount to more than 100 ce. The utilization of this method in the determination of acetone and of 6-oxybutyric acid in blood and tissues appears in a subse- quent paper. 288 Determination of Acetone SUMMARY. 1. The Messinger method for acetone estimation gives correct results. 2. The Scott-Wilson method gives accurate results only when certain modifications in the original procedure are made. It is applicable to very minute quantities of acetone. 3. In distilling a very dilute acetone solution, all of the acetone may be collected in the distillate within ten minutes. NEPHELOMETRIC DETERMINATION OF MINUTE QUANTITIES OF ACETONE. By W. M. MARRIOTT. (From the Laboratory of Biological Chemistry, Washington University, St. Louis, Mo.) (Received for publication, September 4, 1913.) In order to determine very small amounts of acetone such as occur, for example, in a few cubic centimeters of normal blood, it is necessary to have a method more delicate than those at present: in use. As has been shown in the previous paper, the Scott-Wilson method is a delicate and accurate one for acetone determination, but it is not sufficiently delicate for the small amounts of acetone we wished to determine, so another method was devised. The addition of acetone to a silver-mercury-cyanide solution gives rise to an abundant white nebulous precipitate. So delicate is the reaction for acetone that 0.01 mgm. is sufficient to cause a distinct opalescence in 50-100 cc. of solution. Further, the density of the opalescence, as measured by the nephelometer, has been found, within limits, to be proportional to the amount of acetone added. The details of the procedure are as follows: .The acetone solution, which must be free from ammonia, alde- hyde or hydrogen sulphide, is distilled into an excess of the acetone reagent.! The delivery tube must dip under the surface of the 1 The reagent is made up as follows: Mercuric cyanide, 10 grams; Sodium hydroxide, 180 grams; Water, 1200cc. The solution is agitated in a flask and 400 cc. of 0.7268 per cent silver nitrate solution slowly runin. Immediately before use the reagent must be filtered through an asbestos mat, the pores of which have been partially occluded by previous filtration of a little talcum in water. At least 30 cc. of the reagent must be taken for each milligram of acetone present or expected to be present. A little experience enables one to tell by the density of the precipitate formed in the first couple of minutes’ distillate the approximate amount of acetone present. A dense precipitate may call for the addition of more reagent to the receiving flask. 289 2 290 Nephelometric Determination of Acetone liquid in the receiving flask. The distillation is continued for about. - fifteen minutes or until the distillate measures from 75 cc. to 100 ec. After standing for about half an hour, the distillate is trans- ferred to a graduated cylinder and diluted until an opalescence that can be conveniently read is obtained. The turbidity occa- sioned by 0.05 mgm. of acetone diluted to 100 cc. is a convenient strength for this purpose, although considerably smaller or larger amounts give good results. With heavy opalescence it is desirable after diluting to a certain volume, say 250 cc., to remove an aliquot portion with a pipette and dilute this appropriately. A solution containing a known amount of acetone? is distilled into an excess of reagent*® and this distillate which is to be used as the standard is diluted as above. Comparisons of the turbidity of the unknown solutions with that of the standard are made in the nephelometer of Richards.‘ The nephelometer as originally described may be improved by substituting the telescopic attachment of a Duboseq colorimeter for the eye piece instead of the plain brass tube used by Richards. A further modification consists in a partition between the two tubes. This was designed to eliminate reflections of light from one tube to the other. Owing perhaps to inaccurate construction of the instrument the same solution when read in both tubes does not necessarily give identical readings. This source of error may be eliminated by making a series of readings, then reversing the tubes and making another series of readings, averaging the two ratios thus obtained; or more simply, as suggested by Kober,® by reading the standard solution as an “unknown” and taking this value as the potential height of the standard solution. As the suspensions slowly settle out, the readings should be made as quickly as possible after filling the tubes. * A convenient stock solution contains about 0.03 mgm. acetone per ce. The strength of such a solution is determined by titration of 200 cc. by the Meassinger method. ’ *The solution cannot be added directly to the reagent as a lower result is obtained than when distilled. * Richards: Zeitschr. f. anorgan. Chem., viii, p. 269, 1895; Richards and Wells: Amer. Chem. Journ., xxxi, p. 235, 1904. * Kober: this Journal, xiii, p. 485, 1918. W. M. Marriott 291 The instrument is manipulated in a dark room, a small electric flash lamp being used to read the scale. As originally pointed out by Richards® the amounts of precipi- tate are not exactly inversely proportional to the scale -readings. Kober’ has constructed a curve of correction for use with his modi- fication of the nephelometer. When the two solutions for com- parison are of nearly the same concentration, the correction is within the limits of observational error and may be disregarded. Further, by using Kober’s equation for a correction curve it is seen that the difference between observed and corrected values _ becomes proportionately less with readings taken with greater - depths of solution. If the unknown suspension is so diluted as - to be not more than 20 per cent different from the standard and if comparisons are made with scale readings in the neighborhood of 50 mm. or 60 mm., no corrections are necessary. » In doing a series of determinations a single standard suspension _ is used and the various unknown suspensions are diluted in gradu- ated cylinders to approximately the same opalescence. Little difficulty is experienced in thus obtaining suspensions differing from the standard by not more than 10 per cent. It is to be mentioned that the nephelometer used was mechanic- ally crude with no vernier and no ratchet and pinion attachments for adjusting the sliding jackets surrounding the tubes. Greater accuracy could possibly be obtained by using a modification of the Duboseq colorimeter. However, quite satisfactory results are possible as is shown below: Solutions containing varying amounts of acetone were prepared by another member of the staff and determinations made by the writer on these solutions with the following results: Acetone added Acetone found 0.015 0.015 0.022 0.021 0.092 0.083 0.28 0.27 0.63 0.65 1.00 0.92 1.54 1.54 * a eh § = ini, ‘. ao 6 Bete THE DETERMINATION OF s-OXYBUTYRIC ACID IN BLOOD AND TISSUES. By W. M. MARRIOTT. (From the Laboratory of Biological Chemistry, Washington University, St. Louis, Mo.) (Received for publication, September 4, 1913.) The Shaffer! method for the determination of B-oxybutyric acid is applicable to blood and tissue analysis. Before applying the method, however, it is necessary to remove proteins and other disturbing substances. Proteins are removed by a modification of the Seegen? procedure of sodium acetate precipitation. The paired glucuronic acids, glucose and protein remnants are eliminated by a subsequent precipitation with basic lead acetate and ammonia. The details of the method when large amounts of defibrinated blood or of tissues are available, are as follows: A round-bottomed flask of 2 or 3 liters’ capacity provided with a dropping funnel is connected with a condenser, the delivery tube of which dips beneath the surface of a little water contained in a 500-ee. receiving flask. The large flask contains 500 ce. of water, 3.5 ec. of glacial acetic acid and a little powdered tale. The liquid is raised to the boiling point. 100 ce. of blood diluted with 400 ce. of distilled water is then run in through the dropping funnel at such a rate that boiling does not cease.’ Distillation is continued _ until about 300 cc. have distilled. A very small amount of the liquid may occasionally foam over, but this is of no consequence - on account of the subsequent redistillations. This distillate con- tains preformed acetone plus acetone from diacetic acid. Some 1 Shaffer: this Journal, v, p. 211, 1908; Shaffer and Marriott, ibid., xvi, p. 265, 1913. * Seegen: Centralbl. f. Phystol., vi, p. 604, 1893. ’ In using hashed organs the hash is all put into the dilute acetic acid before connecting up the apparatus. Care must be taken to shake the flask from time to time to prevent burning at the bottom. 293 294 Determination of 8-Oxybutyric Acid ammonia may be present, hence redistillation with addition of a little dilute sulphuric acid is performed. A second redistillation* after the addition of 20 ec. of 3 per cent hydrogen peroxide and a slight excess of alkali serves to destroy or hold back hydrogen sul- phide, aldehydes, if any, and volatile acids. The final distillate is used for acetone determination by the Messinger titration in the usual manner. The large flask is removed from the distilling apparatus, and while the contents are still hot, about 15 ec. of 20 per cent sodium carbonate solution are poured in, with stirring. When sufficient sodium carbonate has been added the dark grumous liquid changes to brown, and a flocculent precipitate settles leaving a clear straw colored supernatant liquid, of amphoteric reaction. The flask is held over a ring burner and the contents boiled for a minute or two, then allowed to cool and transferred to a graduated flask or cylinder and made up to 1000 cc. with water. The whole is thoroughly mixed and filtered through dry paper on a Biichner funnel. An aliquot portion (usually 700 ec.) is transferred to a graduated flask, 30 cc. basic lead acetate solution (U. 8. P.) and 15 ec. strong am- monia added and the volume is made up to 1000 ec. The solution is mixed, allowed to stand awhile and filtered on a dry folded filter. 900 ec. of the water-clear filtrate are. boiled to expel the greater part of ammonia and to concentrate to about 500 ce. This is cooled and sufficient dilute sulphuric acid added to precipitate the excess of lead present, the lead sulphate is filtered off, 30 cc. of 50 per cent sulphuric acid added and the whole transferred to a liter Kjeldahl flask provided with a dropping funnel. The contents of the flask are distilled and a solution of potassium bichromate or water is run in at such a rate that the liquid always retains some yellow color, and the volume remains between 400 and 500 ce. It is rarely necessary to add more than 0.5 gram of bichromate and an excess is to be avoided. Slow distillation is continued for two hours and 600 to 800 ce. of distillate collected, the precaution being taken that the tip of the delivery tube is always under the surface of water in the receiving flask. The distillate is redistilled with 20 ce. of peroxide and 5 ee. of 10 per cent sodium hydroxide, and the final distillate titrated by the Messinger method. ‘In a preceeding paper it was shown (p. 283) that ten minutes’ distilla- tion is ample to distil off all acetone. — ¢ W. M. Marriott 295 ‘To test the accuracy of the method, determinations were made on fresh defibrinated beef blood to which had been added pure synthetic B-oxybutyric acid.’ The following results were obtained: - 100 ce. blood alone: 8.4, 7.9, 8.8 mgm. oxybutyrie acid; average, 8.3 mgm. . 100 cc. blood to which had been added 82.08 mgm. of B-oxybutyric acid as determined on the pure solution: 90.9, 89.6 mgm. 6-oxybutyric acid; average found, 90.2 mgm.; amount present, 90.3 mgm. Another experiment on a different sample of blood gave the follow- ing results: 100 ce. blood alone: 7.2, 7.4 mgm. 8-oxybutyrie acid; average, 7.3 mgm. 100 cc. blood to which had been added 87.1 mgm. oxybutyrie acid: 94.3, 92.3, 93.1, 93.9 mgm. oxybutyric acid; average found, 93.4 mgm.; amount present, 94.4 mgm. 50 grams muscle hash alone, 11.3 mgm. 8-oxybutyrie acid. 50 grams of same hash to which were added 173 mgm. 8-oxybutyric acid: found, 182.1 mgm.; present, 184.3 mgm. 50 grams liver hash alone gave 16.5 mgm. 8-oxybutyrie acid. 50 grams of same hash to which were added 173 mgm. of 8-oxybutyric ‘acid: found, 186.5 mgm.; present, 189.5 mgm. Method for small amounts 6f blood. By determining the acetone in the distillates by the exceedingly delicate method of Scott-Wilson,® I have been able to make satis- factory estimations of the acetone bodies in 10-ce. samples of blood, drawn directly from a vein. The details of the method are as follows: 10 ce. of blood drawn from a superficial vein by a sterile graduated syringe are run into about 40 ec. of 0.5 per cent potassium oxalate solution. An 800-cc. Kjeldahl flask, provided with a dropping funnel is connected with a condenser, the delivery tube of which dips beneath the surface of water in a receiving flask. The Kjeldahl flask contains 100 cc. of water and 1 ce. of glacial acetic acid. This is brought to a boil and the diluted blood slowly run in through the dropping funnel. 5 In the form of the purified calcium zine double salt. 6 Scott-Wilson: Journ. of Physiol., xlii, p. 444, 1911; Marriott: this Jour- nal, preceding paper. 296 Determination of 8-Oxybutyric Acid The liquid is kept boiling for about thirty minutes, after the last of the blood has been run in. The distillate is redistilled with a little dilute sulphuric acid and again with 20 cc. peroxide and a slight excess of alkali. The final distillate is caught in small Erlen- meyer flasks containing an excess of the Scott-Wilson ‘acetone reagent.’’? The delivery tube must dip under the surface of the liquid, and it is not necessary to distil longer than ten minutes in order to get off all of the acetone. The resulting acetone-mercuric- cyanide compound is then filtered off on an asbestos mat in a Gooch crucible and acetone estimated as described in a previous paper.® This represents acetone preformed and from diacetic acid. The residue in the Kjeldahl flask is precipitated while still hot, with about 8 cc. of 10 per cent sodium carbonate, boiled a moment, filtered on a Biichner funnel and washed with hot water. To the clear filtrate are added 15 ce. of basic lead acetate (U. S. P.) and 10 cc. of strong ammonia. This precipitate is allowed to settle and then filtered off on a dry folded filter and the filtrate used for B-oxybutyric acid determination in the same way as described above for large quantities of blood, with the exception that the final distillate is caught in excess of “acetone reagent’’ and the . estimation made by the modified Scott-Wilson method, previously described. The following results were obtained. with freshly drawn dog blood. Results are expressed in terms of acetone obtained. 10 cc. of blood alone. ACETONE PREFORMED AND FROM DIACETIC ACID ACETONE FROM 8-oxyButyric AcID . om a! mgm. mgm 0.03 0.32 0.03 0.35 0.03 0.34 Average..... 0.03 | Average..... 0.33 ? See preceding paper. ® Loc. cit, W. M. Marriott 297 10 cc. of same blood to which had been added 1.74 mgm. acetone as oxybutyric acid. ACETONE PREFORMED AND FROM DIACETIC ACID ACETONE FROM OXYBUTYRIC ACID | mgm. | mgm. 2 | 6.06 2.12 0.06 2.14 0.06 | 2.16 Average..... 0.06 | Average found... 2.14 _ Amount present. . 2.04 25-gram portions of hashed muscle of a fasting phlorhizinized dog gave the following results: Messinger method... ps = lg 8-oxybutyric acid. Scott-Wilson........ oy a. ‘ 8-oxybutyri¢ acid. Nephelometric method. _ By applying the nephelometer to the determination of the ace- tone occurring as such or as diacetic acid and also to that obtained from oxidation of the B-oxybutyric acid, it is possible to make a complete analysis using only from 2 to 5 cc. of blood. . For human work, blood is drawn from a superficial arm vein by means of asterile syringe and run into about 50 ce. of 0.5 per cent potassium oxalate solution, contained in a small weighed flask. The diluted blood is run into 100 ec. of boiling water acidified with 1 cc. of glacial acetic acid® contained in an 800-ce. Kjeldahl distilling flask and the procedure carried out as described on pp. 295-6, with the exception that the amount of the precipitate in the mercury reagent is estimated by means of the nephelometer f by the method given on pp. 289-90 in a preceding paper in this ~ number. The question may arise as to whether the cnbetanee giving a precipitate in the acetone reagent is really acetone. In this con- nection it is interesting to note that the results obtained on blood by my method agree closely with those obtained by the Messinger ®Commercial varieties of acetic acid frequently contain substances which behave like acetone. Blank determinations should always be made and correction made accordingly. 298 Determination of 8-Oxybutyric Acid iodimetric titration, in which iodoform was produced and identified microscopically. It is true that ammonia, chlorides, aldehydes, and hydrogen sulphide affect the reagent, but the absence of all of these is assured by the procedures adopted. In view of the fact that the oxidation of oxybutyric acid by chromic acid gives only 90 per cent of the theoretical yield of acetone, as explained in a preceding paper, 10 per cent should be added to the results obtained by titration or by the nephelometer. A few determinations follow, the results being expressed as milli- grams of acetone per 100 grams of blood: Acetone plus dia- §8-Oxybutyric cetic acid acid Normal dog....:...¢. >. 5. ss 0.04 3.2 Normal dog.............5)), sas 0.08 1.4 Normal dog..........)... Js ase 0.06 ye Normal child.'............ 2. eee 0.06 4.4 Normal child.:........: .... 32 eee 0.08 4.4 Dog, phlorhizinized......... Reins Bie RAS ce 7.2 10.4 Child in coma.......... eee ae 23 .4 24.8 Child (following orthopedic operation). . 11.2 28 .0 The methods as given above for the estimation of oxybutyric acid have the advantage over the usual optical methods, in that very much smaller quantities of 6-oxybutyrie acid may be deter- mined with accuracy. The disturbing effect of optically active substances, such as sarcolactic acid, is eliminated. A further ad- vantage in experimental work is that the methods are suitable for estimating optically inactive B-oxybutyric acid. When sufficiently large quantities of oxybutyric acid are present to permit a deter- mination by the optical methods, a comparison of the results with those obtained by the method described in this paper shows the amounts of either dextro, levo, or inactive acid present. This work was undertaken at the suggestion of Professor Shaffer, and I am greatly indebted to him for his active interest and valu- able suggestions. STUDIES OF THE ENDOGENOUS METABOLISM OF THE PIG AS MODIFIED BY VARIOUS FACTORS. I. THE EFFECTS OF ACID AND BASIC SALTS, AND OF FREE MINERAL ACIDS ON THE ENDOGENOUS NITROGEN METABOLISM.! By E. V. McCOLLUM anp D. R. HOAGLAND. (From the Laboratory of Agricultural Chemistry of the University of Wisconsin.) (Received for publication, September 10, 1913.) That there is a fairly constant type of tissue metabolism result- ing from the necessary cellular activity which has received the name endogenous metabolism, and which is in great measure in- dependent of the nitrogen intake, is now generally accepted. That the group of metabolic end products of nitrogenous nature present in the urine of an animal whose diet contains no nitrogen shows relationships which are not found in urines under any other con- ditions is equally well established. It has become the practice, because of lack of knowledge to the contrary, to refer to endog- enous metabolism as a single variety; that is, no effort has been made to resolve this type into factors. The most conspicuous and least ‘variable known constituent of the group of endogenous end products of nitrogen metabolism is creatinine. Mendel and Rose? studying the conditions under which creatine is eliminated have shown conclusively that in fasting rabbits and dogs, or when the animals are on a diet of fat alone there is always an increase in the output of total creatinine (creatinine plus creatine). This rise is always attended by an increase in the total nitrogen output. They hold that creatinine is derived from creatine, and that those conditions which produce a carbohydrate hunger in the cells of the tissues, lead to excessive catabolism of the tissues and con- ? Published with the permission of the Director of the Wisconsin Experi- ment Station. * This Journal, x, p. 247, 1911. 299 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 2 ee -s é 300 Endogenous Metabolism of the Pig sequent liberation of creatine from the muscles. That further, the animal’s power to convert creatine into creatinine, or to destroy it may thereby be exceeded and creatine may appear in the urine. Concomitant with this increased tissue catabolism goes of course an. increased elimination of total nitrogen. In harmony with this theory are the data of Myers and Volovic* published since the work reported in this paper was carried out. Working with rabbits they found that hyperthermia, whether caused by infection or by keeping the animal in an incubator, leads to an increased elimi- nation of creatinine amounting to 36 per cent over the normal output. The view is expressed that creatinine elimination in fever still represents the normal endogenous metabolism which is here proceeding at an abnormal rate. It would seem therefore that, if an animal were placed in a con- dition in which the endogenous type of protein catabolism alone prevailed, and then by some means this type of metabolism could be increased in amount we should anticipate a rise in the creatinine elimination, or an increased creatine output or both. A further question of interest which could be answered by such experiments is the relationship among the constituents of urines carrying nitro- gen derived solely from “accelerated endogenous metabolism.” Do these relationships always remain the same when exogenous catabolism is absent? We have attempted to throw some light on these questions by employing several methods of varying the inten- sity in endogenous metabolism in pigs living on a diet of carbo- hydrate, salts and water. The methods employed in the experi- ments described in this and in the two papers following, which it was assumed would stimulate protein metabolism, were (1) the addition of hydrochlorie acid and acid salt mixtures to the starch diet; (2) the feeding of benzoic acid in the starch diet, and (8) the replacement of stareh wholly or partially by fat. In the present paper we will discuss the effects on endogenous metabolism of giving neutral, basic, and acid salt mixtures, and of free mineral acids with an adequate starch diet. Because of the ease of maintaining an animal during long periods under these experimental conditions we have deemed it best to employ the different methods of accelerating tissue catabolism on the same individuals so that all data would be better correlated and stand- * This Journal, xiv, p. 489, 1913. E. V. McCollum and D. R. Hoagland —301 ardized. For the sake of clearness the data presented in the following tables I to V will be discussed under separate titles. In view of the exceptional advantages of the pig as an experi- mental animal it seemed possible to obtain new data concerning endogenous metabolism which can scarcely be gotten with any other type of animal, because it is very easy to place this animal in a con- dition where the endogenous type of catabolism alone prevails. One of us* has called attention to the fact that a pig will eat a diet of starch, water and salts during a long period (36 days) with no sign of disturbance of appetite or loss of weight. Attention was there called to the fact that after a time, which varied somewhat in different individuals, the total nitrogen output in the urine sank to a level where the creatinine N, which remains constant, forms * about 18.5 per cent of the total. It was pointed out that this ratio could be used as an index to the maintenance requirements of the animal and would serve as a guide to enable one to feed com- parable amounts of any substance to different animals. These observations have now been extended to a considerable number of additional animals and some further comment is necessary on this point. Further experience has shown that some pigs when kept for long periods on the starch, salt and water diet never reach so high a relationship of creatinine N to total N as 18.5 to 100. Others will drop the total N excretion regularly to a point where it is about five and a half times the nitrogen eliminated as creatinine. We have seen animals which even after sixty days on a nitrogen free get had a ratio of creatinine N to total N of only 10 to 109. Whatever the ratio arrived at, it is very constant. Other pigs will have a ratio of anywhere between 10 and 18.5 to 100. In gen- eral it has been the most vigorous and healthy animals which have produced the urimes having the highest percentage of cre- _ atinine N in proportion to total N. Pigs born in the fall in Wis- consin are frequently chilled so much as to lower their vitality and give them a tendency to pneumonia. It is especially in the fall pigs that we have failed to observe the total N to fall to five and a half times the N as creatinine. The reason for this is not clear. It seems certain to be associated in general with lack of vigor. In a number of instances pigs whose breathing gave evidence of lung affections have persisted in putting out in the urine more total ‘McCollum: Amer. Journ. of Physiol., xxix, p. 210, 1911. 302 Endogenous Metabolism of the Pig N than corresponds to the creatinine. Pig No. 34 is an illustration When given starch and an alkaline salt mixture the N as creatinine was only 9.4 per cent of the total. At autopsy, both lungs were | found to be extensively affected with pneumonia. It seems to be true, however, that the ratios between creatinine N and total N under these experimental conditions is not quite the same for all individuals. Thus in the case of No. 38 during the period on an alkaline salt mixture and starch the per cent of the total N as creatinine was 14.8. At the end of the experiment the animal was killed and examined but microscopically nothing - pathological was apparent. If creatinine is a product of the metabolism of muscle tissue alone some variation in the relation between the total end products of tissue catabolism, and creatinine may be due to variation in the relative sizes of the organs as com- pared to the total amount of muscle tissue in the body. More experimental data showing such relationships in a considerable number of animals together with a careful study of the composi- tion of the urines will be necessary to throw light on this question. In an experience with more than thirty pigs it has been found that, the year through, probably three-fourths of all pigs of the sizes employed by us in experimental work, will on a liberal carbo- hydrate diet come to a stage where the creatinine N will make up about 18.5 per cent of the total, and this ratio can be employed to advantage for the calculation of the lowest level of protein metabolism of which the animal is capable. We have not seen the percentage rise above 18.5 per cent except in a single case formerly reported’ and this was possibly due to error in the analytical work. Another factor of importance in determining the ratio of creatinine N to total N after a considerable period on the starch diet is the character of the salt mixture supplied. We have repeatedly ob- served that the lowest level of total-N output is reached only when the salt mixture supplied has an excess of basic over acid radicals. This is due to the fact that when the diet is acid, ammonia is eliminated in quantities sufficient to neutralize the acids present in the diet and this ammonia N is, to a considerable extent, derived from additional protein destruction over what would take place if these acids were not present. This point will be discussed further later on. * McCollum: Loe. cit. E. V. McCollum and D. R. Hoagland 303 In all our experiments agar-agar was given in amounts sufficient to lead to regular evacuation of the intestine so the disturbing factor of the absorption of putrefaction products was kept as low as possible. There is always a regular loss of appreciable amounts of nitrogen derived from the secretions of the tract, and if these residues are not promptly eliminated, some of the nitrogen from this source will be absorbed and eliminated in the urine, changing in some degree the typical relationships between the creatinine N and total N.° EFFECTS OF ACID AND BASIC SALTS AND OF FREE MINERAL ACIDS UPON THE ENDOGENOUS NITROGEN METABOLISM. That neutrality be maintained in the blood and tissues is a fun- damental condition of life. It is therefore essential that acid* radicals, either ingested or of metabolic origin be neutralized. If acidosis obtains both the fixed cations and ammonia take part in the neutralization. With a normal nitrogen intake the ammonia of the urine under these conditions has been observed to rise while the urea-nitrogen is correspondingly decreased. What would be the result if the metabolism were of the endogenous type alone? Experimental. The animals employed in these experiments were young pigs brought to their lowest level of nitrogen elimination through several weeks of starch feeding. They were confined in special cages and the urine collected daily according to the methods of McCollum and Steenbock.’? Upon these samples daily analyses for total N, creatinine N, creatine N, ammonia N and urea N were made. Sulphur determinations were also made during two periods. In the case of the urea,N the Benedict-Gephardt method was used rather than the Folin method as a matter of convenience. The small error involved in the use of this method is of no consequence for the purpose of this investigation. In Table I are found the data for pig 34. The starch diet was begun on January 10 and analyses were commenced with the urine 6 Compare the papers of Folin and Denis: this Journal, xi, p. 167, 1912; Osborne and Mendel: Bulletin of the Carnegie Institution of Washington, No. 156, Part I, p. 39, 1911; Mendel and Fine: this Journal, xi, p. 13, 1912. 7 Research Bulletin No. 21, Wisconsin Experiment Station, 1912. ‘(ID sves yeaqneny oA rod ee: 09 bisreeaidaa - BARS a ql S See R ‘sud ZT ‘(I) 89188 OUTTeATY ‘usy sod ‘[eo og “YorRyg SEIREREE 00!) om oN Se 5 s8 bo} Looene ~ ee SELRIE _— N Ss | ZENE ea OO OO SXSRXSSLSSERBS YRVes uC — - s SSeS 85% rs ‘suu3 ¢ ‘TOeN ‘msy sed ‘eo Og ‘YoIK4g be SCOrADMDAaDAOOCNOCCsH TAMA OHH ANM KN Meta an| awn S38 NU ASOTRARTwannnnoma|/ oone SO) Se eS OS Oo Oe a | Ch Oe ee SRS8/SRESS Sorrol/s 5 ol AAR S838 SSoscoeol|SsScoocoocococoeoooscsloocsa g ARE wD _ _ Sooeoscol!ooeococoocoocoooselocooce _— Secoeeicococococooecoscosos! oocs Z Endogenous Metabolism of the Pig 304 NOILVU won oulIweIO euyUyweIO N &NILvauo 1e8 N@OOULIN TVLOL 40 INGO uad N @NINILVauO N VINONKY ‘aurun u uw hobs fo uoyngrysxq ‘OL ‘UBL poJIv4s SuIposs Youwyg “spunod g-[g “QYysIOM FE ON Fig ——~ —- . SS et et Saez . sees Ome Cocoocooceo . rhah|Sassseeseger — = ot : oS é aiva aomad AQNTHOD A0 “10A ee “TABLE I.—Continued. E. V. McCollum and D. R. Hoagland ° 395 3a a= a Ae ‘ar a -_— 2 - @ | a a 25 Meas 23 os — he L a gat SE 6 = asa 3 Ba 6 Shp a 32 4 aun. 3 ej 3 Aa@astZ eget ~~. S Pigf a @ a gb 83s Cickcich. oO = Bee Up BM 3A bo SSEESR Zz = RD DD PSM ONARS el! ONCeOCr Oe “ANON AMESSAMS A |oodecsosasag | janers Ie © aAgtISOM SRS " aie ao Hw eee PUN no HtaAtaAs Nos S@eaanrxeenanlononcenaene |) [rawtan Hr Mr HOW OWS ONG NS wt wt SER) ~oOooon 3 Srawanggnwolannwoannewte Fl nmaen SBRARSRARSR RARARASRRRE | § SSA a NOMOCHNAMS|HAMOOCANHMOM | MW | DEoHA YSIsSscssl|ceusesssass | &| ssees ENS SESERS 5 | eeeRe LEC ERBBEE) CEELEBEELEE 2 | 288% esescsocoocoisooooooosso cooce wile i i) o ° f=) > S58 8./E 82222885 8 —— a — So esoooocososo : so ao = ow Tr) S ZESSSSSES|ZSELSESESEE 2 | 8E822 esossoosoo|/secocoeoocoosse | cocco ESSSSRRER/SRS28S8RSRR | RZRSR 8 SHA SSASSASRA | RSA Sseesosossoscicosoeoosoeoooosc | coccso SSERSSSAR|/LSBSHE RISERS | SET seecooooes|soseocooseooccsce | ecocco BRAASASRRS | KSRBSARUASS REABR Rope ee) a lS ere ee id vad eek ee —e—) : Cae se g2geaseegieseesaecaee SMM ANNALRANRAR Se NO a: : ue: = eeee L Bm mB Q Saad 8 a =< CGuvde7 . et te Ht Or > mt tae = | Sete e | ae Endogenous Metabolism of the Pig 306 ‘suid QT ‘(AI) 83188 oUT[ey]y “wWsy sed “wo 06 ‘Yo1B4g “sud QT ‘(I) $3188 jeaynoN “wWdy Jed “[va OG “Youwyg NOLMLVU acre | very — = ee “< oe ° ii ea, Dale ical a catia ae idl 7 lL |Z S| 98 | 88S IZL1'0 * @c1'0 ‘0 09°0 ZO O&8t | L : L8t|Ztl| GL | 9°6¢ e120 2910 ‘0 | 89°0 | FIL | OSZT 9 PST | LL IL1'0 0 ITT | O8Z1T | ¢ 9°11 | F9L | PSL FOE SZT'0 991'0 9S1'0 | I¢ 0 101 OS6I + O81 | 6 FL | SZ sr pze'0 C8I'O ZS1'0 | 89°0 | FI | OFLI | € LSI | 8°21 | 2¢ | 8°29 9610 ¥e1'0 0/990 | GOT O88I Zz : PLL| OH | 8'e | $29 [6020 IZI'0 020°0 | $2°0 | OZ T | OFLT =I “ady LZ | 9 8t €°8 | ¥'09 6010 09T'0 |120°0 | ZO 98°O | OLFT | Te 8°91 | 681 801 Fee OLT'O ‘1610 6010 #20 | 10°1 , OF6T | 08 Il I'¢ | 6'ZI | 82 | 1'6F |190°0 . 10 (6E°0 | 62°0 | OST | OFIZ | GZ Le | 221 | ose | Z'6e |10z'0 10 (e¢'0 | 090 | £E°1I | OFLI | SZ 8°61 | TOL | 28% 6 Ih 6920 1'0 698°0 | S¢°0 | IE 1 | OZEL | 2z PLL | SSI | $82 | 91h 6970 | | 9810 Szk'O | 290 | GFT | OST | 92 cl | OIL | Leh F 08 S220 10 ‘0 | F0)| SFI OOLT S 6 LI | LZ 11 | Ze | 6 LE 6920 110 |98F'0 | Z2°0 | OS I OFST | ¥ OSI | LSI | 2:08 | £°6E on IZT'0 (90F'0 | €9°0 | GET | OFST | 8% . €°Sl | 6 IT | Gee | 3 6E ZO ‘0 0g¢'0 | Z9°0 | SCL | O8ST | Zz 8°91 | O'OT | £°9Z | 8°9F |TKO 10 9280 | 29°0 | FI | OLST | Iz ! O'S1 | 901 | SIS | 9 FF 08Z°0 ‘0 j299°0 | 6L°O | L2°T | OOST | 0G ! Z'S1|9'8 | 9°08 | L°Sh 2Es'0 TE1‘0 2960 | OL°0 | EST | O8SI | GT . ¥'1Z| 1:9 | 0'6% | O'zg [Zgg°0 CST 'O |ss¢°O | ¢2°0 | £0°% | OOFT ST / rLL19'8 | 2°6¢ FH BEeO| | 99109280 980 FT) OOST ZTAFIN I : / ‘ews | ew | emd | -ew6 | ews | “sub | “22 ° ° > a a < SC rl?telgeble@ Ela lili lee B| 3 eRe) | | = le] 3 Pew Ras: iad % Z | = La pea eee i os Me 2 N ‘TVLOL 40 INGO Had Zz | ) ) { "IS “G2q pe7s84s ‘Il W1AVL ‘aurun ur uabowju fo u01ngrysiq Zurpoey Yonyg “spunod gg “YF! ?8E “ON Fd E. V. McCollum and D. R. Hoagland ‘Sz]Bs OUTLETS “FST | O'LT [8 8 | GOI | 8°€9 \L2z°0 1° \OFL 0 | WoO ee Ts shup 9 ~AI "S}[BS OUTTEHTS “Yej-yorszg | 6ST L°6 | SOL 6'S9 622'0 OFI'O [S910 | 26°0 | HI skep 21-111 "S}][BS OUTTeALe “Yorwyg | 8°OT SH ZS | 9°69 S8I'0 I9T'O \680°0 | $9°0 | BOT | sXkep FI-II i _ "sq]Bs [eynou ‘oreyg | 2°91 | OL | S1e | & 1h $96'0 £910 |S8h'0 | 19°0 | SOT Pon sXhep §I-] % sporsad fig sabnsvay ‘hunwung PRE | TL | 9ST | 88S [09Z'O |LOL'O \0OL'O [0zz'0 | 80 | IFT | 06% | ¢ ZH |SL | ZI | $99 |61Z'O OOL’O |IIL'O 0610 | GOT | Sot | OLE | F #91 | SOF 16 | 8°S9 |FST‘O |260°0 [00T'0 980°0 09°0 | #60 | OTE | (O81 | GL | LE | 89 je2Z°0 |ZET'0 J0ZT'0 [2800 | FOI | cot | OFF |Z ) "sum3 OT ‘(II1) SH®S) | FST | O08 | 26 | 6°99 ‘0 |SIT‘O |Z01'0 \6zt'0 | 68°0 | Se 1 | Oe | T ABW auleyyy ‘wsy sod ‘peo ce “Quq | G'0Z | S°6 | FOL | 66S \6FZ'0 [SOLO [FILO \4zI'0 | €2°0 | Za T| O9E | OF AI €°91| 88 | ZS. | S19 Pee'O PLI'O 29Z'0 | £21 | 16'T | OOS | 62 O'S | ¢°8 | SIT | 0°29 lOSt’o \60°0 S80°0 SILO | Z9°0 | OO'T | OIF | 8% Z 81106 | 08 | 8'F9 [$820 /FZI'0 [SILO COLO | $80 | SBT | OTL | 22 FOL | TOL | 2°SI | 8'6¢ [91Z'O 9STO [SET'O [ISTO | 62°0 | ZEIT | OF8 | 9% 6 IL | T'1It | 9'6 | #29 \ZSt'O 9F1'O 2Z1'°0 | 68°0 | ZEIT | OL6 | SZ G'6I | 8'ZI | SOL | F'LE |S8T'O OSLO [0ZI‘O \260°0 | $2°0 | 460 | OTOL | ¥% FLL | SIL | SL | 8'S9 S8l'O LITO |IZL'0 9200 | 29°0 | GOT | 096 | & O91 | 221 | G2 | &'F9 |ISZ'0 aro ZITO | TO'T | ZE°1 | OLL | 2 Tet | 86 | FFI | I'19 0120 TO [10Z°0 | €8°0 | 681 | OOL | IZ e poye urure, uod outin { . 8°21 | 96 | et | 29 (8020 910 9€Z°0 | GOT | 89 IT | OZS | SI 88 | LOL | SFr | 2°99 OST'O ZLI'O 8¥2'0| FIT | ILI | 098 | ZT ¢7|L24 |os | L219 eO 6IT'O \2Z1'0 | 6°0 | FET | 066 | OT OST | #8 | GSI | 8'E9 199z'0 1°O [1220 | 60° | ILE | 026 | ST “sui3 QT ‘(IIT) szqesouyey «6| FOL) SS | FO | 889 PIE O €91'0 €21'0 | cel | 261 | OLE | FI -Ty ‘“wsy 10d “peo e, “4ej-qorwig | FLT | 9°01 | OL | 0'S9 6120 eet 0 "0 | 280 | 9%'T | 026 | ST Ill 9°F1 scl 0 80° | OO6T | ZI . Tel | *2r|99 | 629 6EI'0 1g1'0 0 | ZL°O | 90°T | OFLT | IT mesial | 89 |LST'0 Itl'0 0| $40 fs i 308 Endogenous Metabolism of the Pig of February 24. In the first period of four days an ample starch ration with sodium chloride was given. Since no alkaline salts were fed it was necessary for the animal to neutralize the metabolic sulphuric and phosphoric acids by alkali produced within the body. The ammonia elimination is consequently high, being about 19 per cent of the total N excreted, which averaged 1.23 grams per day. (A discussion of the creatinine metabolism will be reserved until later.) There follows a period of eleven days in which a simi- lar starch ration is fed, but with a salt mixture which was of alka-— line character (salt mixture I, p. 315). On the first day there was a slight increase in nitrogen elimination, followed immediately by a decrease which continued at the lower level throughout the period. The ammonia N sank to about one-third that of the previous period. It is to be especially noted that there was a distinct drop in the total N. A period of four days followed, in which the alkaline salt mixture was exchanged for an approximately neutral one (salt mixture II). The total N and ammonia N again increased. While the changes are small in themselves they form a large per cent of the total. It is believed that deductions may legitimately be made from such variations, since the experimental error which is of the magnitude 0.01 to 0.02 gram nitrogen, is small compared with the amounts of nitrogen eliminated daily. No exogenous nitrogen was present, and fluctuations have a significance as indicating actual quantitative changes in the metabolic processes. The experimental work was begun upon pig No. 38 after the animal had received the starch diet for twenty-four days. During the first period of thirteen days a nearly neutral salt mixture was fed (salt mixture IT). Under these conditions the ammonia produc- tion was high, averaging 0.488 gram, while the total N gave a daily average of 1.55 grams. During the succeeding period of two weeks, the salt content of the ration was changed to one of markedly basic character (salt mixture III). As is shown in Table IT the average daily elimination of total N dropped from 1.55 grams to 1.09 grams, the ammonia from 0.488 gram to 0.089 gram, while the urea N and creatinine N remained constant. The very close agreement of the results for urea N in these two periods may be due to coincidence, but there may fairly be deduced from the aver- ages the conclusion that an additional amount of protein has been catabolized in response to the acid character of the diet. One is E. V. McCollum and D. R. Hoagland 309 impressed with the fact that the organism was apparently not able to utilize the nitrogen of the urea fraction to neutralize the acidity and thus prevent an increased nitrogen elimination. There was no change in the creatinine output and so in accord- ance with the present conception of protein metabolism the ad- ditional: protein destruction must have been derived from other sources than muscle tissue. To demonstrate this point more conclusively one animal, No. 39, was fed 10 ce. of hydrochloric acid (1:4) each day during a period of five days (Table IV). The total N increased from an average of 2.86 grams to 4.03 grams, while the creatinine N gave an average in one period of 0.437 gram and in the other 0.424 gram. The probability that this “extra nitrogen” was derived from some tissues other than muscle is further supported by the observa- tions upon the neutral sulphur in the urine of pig 38 during the different periods (Table III). The neutral sulphur remained con- TABLE III. Pig 38. Neutral sulphur in urine. . -PERIOD pace | oe) One | iat ae RATION ce, gms. gms. I Mar. 19 1580 1.53 0.024 | Starch. 20 1800 1°77 0.022 | Neutral salts. 21 1570 1.43 0.028 23 1640 1.35 0.024 24 1840 1.50 0.024 25 1700 1.45 0.026 26 1520 1.49 0.029 27 1320 1.31 0.034 II Apr. 1 | 1740 1.20 0.016 | Starch, alkaline salts. 2) 1880 | 1.05 | 0.019 3 1780 1.24 0.021 4 1950 1.01 0.023 5 —-1780 1.11 0.030 6 1780 1.14 | 0.033 | 7 1830 1.02 0.035 | Summary. Averages by periods. . Endogenous Metabolism of the Pig 310 64% 091} ¢°L | OIE 969°0 rr 0 ot 9 FI | OF S| COST | SZ 0'0€ | O'L1 | HL | 9'Sh S8L'0 SHO L610 | GI'l 19°% | OzoE | % . 6°96 | $81 | 9'L | I°L¥ 069°0 9270 FEL'0| IZ 1 | 2¢°% OOLZ | & ) “A]Vep plow otozuoq “suid F-+-uoIyws ouI’E | ZZ | BLT | LS | HIG ZS9'0 Eze 0 Se2°0 | IS I | 6's 099% | zz Il £6 | 02/29 | 0°29 SIZ'0 H0S'0 ecto | GI | 62'S | OFZ | Iz G91 | 0°61 -8'F | 6S I8E'0 SFO TIT'0 | SET | Ie's oszioc | |Z 6r | 6g 820 2910 | $8°S | OL6Z | 6I SIL | €8l| FE | S29 S980 S690 O1L'O | 61'S 92'S | OZ | SI ‘OOON ‘su O+(AT) St18s ‘smB OZ O'ZI FEZ! 0G | ZS 2920 ZEF'O TEI“O | ZIT | ZZ | OOFE | ZI | rerepet an Oo 8°91 | 1'2S| 9 | OSS GEO FIS'O FIO | SVT | SES | OOS | St 102 | 0°61 | 29 | 88 |IIS'O Z8F'0 Ucl0 | GEL HS OLzE! FH G11} #81 | 99 | OLE SOF'O Fr'O ecto | E81 | Ig's | O@wzZ EI 6°02 | FFL | LSI | 1S PI9'O Zer'0 FEO | SST | HOS | OFEE | ZI ‘suid Gg ‘(AI) =| 6°LZT-| SLT | 8 LT | 9 LH OPO ZZb'0 Oze'O «IS 1 c2% | OOSe | TT S4[8S oUTTBY[Y “wWisy Jed “;wo gy “YouRIG | 62S O'ST | L'8 | 20S ie 0 | 81 | 92 | Ooze | OL “Ady |-T 2 “sub *su6 | -swu6 | “smd | ems “29 > a yw r= | < i] PIPL ELFI ELE ELE SES Te B | 8 Soe) Ble] 2 | 8 | - NOWVU 5 = % 2 = Bt | ) | @ ) | | a Oa | . N IVLOL 40 INGO Ud ‘qurin ur uabosjru fo uo1ngrysiq, "] Youvyy pojieys Zurpoay youeyg “spunod ZOT “FysIOA 268 Bd “AI ATAVL o ‘N plow oyanddyy sepnyouy syq.y, , oe ‘SUI3 91 “OH ‘Plow o1ozuaq “youwyg | + OF | OL | 8 gE | PEL E91 FeO LT leo oF, skup ¢ ~A ‘Suu QT “plow olozueq “940 “Yorwyg | 12g |S St | WET ZI Z6r'I wer 0 SEO | ss°0- ae ad Bare shup ¢ -AI ‘SWF QT ‘plow ofozueq “949 ‘Young | ZF) 212/66 0°92 SHO FSF'0 Oez0 S60 Gazal skep 2 -I1I ‘SWS F ‘plow dfozuoq “940 ‘Yorngg | 69S | FLT SL | 1°6F 189'O 9EF'O 0S'O Gat es) Shep ¥ ~II “syyes ourpeyye “yoreyg | 9°91 1.1 °6T | 0-8 | 6:99 FeF'O 88F'0 jo0z'0 lert| oer shep ZI-I *sporsad fig sobv.0ay ‘fupwmuny : | | a) q — on 13°) © me . | ;. PZE 66 | OSh SH ze l OPO OSL 1 | 8o°0 6OF OSI ZI 4 se sor | 91 TOT pees ‘0 082°T | £0 82h 00 | IT . OIF OIL FTE O'OT FID'T 9Zh'O 0ZZ'l | 290 88° | OSS | OI = ‘Ajrep ‘suiZ g] ‘prow Siozueq ‘09 = | Zh SOL OF ZE. FGI OSL I OPO 0821 GF'0 | 96°E | 099% 6 . a Ot ‘(F: 1) IOH *suZ OF ‘(II) S3Tes “Yoreyg | 8h TIT | 8°86. FFI ZS ‘0 OFT | LO 96E OF 8 A S : ee } / a> - Fie LEI GFL OST WORT HE'O SEO 08-0 00" 0c8e | 2 = F¥S | SIT Sk | 60% Z6L'T 0880 8EF'0 | 690 OF'E Oz 9 = OTS ILL | OFT | OST ee 0 ZOr'O | Z6°0 | 68'S (0964 | ¢ oe (36h L991 | OST | SEL ZIP T SZF°0 OFF'0 Gg'0 | 28°% | 0662 | F Ee a ATR PIVe O1osueg “suid OTE WON es Wate FIs | SOL FOL) Gor ete T 06F'0 $9Z°0 | GF'O | 99'S | OLES | E AI a —— : Por | aes) a2 | $22 LOT [9LE°0 L9T'0 | Ze°0 le °G | 0862 z > oop | 11z| ZOOL | €'ez [6160 oer’ 0 GOZO 2F0 | O'S OFS | I AEN x) Zeb | 8% | LIL | 8'1z |IS6'O |Fes"0 0 | 8F'0 | 02% 0908 | OF (9 Fe | 0'9S EL. | G'TE 98E°O OFF'D FZI'0 | FO 69'T O8GZ | 64 FOF | 2°81 | 86 | Ss ELT 0940 17G'0 | 9'0 9F'S | OFEE | 8Z e ‘A[wep PH’ olozueq ‘swMBQT+uUONeIOMIEg 98g | LAT | PS PIE I |82F'0 9€8'0 | S80 | 12'S | O€ZE | Lz : (6% 611 | $6 bet bret 0 sIs"0 €0'€ Ws Fee 9¢ tdy III 312 Endogenous Metabolism of the Pig stant whether the diet contained a neutral or an alkaline salt mixture. It might be suggested that the extra nitrogen catabolized with the acid ration had its origin in the liver and that no extra muscle protein was decomposed, which would be in harmony with the constancy of creatinine elimination in the different periods. Hedin’ has shown that outside the body an acid reaction causes increased proteolysis in liver tissue. Arinkin® has obtained similar results. If analogous conditions obtain in the functioning organ the in- creased nitrogen with constant creatinine elimination would be explained. In this connection it may be recalled that in certain pathological conditions, as in phosphorus poisoning, with its ac- companying degeneration of the liver, an increased nitrogen elimi- nation takes place without increased creatinine.!° Schryver," as a result of his studies upon the autolysis of organs, has advanced the hypothesis that the stability of the liver is the result of the mass action of three sets of bodies, the tissues, the metabolites or bodies derived thereform, and the autolytic enzymes. An acid reaction causes increased activity of the autolytic enzymes, with the degra- dation of protein. The resultant amino-acids are the sources of the ammonia necessary to restore neutrality. In cases of starva- tion or of low nitrogen intake these proteolytic enzymes serve auto- matically to adjust the destruction of tissues to the requirements of acid neutralization. If this conception be correct we should reason that a nitrogen-free ration of sufficiently alkaline reaction would reduce the endogenous output to a level lower than could be ob- tained by any other combination of food ingredients. Conversely we should assume that the feeding of a non-oxidizable acid would accelerate the proteolysis. Pathologically, any conditions which would increase the acid formation within the body should also increase the liver autolysis and consequently the amount of nitrogen eliminated. Schryver has cited the examples of insufficient oxygen intake and of phosphorus poisoning. In each instance the dimin- ished oxidation results in the accumulation of intermediate prod- * Festschrift fir Hammarsten, 1906. 9 Zeitschr. f. physiol. Chem., liii, p. 192, 1907. 1° Mendel and Rose: this Journal, x, pp. 213-264, 1911, give a full discus- sion of the literature on this subject. " Biochem. Journ., i, p. 123, 1906. SE Se ae ee ee - Distribution of nitrogen in the urine. E. V. McCollum and D. R. Hoagland 313 he an +: 2 z 23 |i | mi a am |, Vee $2 rs g@ jo™ | os ] St 6 ery : S% 3 gs [ES | ta = ae leg 3% = g aa | 38 = 71) ee « e loa = aE Reet = m= he ~ ~ i on Pie a ee hE 60g 35 —_ - - aN & o =e Fr oe ee — i | x ° ao 2s SESINE 2 | | aBag =~ Ole 2 | } = aaro 4 4 SIS Ss aa & to a “) sezieee | | 2052 = e gaz on | , O10 —4 et CF gle &™ | 28t+ &-- > Sis ak eaee ou ios 44 mi 2 om as Zz < t gu | POROMON| BIA AQNOSCA\OWMIENOadt || Cnen P Cee) LSet Seas MSetNeSSie™! ane > a - fe Biss A CO HOD SHH OAD HOD OR HOOT HMIOrUINOCHON 3 ImMeoowo erates oo eee ee | eroae memes | © eee SaISISaGBSSSSSSSRABSSRaVs |<“) SSS naum | SEESRRSERSSR SS SS eSseseRs | P| Reas “ammaGND) Sscoooccloccoosococolssclossco : thi) ID $H O9 — as x7) era AU ERE) FEREEE BREE SEER RTE | SNINUIVEW) SsSscoocclosceoscoSoooScl|lssslosccsco esoco N vinonny| SSRRKSSSIRRSARAA AAS ASSAS RAZA Ssescsesescsloscoooscsocolsssiloscsco soso nona| SRSSRBERSBSessaassReaaaa| | axes | SSSSSSSlOHOHOOCSOHOlSSSlooscsos | cose | Peery n won| SSSSSRRSRISTHIRIGRABSSSSS | | SSXs numa, | SERRSERRSRESTSSSSETRSTSE | 40 ENOTOA) SNS wate SBS2SSE 522 2aANR | b.. -ai oy: ae } ° - . e LEAANRAIRRNRRSSTAP TS SRees EE: adiva| o | onmm 2 z : eae | (2288 | | | y. b> > ae | coral = i= ios | piot=l j ; _ 314 Endogenous Metabolism of the Pig ucts of combustion (e.g., lactic acid), and it has been demonstrated experimentally that there is also increased nitrogen catabolism. In pathological cases where the liver becomes strongly acid its degen- eration is rapid. The experimental data reported in this paper are in harmony with this hypothesis. The constancy of the creatinine N as con- trasted with the marked changes in the total N finds its explana- tion in the sources of the nitrogen fractions. According to the present theories of protein metabolism, practically all of the creatinine N originates from the muscle tissue, but an acid accel- ‘eration of catabolism does not materially affect the creatinine output. It seems probable that the extra nitrogen eliminated under the influence of acid is derived from the liver. If this is the case we should not expect an increase in creatinine output to follow even a marked increase in the total N elimination m acidosis. The experimental data available indicate that the endogenous metabo- lism of certain tissues can be selectively accelerated by the intro- duction of acid salts and of hydrochloric acid into the diet. Two possible explanations seem available for the great excess of nitrogen eliminated by pigs under the influence of acids. It may be assumed that the animal cannot use the nitrogen, which would appear as urea if the diet contained alkaline salts in excess, for the production of ammonia necessary to maintain neutrality in the body. It is also possible that the nitrogen of the urea fraction is utilized in the first instance to produce ammonia and that the nitrogen catabolism of the tissues is stimulated by the presence of the ammonium salts thus formed. Our data do not afford an answer as to the correctness of either of these views. Since these experiments were carried out Underhill” has published results which seem to indicate that ammonium salts do stimulate nitrog- enous metabolism. There was a large exogenous factor in his experiments and the available data do not seem to warrant a con- clusion as to the correctness of the view when the endogenous type of metabolism alone prevails. ® Underhill: this Journal, xv, p. 327, 1913. E. V. McCollum and D. R. Hoagland 315 SUMMARY OF CONCLUSIONS. 1. Data are presented which show that the endogenous metabo- lism of the pig reaches its lowest level when the animal has an abundant supply of carbohydrates together ‘with a salt mixture of an alkaline character. 2. The total output of nitrogen derived from endogenous sources can be greatly increased without changing the output of creatinine. 3. The additional nitrogen which is eliminated on an acid over what appears on an alkaline diet is in the form of ammonia. The animal is not able to use the nitrogen of the urea fraction to neutralize the acids present in the diet, but draws additional nitro- _ gen from the tissues for ammonia production. Composition of salt mixtures. Salt mixture I. _ Fe,0; added. Salt mixture IT. per cent per cent de as... 5 wea eee 0.8 Ca lactate...:ssc sedans... sae 17.4 SE | 13.2 MgSO, (anhydrous).......... 20.7 SEEMS Pigs cs. fs eames 22.3 K,HPOg) #239 ee cs ee 48 .6 } CaHy4(PO,)s A srr 37.0 NaCl 0 0 a kip ahels.6'\« a Res «2 ¢ 59.8 2.8 MgSO, (anhydrous)........... 2.4 NaSO, (anhydrous).......... 10.4 TS INPMEO is iocs....- +s alee 17.7 Fe.O; added in small amount. ES Ok... 6.6 Salt mixture III ps Sait mixture FY. ee ee 10.0 a — alt mixtuteelc.:..|" ||") =) = jl -— |= +@ Chloroform ext. of Bees acid aol 2..:.:. 9. pink |—|—|—|— | = —- |-| —- -|-| =) = Ether extract of | acid sol......... pink + —| | - -| - je} + Petrol. ether ext. | oe | of ammon. sol... — | 8 —| =i eel toed ae jaca aie ate +@ Benzene extract of / * ammoniacal sol.| — = (——|—|—| — - j-|,- |- - - Chloroform ext. of | ammoniacal sol.) — ||| ee | — Je Se hele - Amy! alcohol ext. | me of ammon. sol... — --—- —| —| —| —|brown| yel- |—| pink |—| —| vio- | gre ) | low let i 0 ‘puiphate added | Me eta aaa q Petroleum ether, | | | . ext. of acid sol... " | a =) ae = - - j-;-| - - Benzene extract of | | | | 4 acid solution... — | -|-| | = — |i eg + Chloroform ext. of | | | | | : acid solution .... pink — — FY —-|—fe] —ol — |=) Se) ee Ether extract of | - acid solution . | pink |~ -|-||—| — ee) Se ae | lee + Petrol. ether ext, | of ammon. sol... = ae -|= - —- |= - -|-l- +; Benzene ext. of, _ J | | . ammon, sol.....Ja [—|-ii—| —| 1 | | - i} - -}-]) - ) = Chloroform ext. of |] | ' ammoniacal sol.) + +++ + +| +/+ | oa + Amy! alcohol ext. | ) of ammon, sol... + tite i+, +) i+ —_———— oe * These resulte also a the absolute worthlesibess bad uf the Prussian blue test. + et | ant + 4 BACTERIAL AND ENZYMIC CHANGES IN MILK AND CREAM AT 0°C.1 By M. E. PENNINGTON, J. S. HEPBURN, E. Q. Sr. JOHN, E. WIT- . MER, M. O. STAFFORD anp J. I. BURRELL. (From the Food Research Laboratory, Bureau of Chemistry, United States Department of Agriculture.) (Received for publication, October 7, 1913.) INTRODUCTION. 7 id Previous work? in this laboratory has den I that raw milk, held at or a little below a temperature « C., undergoes a marked proteolysis, which is pronounced at the end of two weeks. One function of the present research was to determine what part of this proteolysis must be ascribed to the native enzymes of the milk and what part to the bacterial flora, and, finally, to determine the combined action of these two agents. The action of the milk 1 This paper was presented at the Third International Congress of Re- frigeration, Chicago, 1918. It is the report of work done during the winter of 1908-09, when the laboratory was confronted with the task of studying the effect of low temperatures on flesh foods. There were no precedents or methods to guide us, and the apprehensions of the public demanded that results be obtained promptly lest the public health suffer. We, therefore, made a series of observations on the chemical and bacteriological changes occurring in milk and cream, to determine as quickly as possible the gen- eral trend of the decomposition at low temperatures, and conducted our work on flesh changes accordingly. Part of the preliminary work has al- ready been published in this Journal; the remainder constitutes the present communication, issued in the hope that the fundamental facts which devel- oped as the work progressed, may be of service to others as they were to us. It would have been impossible to execute such a study promptly had there not been unusually efficient team work on'the part of the investigators, and an intelligent cross interest of chemists and bacteriologists in the entire _- scope of the work. Dr. J. S. Hepburn, in addition to his share of the chem- ical analyses, has correlated and presented the data, and to him the thanks of all the authors are due.—M. E. PENNINGTON. * This Journal, iv, p. 353, 1908. hax T im, 33 r f : 5 i rae THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3. 332 Changes in Milk and Cream at 0°C. enzymes took place in formolized raw milk, of the bacterial flora in reinfected sterile milk, and of the two agents in combination in raw untreated milk. Sterile milk was also studied as a control. Since in the previous studies, it was shown that the acidity of raw milk increases* and that the lactose content decreases,‘ when raw milk is held at 0°C., it was considered advisable to include within the scope of the present investigation, determinations of acidity and lactose, and to line up the chemical changes with the changes in the freezing point. At the same time, a similar set of experiments was carried out on cream, the chemical analysis including the various fat constants, the lecithin and the freezing point. Certain zymo- 1e research cited showed that bacterial growth to a high degree had taken place in the milk during its progress,> so the number, groups and species of organisms, present in the various samples of milk and er e dete ed. The bacterial work possessed an added interest, uch as it rendered possible a comparison of the relative rate of growth of ‘the milk organisms in reinfected sterilized clean milk (or cream) and raw untreated clean milk (or cream) at a temperature of 0°C. ic investigations were made on both the milk and the cream. METHODS. Preparation of the samples. The raw milk was fresh clean milk from a high-grade dairy and was strictly comparable with the milk certified by medical milk commissions. The portion for the study of sterile and reinfected sterile milk was received at the laboratory two days in advance of the remainder of the sample, and was sterilized by heating inan Arnold steam sterilizer for thirty minutes on each of three successive days. The raw untreated milk was stored without treatment of any kind. The formolized milk was prepared by adding suf- ficient formaldehyde to raw clean milk to make 0.1 per cent. This is the quantity of formaldehyde that was used by Tice and Sherman® in their study of proteolysis of milk at the temperature of the laboratory. The reinfected milk was prepared by the method of St. John and Penning- ton.’ The organisms were precipiusred froma pore of the raw milk by * Pennington: loc. cit. * Hepburn: Journal of the Franklin Institute, clxxii, p. 187, 1911. * Pennington: loc. cit. * Tice and Sherman: Journ. of the Amer, Chem, Soe., xxviii, p. 189, 1906. 78t. John and Pennington: Journ. of Inf. Dis., iv, p. 647, 1907. Pennington and Collaborators 333 centrifugalization at high speed in sterile glass tubes, which have the gen- eral shape of tubes for the centrifugal collection of urinary sediment, but have a capacity of 250cc. The centrifuge carried eight such tubes and was run at a velocity of 3,000 revolutions per minute. A picture of the appara- tus has been published. The supernatant milk was removed by means of sterile pipettes; sterile physiological salt solution was added and mixed with the bacterial sediment; then the tubes were returned to the centrifuge and whirled. The supernatant solution was removed from the bacterial sed- iment with sterile pipettes, and the entire procedure was repeated several times in order to obtain organisms as free as possible from milk serum. The organisms were then sown in a portion of the sterile milk at the tempera- ture of the room; care was taken that the total count per cubic centimeter of the reinfected milk should be approximately the same as the total count per cc. of the raw untreated milk. Each of the four samples—raw untreated, formolized raw, reinfeeted sterile and sterilized—was kept in a sterile flask in a mechanically refriger- ated chill-room at 0°C. At intervals of one week, the contents of each flask were mixed intimately by thorough shaking, and a sample was withdrawn by means of a large sterile pipette, placed im a steri k and submitted to analysis—bacterial, zymochemical and chemical. The sterile sample was analyzed at the b ing and at the end of the experiment. Its ster- ility was demonstrate ‘at both times by plating. Needless to remark, on the first analysis of the fresh samples, one analysis sufficed for both the raw untreated and the formolized raw milk, and one analysis for both the sterilized and the reinfected sterile milk. The source, preparation and sampling of the cream were the same as in the case of the milk. The milk was held at 0°C. for a maximum period of thirty-five days, the cream for a maximum period of twenty-eight days. Bacterial and zymochemical studies were made on both milk and cream. For the chemical analysis, the milk served for a study of changes in freez- ing point, lactose, acidity and distribution of the nitrogen, while the cream served for astudy of the freezing point and of the lipins, including the de- termination of the various fat constants and of the quantity of lecithin. Chemical technique. The total nitrogen, casein nitrogen, and albumin and syntonin nitrogen, were determined by the method of Van Slyke and Hart® as modified in this laboratory.!° The caseose nitrogen was determined in the filtrate from the 8’ Hepburn: Journal of the Franklin Institute, clxxi, p. 595, 1911. ® Van Slyke and Hart: New York Agricultural Experiment Station, Gen- eva. Bulletin 215, 1902, p. 101. 10 Pennington: This Journal, iv, p. 360, 1908; Hepburn: Journal of the Franklin Institute, clxxii, p. 390, 1911. 334 Changes in Milk and Cream at 0°C. albumin and syntonin by the method of Bémer" for proteose nitrogen. The amino-acid nitrogen was determined on a separate portion of the milk by the method of Bigelow and Cook.” In all cases the actual determination of the nitrogen in its various forms was by the Gunning method. The determination of total nitrogen including nitrate nitrogen, was car- ried out according to the modified Gunning method" of the Association of Official Agricultural Chemists, using salicylic acid and sodium thiosulphate. The free ammoniacal nitrogen was determined according to Berg and Sherman." In the tabulated results (Table I) the data included as undetermined nitrogen may be taken as a measure of peptone nitrogen. The percentage of each form of nitrogen in terms of the milk is given in the tables in Roman type, while each form is also expressed as per cent of total nitrogen of the milk by means of italics. The latter mode of ex- m of nitrogen results is of great help in the study of proteolysis. ——— ty was determined by titrating 10 cc. of milk at room temperature with +5 sodium hydroxide, using phenolphthalein as the indicator. The a8 Cc. of 4 to sodium — required by 100 ce. method, after clarification by The lecithin was extracted 4 as direoted by Nerking aa aensel!® and was burned by means of sodium peroxide, as described by Le Clere and Dubois"? and by Dubois.'* The lecithin phosphoric anhydride was then de- termined volumetrically by solution of the ammonium phosphomolybdate_ in a known volume of standard alkali, and titration of the excess of alkali. — The per cent of lecithin was caleulated by multiplying the per cent of leci- — thin phosphoric anhydride by the factor 11.41. The fat for the determination of fat constants was extracted from the cream by the method used in this laboratory for the extraction of fat from egg yolk.*® The cream was mixed with several times its volume of 95 per cent alcohol, and the precipitate was collected ona filter and dried over cal- 1 Bémer: Zeitschr. f. anal. Chem., xxxiv, p. 562, 1895. 2 Bigelow and Cook: Journ. of the Amer. Chem. Soc., xxviii, p. 1485, 1906. “VU, 8. Department of Agriculture, Bureau of Chemistry, Bulletin 107, revised, p. 8. Berg and Sherman: Journ. of the Amer. Chem. Soc., xxvii, p. 124, 1905. '" U. 8. Department of Agriculture, Bureau of Chemistry, Bulletin 107, revised, p. 118, ' Nerking and Haensel: Biochem. Zeitschr., xiii, p. 348, 1908. ? Le Clere and Dubois: Journ, of the Amer. Chem. Soc., xxvi, p. 1108, 1904. Dubois: ibid., xxvii, p. 729, 1905. '® Pennington: this Journal, vii, p. 115, 1910. Pennington and Collaborators 335 cium chloride in a desiccator. The filtrate was evaporated to dryness in vacuo, using a water bath as a source of heat; and the residue was combined with the precipitate, then extracted for two days in a Soxhlet extractor with freshly distilled petroleum ether of boiling point 40-60°C. The sol- vent was removed by distillation on the water bath, and the residue of but- ter fat was used for the study of the fat constants, which were determined by the methods of the Association of Official Agricultural Chemists.2° The procedure of Hanus was used for the determination of the iodine number, while that of Leffmann and Beam served for the determination of the Reich- ert-Meiss| number. The index of refraction was taken on an Abbe refrac- - tometer which was provided with a water jacket. The ester value was determined by difference by subtracting the acid value from the saponi- fication number. For the determination of the freezing point, the apparatus of Beckmann was used. The fixed point of the Beckmann thermometer was ie and determined by means of dimethylaniline. Zymochemical methods. | For the detection of catalase 5 ec. of . - creat 0.3 cc. of M, hy- drogen peroxide were mixed in a sterile Erlenmeyer flask of 50 cc. capacity, provided will a ster elivery tube which dipped beneath a eudiometer ums atic troug The oxygen evolved was thus collected over water ; upward, wet displacement. The period of incubation was forty-eight hours at 37.5°C. The volume of oxygen was measured at room temperature, sh was noted as well as the barometric pressure. The difference of 4 Novel of the water within and without the eudiometer was disregarded. The volume of oxygen has been reduced to a temperature of 0°C. and to a pressure of 760 mm. of mercury for insertion in the tabulated results. In the study of reductases, the reagents—methylene blue and methylene- blue-formaldehyde—were prepared and used as directed by Schardinger.*! No attempt was made to obtain anaerobic conditions; the period of incu- bation was from twelve to twenty-four hours at 37.5°C. A bleaching of the lower portion of the solution during that period of time was considered proof of the presence of reductase, even though a blue ring remained on the surface of the substratum. The oxidase reagents were tincture of guaiac U.S.P.2? and trikresol—a 3 per cent aqueous solution. Ten cubic centimeters of milk and 1 cc. of the reagent were mixed and incubated at 37.5°C. for twelve to twenty-four 20U. §. Dept. of Agric., Bureau of Chemistry, Bulletin 107, revised, pp. 131-142. 21 Schardinger: Zeitschr. f. Untersuchung der Nahrungs und Genussmit- tel, v, p. 1113, 1902. 22 Pharmacopoeia of the United States of America, 8th Decennial Revis- - jon, p. 467. 336 Changes in Milk and Cream at 0°C. hours. The production of a blue color by the guaiac and of a violet or pur- ple coior by the trikresol during the period of incubation was considered to indicate the presence of an oxidase. In the tabulated results, the presence of either of the reductases or of either of the oxidases is designated by a plus sign (+-) in the proper column, while the absence of the enzyme is recorded by a minus sign (—) in the proper column. Bacteriological technique. To determine the total number of organisms per cc., plain nutrient agar was sown with the milk or cream. To determine the number of acid-formers present, plates of litmus lac- tose agar were sown and a count was made of all acid-forming colonies. ° determine the number of anaerobes and facultatives present, plates nutrient agar were sown according to Wright’s anaerobic plate ia In Sexy case the dilutions were carried high enough to insure between one hundred and two hundred organisms on each plate, and duplicate plates were always made of every dilutior The plates were incubated at 37' s for two days, 20°C. for five days, and 0°C. for four weeks. A Stewart cou x chamber and a 1} inch lens were used for counting; the results were Biecordad as directed by the Amer- ican Public Health Association.** bd To estimate the number of liquefying organisms present, sowings were made in nutrient gelatin; the plates were incubated at 20°C. for five days and the number of organisms liquefying the medium were counted and the — results recorded as stated above. ae Isolation and identification of organisms. For the isolation, identification and comparative rate of growth of the organisms in the milk and the cream, plates were selected where the colo- nies had sufficient space for free development. All the colonies were counted, and those with like cultural characteristics were grouped and studied as to their morphology and relative rate of growth and appearance of the agar streak, when incubated at 37°C. (incubator), 22°C. (room), and 20°C, (refrigerator), respectively. The optimum temperature for each organism isolated was used for the further study of its morphology and bio- chemical characteristics. Characteristic ‘growth on plates of plain nutrient agar, litmus lactose agar, and nutrient gelatin, respectively, were studied, also growth in stab culture on nutrient agar and nutrient gelatin. The characteristic growth and reaction produced by the organisms in neutral * bouillon and in litmus milk were noted, as were indol production, reduction * American Public Health Association: Standard Methods for the Exami- nation of Water and Sewage, 2nd edition, 1912, p. 79. Pennington and Collaborators 337 of nitrate to nitrite, digestion of casein and of gelatin, chromogenicity and fluorescence. The aerobic, anaerobic and facultative properties were stud- ied as well as morphology, motility, arrangement of flagella, spore-forma- tion and reaction toward the Gram stain. THE CHEMICAL CHANGES IN THE MILK DURING HOLDING. The distribution of the nitrogen. On the first analysis of the reinfected sterile and the sterilized milks, it was seen that the heat during sterilization had given rise to a partial coagulation of the albumin; this coagulated albumin precipitated with the casein in the analytic separation of the two proteins, hence the casein nitrogen was high and the albumin and syntonin nitrogen low for a fresh milk. The milk had also been slightly concentrated by sterilization as shown by the total nitrogen which was slightly higher in the sterile reinfected : milks than in the raw untreated and the formolized raw milks. In the raw, untreated milk, the casein nitrogen underwent a progressive decrease; the albumin and syntonin nitrogen was almost the same at the beginning and end of the experiment, its fluctuations during the intermediate analyses being doubtless due to formation and decomposition of metaprotein at the expense of the casein. The digestion of the casein must account for the large increase in caseose nitrogen during the latter half of the period of holding. That the caseoses may undergo digestion is shown by the progressive decrease in that form of nitrogen during the first half of the experiment. The amino-acid nitrogen showed a marked tendency to increase at the expense of the protein nitrogen. Dur- ing the proteolysis, peptones were doubtless formed as is shown by the variations in the data listed as ‘undetermined nitrogen,’* which is a measure of peptone nitrogen by difference. The quantity of _ free ammoniacal nitrogen was but slight and fluctuated wildly. _ Nitrates were not formed during holding, for, at the end of the experiment, the total nitrogen and total nitrogen including ni- trates were practically the same, hence, nitrogen fixers were absent from the milk. Ravenel, Hastings and Hammer” analyzed a clean milk and a _ fair grade of commercial milk which had been held at low tempera- *4 Ravenel, Hastings and Hammer: Journ. of Inf. Dis., vii, p. 38, 1910. 338 Changes in Milk and Cream at 0°C. - tures for a period of 203 days. In addition to the total nitrogen, they record the “water-soluble” nitrogen which was obtained by diluting the milk with water, adding “a small amount” of acetic acid at the temperature of the water bath, filtering and determining the nitrogen content of the filtrate. During holding, the soluble nitrogen—expressed as per cent of the total nitrogen—became higher than in fresh milk, being 17.97 in the clean milk and 22.38 in the commercial milk kept at —9°C.; and over 72 in both milks kept at 0°C. In the latter experiments the total nitrogen decreased to a marked degree; the loss is ascribed to the liberation of elemen- tary nitrogen. These investigators refer the proteolysis at 0°C. to i terial action; that at —9°C. to the action of the native milk a Tet: galactase. Since their method of chemical analysis dif- ered widely from that used in the present research the results obtained in the two studies are not strictly comparable. In the reinfected, sterile mill ‘casein nitrogen decreased pro- gressively but to a far less de; lan did the casein nitrogen of the raw untreated milk. The albumin and syntonin itrogen varied within narrow limits and showed no marked change ¢ gtheperiod | of keeping. The caseose nitrogen showed a progressive increase most marked during the last third of the experiment. The amino- . acid nitrogen increased to some extent, and the peptone nitrogen. tended to decrease. Since, on the final analysis, the total nitrogen q and the total nitrogen including nitrates were the same, nitrates were not formed during holding. Hence, nitrogen fixers were not present in the milk. The amount of free ammoniacal nitrogen was small and apparently tended to decrease. In the formolized, raw milk, the casein nitrogen remained practi- cally constant during the entire period of holding. The albumin and syntonin nitrogen tended to decrease and the caseose nitrogen to increase. The peptone nitrogen showed a tendency to decrease, and the amino-acid nitrogen a tendency to increase. The free ammoniacal nitrogen was a negligible quantity. In this connec- tion it should be mentioned that Tice and Sherman, during a study of formolized raw milk held at room temperature for periods as long as thirty-seven months, noted that “the albumin was largely digested before the original amount of casein was appreciably reduced.’”’ Hence, at the temperature of the chill-room and of the % Loc, cil. Pennington and Collaborators 339 room, the same type of proteolysis occurs in formolized raw milk, and is produced mainly, if not entirely, by galactase—a native milk enzyme; possibly proteolytic enzymes derived from the dead bacteria, may also participate in the digestion of the protein. On the other hand, Sherman, Berg, Cohen and Whitman*’ reported that the free ammoniacal nitrogen increased in formolized raw milk kept in the room at 15°C. for three months, while in the pres- ent research the free ammoniacal nitrogen was found to be an absolutely negligible quantity in formolized raw milk held at 0°C. In the sterilized milk, during the entire period of storage, the changes in the distribution of the nitrogen were but slight, and, on the whole, lie within the limits of analytic error. ' The study of the nitrogen results leads to the followi sions. The proteolysis of the casein is, primarily, uC origin, since it occurred in the reinfected a yut i the formolized raw milk. The e digestion 0 of the albumin and syn- tonin is, primarily, due to m ative enzymes of the milk, since it took place in th n lized raw but not in the reinfected. sterile milk. In raw untreated milk, however, the native enzymes and bacterial flora act in combination in giving rise to more rapid . oteolytic changes, since in the same period of time—five weeks— over twice as much casein was digested in the raw untreated milk as in the reinfected sterile milk. The general trend of the pro- teolysis, enzymic, bacterial and combined, is toward a tryptic digestion, that is, the passage through caseose and peptone to amino-acids which accumulate as the period of holding lengthens. The changes in the ammoniacal nitrogen are negligible. Acidity. The acidity of the raw, untreated milk increased more or less progressively to the highest values of the entire series of experi- ments. In the reinfected, sterile milk, the acidity increased pro- gressively and finally attained values which were second only to those obtained in the raw untreated milk. In the formolized, raw milk the acidity first increased, then decreased to a value which remained fairly constant to the very end of the experiment. The initial rise was possibly due to the bacterial enzymes of the dead 26 Sherman, Berg, Cohen and Whitman: This Journal, iii, p. 171, 1907. 340 Changes in Milk and Cream at 0°C. TABLE I. Chemical changes in raw and treated clean milk kept at 0° C. (The per cent of the various forms of nitrogen in the milk are printed in Roman; the various forms of nitrogen are also printed as per cents of the total nitrogen in italics. Lactose is expressed as per cent and acid- ity as ec. of 4 NaOH required to neutralize 100 cc. of milk.) 2 DISTRIBUTION OF THE NITROGEN , Bs | Bin] aren 4 3 8S “a | = pommel. & : - * Z se | = o az = 3 3 2; = A} ES 8 =z af g & iS Bes |g | Be) es| 2 |8| & 3 Sijee | © [toy oS g |. kad 4 & Raw untreated milk. deg. C. 7, 0.033) 0.007; 0.019) 0.00047 |5.3) 18.0 | —0.550 8.265 8 0. 16.5 | —0.540 : 32 af i. 46.5°|2 56.0 Fresh | 0.608 0.528) 0.018! 0.029| 0.010! 0.0231 0.0029 |4.5| 25.8 | —0.735 186.84 | 2.96 | 4.77 | 1.64 | 8.78 | 0.477 7 | 0.595 0.516 0.021) 0.016, 0.030 0.012, 0.00356 |5.1) 26.0 | —0.550 86.72 | 3.53 | 2.69 | 5.04 | 2.02 | 0.598 : 14 | 0.585, 0.522; 0.015) 0.029) 0.033:40.014) 0.00410 |4.2) 32.0 | —0.560 89.23 | 2.66 | 4.96 | 5.64 | 2.39 | 0.701 21 | 0.590 0.507) 0.013 0.027 0.00210 |4.4| 30.8 | —0.585 85.93 | 2.20 4.58 0.356 28 | 0.592, 0.463, 0.015, 0.080, 0,014) 0.020, 0.00168 3.3) 41.3 | —0.570 78.21 | 2.63 13.61 | 2.86 | 3.38 | 0.284 35 | 0.506, 0.444) 0.023, 0.105, 0,026 t0 002 0.00108 |3.5) 51.8 | —0.680 74.60 3.86 [17.62 | 4.86 | 0.84 | 0.181 *Total nitrogen, including nitrate pitr at 35th day of hol raw untreated milk 0.565; ss storile mil vet — yt : flo this . the sum of the nitrogenous constituents determined exceeds the total nitrogen of tthe (alli y this amount, Pennington and Collaborators 341 TABLE I—Continued. RR ae DISTRIBUTION OF THE NITROGEN Meise A Bz ree 4 | : s) q g g Saee bs atom | 2 1 et E : aeciee fb 2 | 28 ¢ | See fg z cela|i |e i 2/32) og ele] § pele é |2° | S* | Seeeray ee Formolized raw milk. days , deg. C. Fresh | 0.585 0.033} 0:007, 0.019| 0.00047 |4.8| 18.0 | —0.550 5.64| 1.20 | 3.25 | 0.080 7 | 0.573 0.023, 0.015) 0.029, 0.00060 5.1} 19.0} —0.660 4.01 | 2.62 | 5.06 | 0.010 © 14 | 0.564 0.031} 0.026 0.000 0.00000 4.4) 27.8 | 5.50 | 4.61 | 0.00 = 21 | 0.550 0.025 .0| 19.8 | —0. 4.55 : 28 | 0.561 0.048|.0.012, 0.004) 0. 21.0 0.| 8.56 | 2.14 | 0.71 | 0. 35 | 0.602 3) 0.047] 0.019 0.005 21.5 | —0.680 48 | 7.81 | 3.16 | 0.83 is Sterilized milk.t “er Fresh| 0.608) 0.528) 0.018) 0.029 0.010 0.023, 0.0029 4.5) 25.8 | —0.735 : 86.84 | 2.96 | 4.77 | 1.64 | 3.78 | 0.477 35 | 0.637| 0.547 0.012 0.027 0.027] 0.024) 0.00298 4.6) 26.0 | —0.615 85.87 | 1.88 | 4.24 | 4.24 | 3.77 | 0.468 tNo analyses made during first four weeks of holding. organisms, which had been killed by the formaldehyde. The subsequent decrease in acidity may be ascribed to either a neutrali- zation or further decomposition of the lactic acid. In the steril- ized milk the acidity remained constant during the entire period of holding. Ravenel, Hastings and Hammer’ report a decrease in the acidity of a clean milk and of a fair grade of commercial milk, held at —9°C. for a period of 203 days; and an increase, decidedly progressive, in the acidity of both the clean milk and the commercial milk held at 0°C. for that period of time. 27 Loe. ett. 342 Changes in Milk and Cream at 0°C. Lactose. The lactose content of the raw, untreated milk decreased pro- gressively, the greatest loss occurring during the earlier portion of the period of storage. In the reinfected, sterile milk the decrease in lactose tended to parallel that of the raw untreated milk. The tendency was for the lactose to decrease but little, if at all, in the formolized, raw milk; a similar tendency was noted by Tice and Sherman*’ in raw formoliaee milk kept at the temperature of the laboratory. The sterilized milk showed no change in lactose con- tent during the period of holding. The fermentation of the lactose ith the formation of lactic acid was then largely, if not exclusively, to the activity of bacteria. Freezing point. The decdteiost ean, of the uents of the milk, eS the carbohydrate and the fF wi small molecules from one large mo Bile must give rise to a higher molecular concentration of the solutes of the milk and, therefore, should be accompanied by a depression, or lowering, of thefreezing point of the milk. The analytic findings are in perfect harmony — with this theory. In the raw, untreated milk, the proteolysis and — the fermentation of the lactose were accompanied by a lowering of : the freezing point, which tended to be a progressive change. The — changes in distribution of the nitrogen in the formolized, raw milk and the changes in protein and lactose in the reinfected, sterile milk, were likewise accompanied by lowerings of the freezing point. THE CHEMICAL CHANGES IN THE CREAM DURING HOLDING. The fat constants. The iodine number remained practically unchanged in all the experiments. The index of the refraction underwent no change in any of the samples. The Reichert-Meiss! number showed no marked change. In the raw, untreated cream, the saponification number, Hehner number and acid value seemingly increased progressively; the greatest rise in the saponification number oceurred during the * Loc. cit. . Pennington and Collaborators 343 first week. In the. reinfected, sterile cream, the saponification number showed an increase, followed by progressive decreases. The Hehner number and the acid value increased, with a tendency to do so progressively, the greatest increase being during the first week. In the formolized, raw cream the saponification number increased, especially during the first week, the Hehner number also increased with a tendency to do so progressively, although the greatest increment was during the first week. The acid value increased to a very slight extent. In the sterilized cream, the saponification number and the Hehner number increased, while q the acid value suffered a slight decrease. If the values of the saponification and Hehner numbers, and, the acid value obtained on the initial analysis, be compared wi _ the highest values obtained in the subsequent analyses of ach series, it will be observed that the greatest i in acid value occurred in the reinfected sterile cream, the least in the formolized raw cream, with the raw untreated cream occupying an intermedi- ate position. Thi o1 d tend to show that the fat-splitting is of bacterial origin rather than due to enzymes of the cream. The simultaneous increase in both Hehner and saponification _ numbers occurred even in the sterilized cream, showing that this . reaction may depend simply on oxidation and the fine state of - division of the butter-fat, possibly aided by a thermostable inor- _ ganic catalyst. Since this change was most pronounced in the _ formolized raw cream, native enzymes of the cream must also play a prominent réle in its production. The progressive decreases in the saponification number, accom- panied by an increase in the Hehner number, in the reinfected sterile cream denote a type of fat decomposition, different from that observed in the sterilized, formolized raw, and raw untreated cream. Apparently, the raw untreated cream represents the resultant of the bacterial changes, revealed by the reinfected sterile cream, of oxidation and catalytic changes seen in the sterilized cream, and of the enzymic changes occurring in the formolized raw cream. The fine state of division of the butter-fat in the cream also, doubt- less, plays a réle in all the samples by exposing a large surface for oxidation. 344 Changes in Milk and Cream at 0°C. TABLE II. Chemical changes in raw and treated clean cream kept at 0°C. FS FAT CONSTANTS tel g 1} 2 ° =] 3 ne Lal 48 os mie (dei a | a led! (ze lea] a | cei 2s|2\2| § |oe8| ap lcesigee| 6 | 3 ce | a° 3S | Sete tee Se Raw untreated cream ¥ eS per cent, deg. C. 32.0 | 206.0,0.5 | 205.5 0.25 | 76.8 | 31.1 |1.45450.1159) —0.543 32.5 | 228.7; 1.8 | 226.9) 0.91 | 81.75) 34.6 |1.4546)0.1061) —0.620 33.1 | 221.2) 1.2 | 220.0 0.60 | 85.0 | 31.0 |1.454310.0917) —0.685 33:3 | 229.8) 1.3 | 228.5) 0.65 | 85.6 | 33.9 |1.454310.1220| —0.720 31.4 | 226.5, 2.1 | 224.4 1.06 | 83.0 | 33.0 |1.4548/0.1388| —0.712 Fresh | 32.1 ik; i' | 30m WE 14 | 32.9 i A 21 | 33.4 yh Fresh | 32.0 | 206.0 0.5 | 205.5 0.25 | 76.8 | 31.1 |1.45450.1159 7 | 32.8 | 233.7| 0.52 | 233.2 0.26 | 83.6 | 27.4 |1. 1249, —0. 14 | 33.0 | 232.9] 0.43 | 232.5 0.22 | 85.4 | 22.0 1, 1342) —0.655 21 | 32.5 | 238.6 0.33 | 238.3: 0.17 | 85.9 | 35.5 {1.45440.1520, —0.650 28 | 30.9 229.3 0.53 | 228.8) 0.27 | 84.0 | 35.4 |1. 0.1222 —0.660 Sterilized cream.* Fresh | 32.1 | 213.2 0.7 | 212.5 0.35 | 79.82 26.4 |1.454610.1067/ —0.565 28 | 33.0 | 235.2 0.3 | 234.9| 0.15 | 85.6 1.4551/0.1158) —0.622 *No analyses made during firat three weeks of holding. Pennington and Collaborators 345 The lecithin. This lipin apparently has not been decomposed in any of the four series of experiments. Freezing point. The raw, untreated cream showed a progressive lowering of the freezing point, as did the reinfected, sterile cream, though to a less degree. The formolized, raw cream showed a depression during the first week, then remained fairly constant. The sterilized cream showed a slight depression, but less than took place in the other samples. These changes were probably due to digestion of the protein and lactose of the cream, rather than to fat decomposi- tion, for the Reichert-Meissl number, which r be accepted as a fair measure of soluble fatty acids, had undergone no marked in- crease; therefore, no soluble decomposition products of the butter- fat had been formed to exert an influence on the freezing point. ENZYMES OF THE MILK AND CREAM. The results of the zymochemical experiments have been collected in Table III. In the raw, untreated milk and cream, reductases which attack methylene blue were apparently absent from both the fresh milk and the fresh cream, but were invariably present after the first week of holding at 0°C. Reductases which act upon methylene blue, plus formaldehyde, were present in both the fresh milk and the fresh cream and retained their activity through- out the period of storage. Oxidases which give rise to the oxida- tion of trikresol were always present in both the milk and the cream, while oxidases which are reactive toward guaiac were found in but three of the six milk analyses and in but two of the five cream analyses. In the reinfected, sterile milk and cream, reductases which destroy the color of methylene blue, as well as reductases which decolorize methylene blue, plus formaldehyde, were present in all the samples of milk save that tested after holding for one week, and were invariably present in the cream. Oxidases which produce oxidation of trikresol were always present in both the milk and the cream, while oxidases which cause a coloration of guaiac were always absent from the cream and were found only in the fresh 346 Changes in Milk and Cream at 0°C. milk. In the formolized, raw milk and cream, reductases which decolorize methylene blue were absent from the fresh milk and from the final sample (held for thirty-five days at’0°C.) but were present in the other milk samples and invariably present in the cream. Reductases, which act on methylene blue plus formaldehyde, were always present in the cream and occurred in the fresh milk but were not found in any of the samples of milk which had been held at 0°C. for one or more weeks. Oxidases which attack trikresol were invariably present in both the milk and the cream, while oxidases which are reactive'toward guaiac were invariably absent from both the milk and the cream. nfortunately, the data on catalase are incomplete. However, his enzyme apparently occurs as a true milk enzyme and is also secreted by the microérganisms. The raw untreated milk was always able to liberate more oxygen from hydrogen peroxide than was either the reinfected sterile the formolized raw milk, yet this rule did not hold good in’ of the cream. At the end of two weeks’ storage at 0°C. the cream showed greater catalytic activity than did the milk, but the reverse was true at the end of the third week. » To sum up, reductases which attack methylene blue were ‘normal constituents of the raw untreated, reinfected sterile, and formolized | raw samples of milk and of cream. Reductases which decolorize methylene blue in the presence of formaldehyde were invariably present in all three kinds of cream, were normally present in the — raw untreated milk and in the reinfected sterile milk, and were usually absent from the formolized raw milk. It is, therefore, probable that the aldehyde reductases of the milk were of bacterial origin. Oxidases, which give rise to an oxidation of trikresol, were in- variably present in both milk and cream—raw untreated, rein- fected sterile, and formolized raw. Oxidases, which produce a color with guaiac, were present in about half the experiments on raw untreated milk and cream, and were absent from all the other samples with the single exception of the fresh reinfected sterile milk. These results point to the conclusion that the simple reductases and the trikresol oxidases of the milk may be enzymes native to the milk and may also be of bacterial origin. The aldehyde reductase hn III. t and cream kept at 0°C. : . + and (The presence or absence of enzymes is expressed by the Enzymes of raw and — respectively, except in the case of catalase where the figures represent cc. of oxygen evolved.) Pennington and Collaborators 347 gj meer) +444 ae] ++ i=} 5 | ¥ owen | + | |. hee ee —_— Siemmeng | + bee +++ § | euerAqioy | b B | 8 | oun, ti tt+t| +++ o asvIvivo 88 3k ben | a orn é. PReARL| + + ee E 5 omen | | dt dd a g pa + 1 | . a |g shail te ° if en aa ae S| ouapsone L++++ 1 GSVIVLVO SR SR alt Sood MN 19 pre t++tt+++ /++4+4++ : 8 omen) + i++ i) i tit a apay E G feong | H+ t++++i++4+4++ P | 6 |euémen eS Sa & b 18 | neat let+te|i +444 } =O 8 BS R& GSVIVLVO Aan No f=! Skiarona | @™*~N RSE AAR SXyd NI Golda) f ‘ay te SY | I f SE : ft wa f 5 5 : g THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3 348 Changes in Milk and Cream at 0°C. of the milk was probably of bacterial origin, and apparently the guaiac oxidases arose from the same source. In the cream both varieties of reductase and the trikresol oxidases were apparently native enzymes and were also secreted by the microérganisms, while the guaiac oxidases were probably of bacterial origin. The cata- lase of both the milk and the cream was a true milk enzyme and was also due to the activity of microérganisms. All five enzymes— the two varieties of oxidase, the two varieties of reductase, and the catalase—retained their power to act as catalytic agents in spite of the prolonged exposure to a temperature of 0°C. THE BACTERIAL CHANGES IN THE MILK DURING HOLDING. Conn and Esten?* report three experiments in which milk was kept at 1°C. for periods as long as 42 days, bacteriological analyses being made at intervals of 2 days to 9 days. During the first 6 or 8 days of holding, scarcely 4 terial development occurred, then the organisms increased steadily until very great numbers were present. Since the usual ‘lactic acid organisms were not in the majority, the milk did not curd. A comparatively large number of liquefying organisms, and of neutral organisms, which produce neither acid nor alkali, were found in the milk. There was also a_ tendency for certain organisms to disappear during the period of holding. Ravenel, Hastings and Hammer,” studied bacterial growth i in milk held at low temperatures. The milk was of two kinds, “barn milk,” the best milk obtainable, and ‘‘dairy milk,” a fair commercial article. One sample of each kind of milk was kept at —9°C., and at 0°C., for a maximum period of 203 days, analyses being made at varying intervals. Up to and including 160 days, plates of lactose agar were sown and incubated at 37°C.; plates of plain gelatin were sown and incubated at 12 to 15°C. up to and including 203 days. During holding at —9°C., the number of organisms growing on agar at 37°C. remained fairly constant in both kinds of milk, the variations being insufficient to permit of any definite conclusions. The number of organisms developing ** Conn and Esten: Sixteenth Annual Report of the Storrs Agricultural Ha- periment Station, 1904, p. 27. © Loc. cil, i Pennington and Collaborators 349 on gelatin at 12 to 15°C. was fairly constant in the barn milk, but decreased considerably and with a marked progressive tendency in the dairy milk. During holding at 0°C., the organisms growing on agar at 37°C. underwent a quiescent stage in both kinds of milk for six days, then increased progressively, in the barn milk to the end of the period of holding, in the dairy milk up to and ineluding seventy-four days, afterward decreasing progressively. The organisms devel- oping on gelatin at 12° to 15°C. were characterized by progressive increases to a maximum count followed by decreases; the maximum was attained in the dairy milk much earlier than in the barn milk. In the raw, untreated milk on the plates incubated at 37°, ae total count per cc. increased progressively up to the twenty- rst day, then dropped rapidly and progressively. The anaerobes and facultatives rose during the first week, then fell; the maximum count, however, was at-21 days, after ) vhich a rapid progressive decrease occurred. The acid formers underwent a progressive increase up to the twenty-first day, then a rapid progressive de- crease took place. On the plates incubated at 20°, the total count i per ce, showed a progressive increase throughout the period of _ fholding. There was a decided tendency for the anaerobes and Hacultatives to increase, the highest counts being at 21 and 28 days. The acid formers and the liquefiers increased progressively. On the plates incubated at 0°, the total count per ce. showed pro- gressive increases and reached a maximum at 35 days. The anaerobes and facultatives tended to increase, and while this _ increase was not absolutely progressive, the highest counts were at 21 and 35 days. The acid formers were characterized by an increase, which, on the whole, was progressive. In the reinfected, sterile milk, on the plates incubated at 37°, the total count per cc. rose to a maximum at 7 days, then suffered a , progressive decrease throughout the experiment. The changes in ~ the anaerobes and facultatives paralleled those of the total count, the maximum being at 7 days, after which there was a progressive decrease until no growth was obtained on the final analysis. The changes in the acid formers ran parallel to those of the total count _ and of the anaerobes and facultatives, the maximum at 7 days, then __a progressive decrease to the end of the period of holding. On the - plates incubated at 20°, the total count per ec. tended to rise 350 Changes in Milk and Cream at 0°C. progressively, the maximum being at 28 days. The number of anaerobes and facultatives showed wide fluctuations, the greavest number being found at 14 and 35 days. The acid formers showe4 a decided trend to increase progressively, the maximum being! 28 days. The liquefiers increased progressively throughout ti. experiment. On the plates incubated at 0°, the total count per cc. rose progressively throughout the experiment. The anae- robes and facultatives and the acid formers showed a general ter ~' ency to increase progressively, reaching their highest counts / 28 days. v. While a few organisms, including anaerobes and facultatives aad cid formers, were present in the fresh formolized raw milk, yet in the subsequent analyses the plates were almost invariably sterile. The formolized raw milk of Tice and Sherman, which was kept at the temperature of the laboratory, at times contain’ liquefying cocci, which formed yellow colonies, as well as cocci and bacilli, which formed white colonies. The sterile milk retained its sterility throughout the period ot holding, as was demonstrated by platings made on the initial and final analyses. x Comparison of the raw and reinfected milk. As a general rule, the total counts on the raw untreated mil were higher than on the reinfected sterile milk at all temperatures of incubation through the experiment, although the initial total count was practically the same. The initial counts of anaerobe: and facultatives were fairly close in the two kinds of milk; during the first 2 weeks the reinfected sterile milk showed the highest count at all temperatures of incubation; at 3 weeks the raw un- treated milk had the higher count at all temperatures of incuba- tion; the raw untreated milk also continued to have the higher count at 37° throughout the rest of the experiment. The initial count of acid formers differed in the two experiments, so the counts throughout the experiment can scarcely be compared. It may be noted that the relation of the acid formers to the total count varied throughout both series. At times the number of liquefiers was higher in the raw untreated milk, at times in the reinfected sterile milk. Loc, cit. Pennington and Collaborators 351 Influence of temperature on the organisms. On comparison of the counts, incubated at the three tempera- tures on the same analysis, the total count per cc. in the raw un- treated milk had either 20° or 0° as the optimum temperature after the first week, in the reinfected sterile milk, 20° after the first week and 0° at the fifth week. The anaerobes and facultatives in the raw untreated milk passed from an optimum temperature of 37° through 20° to 0°, while in the reinfected sterile milk the optimum growth was at either 20° or 0°. The acid formers almost invariably had a maximum count on the plates grown at 20°. Ratio of increase of the organisms. The maximum count obtained during the period of holding with each group of organisms at each temperature of incubation is always compared with the count obtained with that group of or- ganisms at the same temperature at the beginning of the experi- ment, using the latter figure as unity in the ratio. Raw, Unrreatep Miix. Total cownt per cc. At 37° the maximum was attained at 21 days and was 463.9 times the count on the fresh milk. At 20° the maximum was at 35 days and the ratio of increase 1:3563.2; at 0° the maximum was at 35 days and the ratio of increase 1:290909.1. Anaerobes and facultatives. At 37° the maximum was at 21 days and the ratio of in- crease 1:118.9, at 20° the maximum was at 21 days and the ratio 1:340.0, at 0° the maximum was at 35 days and the ratio 1:533.3. Acid formers. At 37° the maximum was at 21 days and the ratio of increase 1:1913.0, at 20° the maximum was at 35 days and the ratio 1:12258.1, and at 0° the maximum was at 35 days and the ratio 1:357894.7. The maximum count of liquefiers was at 21 days and the ratio of increase 1:19000.0. REINFECTED, STERILE Mitx. Total count per cc. At 37° the maximum was at 7 days and the ratio of increase was 1:13.0, at 20° the maximum was at 28 days and the ratio 1 :1437.5, at 0° the maximum was at 35 days and the ratio 1:19200.0. Anaerobes and facultatives. At 37° the maximum was at 7 days and the ratio of increase 1:34.6, at 20° the maximum was at 28 days and the ratio 1:505.9, at 0° the maximum was at 28 days and the ratio 1:454.5. Acid formers. At 37° the maximum was at 7 days and the ratio of increase 1:9.9, at 20° the maximum was at 28 days and the ratio 1:3035.7, and at 0° the maximum was at 28 days and the ratio 1:13333.3. The maxi- mum count of liquefiers was at 35 days and the ratio of increase 1:5000.0. 352 Changes in Milk and Cream at 0°C. Taste IV. Bacterial content a (The figures in Roman represent bacterial ¢ RAW UNTREATED MILK PERIOD OF PLATES KEEPING INCUBATED IN DAYS at °C Total count | ABaerobes and Facultatives per cc. per ce. per cc. Acid Formers Liquefiers per cc. Be Oe 97,000 18,500 11,500 1 15,500 1,000 1 1 475 | 1 3,300,000 28.7 4,700,000} 1,500,000 803.2 1500.0 4,400,000 9263.2 7,600,000 660.9 46,000,000 00,000 2967.7 2700.0 25,000,000] ils 52,631.6 ea 22,000,000 1913.0 120,000,000 | 19,000,000 7741.9 19,000.0 140,000,000 | 294,736.8 80,000 Fresh.i<.2.... 20 Ye eer. = aueemeend —) sala ae a) —) 67,000,000 Ss 8 190,000,000} 15,000,000 12,258.1 | 16,000.0 357,894.7 Pennington and Collaborators 353 ease in milk kept at 0°C. . ares in Italic represent ratios of increase.) REINFECTED STERILE MILK agen perpen noses count “Pacultacives ee? Formers saeeetere me count 100,000 26,000 | 91,000 400 1 1 1 160,000 17,000 56,000 5,000 140 1 1 1 1 12,500 4,400 9,000 0 1 1 | 1 1,300,000 900,000 900,000 13.0 34.6 | 9.9 ge 1,500,000 600,000 —:1,000,000 =. O 9.4 35.3 17.9 58,000 27,900 | 0 4.6 8.0. 900,000 120,000 330,000 0 9.0 vent, KB 3.6 10,000,000 | 1,400,000 9,000,000 0 62.5 82.4 160.7 9,100,000 460,000 4,600,000 0 ue «(788.0 104.5 511.1 we” 200,000 50,000 120,000 5 tl 2.0 1.9 | 1.3 72,000,000 330,000 54,000,000 12,000,000 450.0 19.4 964.3 2400.0 67,000,000 550,000 32,000,000 5400.0 125.0 3711.1 29,000 10,000 27,000 0 0.29 0.88 0.30 230,000,000 | 8,600,000 —_| 170,000,000 14,000,000 0 1437.5 505.9 3035.7 2800.0 180,000,000 | 2,000,000 120,000,000 | 0 14,400.0 454.6 13,333.3 3,000 0 1,000 0.03 0.01 160,000,000 | 1,300,000 72,000,000 25,000,000 1000.0 76.5 1285.7 5000.0 240,000,000 | 1,100,000 62,000,000 19,200.0 250.0 6888.8 354 Changes in Milk and Cream at 0°C. The general trend was for organisms which grow best at 37° to reach a maximum during the earlier stages of the period of hold- ing, while those growing best at 20° and at 0° continued to increase and reached their highest values during the later stages of the experiment. If the ratio of increase in the raw untreated milk at each period of analysis and at each temperature of incubation for each group of organisms be compared with the similar ratio in the reinfected sterile milk, it is seen that, in 42 of the 46 cases, 91.3 per cent, the ratio is higher in the raw milk than in the reinfected milk. There- fore, at 0°C. the raw milk has been the more suitable medium for 7 oii oe Co and has given rise to a greater rate of proliferation. _ The relative rate of growth of milk organisms in raw and pas- teurized milk has been studied by several investigators. Accord- ing to St. John and Pennington® when raw clean milk and rein- fected pasteurized clean milk of a imately the same bacterial content are held at the temperature of the room and of the ice box, the milk organisms proliferate more rapidly in the reinfected than in the raw sample. Rickards* likewise found that bacteria increase more rapidly in pasteurized than in unpasteurized milk at the temperature of the ice box. Ayers and Johnson* state that their results, obtained on milk kept at 10°C., ‘‘tend to prove that bacteria do increase approximately the same in pasteurized as in raw milk, provided their initial counts are practically alike.” The mode of preparing the reinfected sterile milk in the present research was the same procedure as used by St. John and Penning- ton in their study of reinfected pasteurized milk. The rate of bacterial growth was greater in the reinfected pasteurized milk than in the raw milk but was greater in the raw milk than in the reinfected sterile milk. On the other hand, in the experiments on cream described later in this paper, the rate of bacterial increase was greater in the reinfected sterile cream than in the raw cream. The following explanation is offered for these % Loc. cit. : % Rickards: Amer. Journ. of Public Hygiene, xix (New Series, v), p. 507, 1909. * Ayers and Johnson: U.S. Dept. of Agric., Bureau of Animal Industry, Bulletin 126, 1910, p. 52. r | Pennington and Collaborators 355 phenomena. The pasteurized milk underwent no marked change in color and suffered no chemical change save, possibly, a partial coagulation of the lactalbumin; pasteurization had been attained by holding the milk at 79°C. for 20 minutes. The recent work of Rupp® demonstrates that the only appreciable chemical change produced in milk by pasteurization is a partial coagulation of the lactalbumin. Thus Rupp found that pasteurization of milk by holding at 71.1°C. for 30 minutes coagulated 30.87 per cent of the total lactalbumin of the milk. This investigator also noted that an increase in the temperature of pasteurization, the time ) factor remaining constant, gave rise to an increased coagulation of ) lactalbumin. I! The sterilized milk had a marked light golden color; the major portion of the lactalbumin had been coagulated; and a portion of li the lactose must have undergone decomposition, for the lactose | content should have risen to a slight extent on account of the con- centration during sterilization, yet the sterilized milk contained | less lactose than did the fresh raw milk. One evidence of this decomposition of lactose is the greater acidity of the sterile than of the fresh raw milk on the initial analysis. The increase in lactic acid, however, is not sufficient to account for the decrease in lactose, and it appears probable that certain other decomposition products of the lactose may exert an inhibitory action on the bacterial growth in the reinfected sterile milk. In the cream, the change in color as the result of sterilization was not noted, and moreover the lactose content of cream would be much less than that of milk, so the inhibitory substances would be formed to but a slight degree, if at all. During the process of sterilization, however, the complex mole- cules of the organic constituents of the milk may possibly be re- arranged in a manner which escapes detection by the ordinary methods of milk analysis, and the value of the milk as a nutrient medium for the rather fastidious organisms may be lessened, thus accounting for the slow rate of proliferation in the reinfected sterile milk. 3° Rupp: U. S. Dept. of Agric., Bureau of Animal Industry, Bulletin 166, 1913. 356 Changes in Milk and Cream at 0°C. The bacterial flora of the raw untreated milk and of the reinfected sterile milk. The organisms which were isolated in pure culture from each sample of milk at each period of analysis are enumerated in Table V, according to the temperature of incubation of the plate on which each was found: The predominant species of organism at each temperature of incubation and each period of analysis of each sample is designated by a double asterisk (**), the species next numerous is designated by a single asterisk (*); at times, two or- ganisms shared the premier position, in such cases both are des- ' ignated by a double asterisk, and the single asterisk is omitted. . —-. the fresh raw, untreated milk, on the plates incubated at 37°, ) he predominant species was Micrococcus ovalis (Escherich) ; throughout the period of holding this organism tended to retain its position, although it twice shared that position with Micro-_ coccus aurantiacus (Cohn), and once was displaced by the latter. On the plates incubated at 20°, M. aurantiacus held the premier position most of the time throughout the experiment, although it was once equalled and twice exceeded by M. ovalis. On the plates incubated at 0°, M. aurantiacus predominated until the very last analysis when it gave place to M. ovalis. At all three tempera- tures, whenever M. ovalis predominated, M. aurantiacus held the — second place with respect to frequency of occurrence, and vice versa. Both these micrococci are acid formers. Apparently, M. aurantiacus preferred the lower temperatures of incubation, while M. ovalis became acclimated to those temperatures during the period of holding at 0°C. This process of acclimatization was also undergone by other organisms, thus Bacterium flecuosum was not found on the plates sown with fresh milk and incubated at 0°, but occurred on the 0° plates beginning with 7 days of holding and continuing to the end of the experiment. | On the other hand, some organisms soon disappeared from the milk during holding. Thus Micrococcus acidi lactici was re- covered on the plates incubated at 37° from both the fresh milk and the milk after storage for 7 days, but thereafter could not be isolated from such plates. This organism was never found on the plates incubated at 20° and at 0°. In the fresh reinfected, sterile milk, the predominant organism at 37° was Micrococcus ovalis (Escherich), at 20° and at 0° Micro- Pennington and Collaborators TABLE V. 357 Organisms isolated in pure culture from raw untregted and reinfected sterile milk, kept at 0° C. - p? O71 lis (Escherich) pony FROM THE HOLDING tae gl RAW UNTREATED MILK REINFECTED STERILE MILK —O ar °C. . M. acidi lactici (Linder) | M. albus liquifaciens M. albus liquifaciens **M. ovalis (Escherich) 37 *M. aurantiacus (Cohn) | S. farcinica **M. ovalis (Escherich) 8. flava 8. flava *Penicillium . Fresh M. albus liquifaciens **M. aurantiacus ee **M. aurantiacus (Cohn) *M. ovalis (Escherich) pe 20 *M. ovalis (Escherich) B. flexuosum t) B. flexuosum (Wright) . flava te _S. flava | li » **M. aurantiacus (Cohn) |**M. aurantiacus (Cohn) . ovalis (Escherich) 14 _ M. acidi lactici (Linder) M. albus liquifaciens *M. aurantiacus (Cohn) **M. ovalis (Escherich) B. flexuosum (Wright) B. detrudens (Wright) . albus liquifaciens *M. aurantiacus (Cohn) **M. ovalis (Escherich) B. flexuosum (Wright) **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. flexuosum (Wright) 8. flava M. albus liquifaciens **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. flexuosum (Wright) **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. flexuosum (Wright) 8. flava *M. \urantiacus (Cohn) **M/ ovalis (Escherich) 37 **M. aurantiacus (Cohn) **M. ovalis (Escherich) B. flexuosum (Wright) . aerius . aurantiacus (Cohn) . ovalis (Escherich) 20 *M. aurantiacus (Cohn) **M. ovalis (Escherich) S. flava . aurantiacus (Cohn) **M. ovalis (Escherich) B. flexuosum (Wright) S. flava ** The predominant species of organism at each temperature of incubation and at each period of analysis. *The second most numerous species of organism at each temperature of incubation and at each period of analysis. 358 Changes in Milk and Cream at 0°C. TABLE V.—Continued. pao IN FROM THE Ponape ed Pepa RAW UNTREATED MILK REINFECTED STERILE MILK MILK AT at °C °c. **M. aurantiacus (Cohn) /**M. aurantiacus (Cohn) 4 0 *M. ovalis (Escherich) *M. ovalis (Escherich) 1 B. flexuosum (Wright) B. flexuosum (Wright) S. flava S. flava aus **M. aurantiacus (Cohn) (|**M. aurantiacus (Cohn) 97 *M. ovalis (Escherich) *M. ovalis (Escherich) , B. flexuosum (Wright) S. flava **M. aurantiacus (Cohn) M. albus liquifaciens **M. ovalis (Escherich) **M. aurantiacus (Cohn) 20 B. flexuosum (Wright) *M. ovalis (Escherich) 8. flava B. flexuosum (Wright) de 2 a S. flava **M. aurantiacus | 3 **M. aurantiacus (Cohn) 0 | *M. ovalis (Escherich) *M. ovalis (Escherich) B. flexuosum (Wright) B. flexuosum (Wright) 35 **M. aurantiacus (Cohn) /|**M. aurantiacus ( hn) **M. ovalis (Escherich) *M. ovalis (Escherich) bcs a *M. aurantiacus (Cohn) M. aurantiacus (Cohn) | 98 20 |**M. ovalis (Escherich) |**M, ovalis (Escherich) S. flava *B. flexuosum (Wright) **M. aurantiacus (Cohn) | *M. aurantiacus (Cohn) 0 | *M. ovalis (Escherich) |**M. ovalis (Escherich) B. flexuosum (Wright) B. flexuosum (Wright) *M. aurantiacus (Cohn) |**M. auranticus (Cohn) 37 |**M. ovalis (Escherich) *M. ovalis (Escherich) 8. flava *S. flava **M. aurantiacus (Cohn) |**M. aurantiacus (Cohn) 35 20 *M. ovalis (Escherich) *M. ovalis (Escherich) 4 B. flexuosum (Wright) B. flexuosum (Wright) 8S. flava S. flava *M. aurantiacus (Cohn) |**M. aurantiacus (Cohn) 0 \**M. ovalis (Escherich)’ *M, ovalis (Escherich) | ** The predominant species é organiam at each temperature of incubation and at each period of analysis, * The second most numerous species of organiam at oach temperature of tnoubation and at each period of analysts. | | \ B, flexuosum (Wright) Pennington and Collaborators 359 ~ coccus aurantiacus (Cohn). Throughout the period of holding, at each of the three temperatures of incubation, M. ovalis predomi- nated twice, M. aurantiacus thrice. The two lower temperatures, at the beginning of the experiment offered a more favorable envi- ronment to M. aurantiacus than to M. ovalis, yet the latter exhibited a tendency to become acclimated during holding of the milk at 0°, as was shown by its predominance at 28 days on the plates in- cubated at 20° and at 0°. These two acid-forming micrococci— M. aurantiacus and M. ovalis—almost invariably were the two more numerous species. On the plates at 37°, however, the second place in the fresh sample was occupied by Penicillium, and in the final sample was shared by M. ovalis and Streptothrix flava, while on the plates at 20° the second place at 28 days was oceupied by Bacterium flecuosum. The tendency to become acclimated to the low temperature was also shown by Bacterium flecuosum which disappeared from the 37° plates after the first week of holding, appeared on the 20° plates beginning with the first week and on the 0° plates beginning with the second week. The opposite ten- dency—to disappear entirely from the milk during holding—was exhibited by Streptothrix farcinica which was found only on the 37° plates of the fresh sample; it failed to develop on the 20° and 0° plates of that sample and was never found during the subsequent analyses.. THE BACTERIAL CHANGES IN THE CREAM DURING HOLDING. In the raw, untreated cream, on the plates incubated at 37° the total count per cc. increased progressively up to 14 days, then de- creased; the anaerobes and facultatives rose during the first week, then decreased; the acid formers underwent a progressive increase for 14 days, then decreased. On the plates incubated at 20°, the total count per cc. increased irregularly to a maximum at 21 days, followed by a decrease on the final analysis; the anaerobes and facultatives attained their highest value at 7 days, then exhib- ited a marked tendency to decrease progressively; the acid formers reached their highest values at 7 and 21 days, there being a trend toward a maximum during the middle of the experiment, then a decrease toward the close of the experiment; the liquefiers increased more or less regularly throughout the entire experiment. On the ite. 360 Changes in Milk and Cream at 0°C. TasLe VI. Bacterial cor (The figures in Roman represent bac | RAW UNTREATED CREAM PERIOD OF | PLATES ‘ KEEPING INCUBATED An d ; IN DAYS | at °C = count Fucutatves anes Formers ——— . 37 | 5,900,000! 1,400,000! 550,000 ; 1 1 1 Fresh 20 6,100,000 1,700,000 1,800,000 320,000 plates incubated at 0°, the total count per cc., anaerobes and faculta- tives, and acid formers all showed a distinct tendency 19,000,000 59.4 progressively during the entire period of holding. In the reinfected, sterile cream, on the plates incubated at 37°, the total count per ce., anaerobes and facultatives and acid formers rose progressively up to 14 days, then decreased on the final analy- to increase r , R eae | Pennington and Collaborators 2,000,000 59.7 361 in cream kept at 0°C. n Italic represent ratios of increase.) REINFECTED STERILE CREAM hace: pnmcigcad Anaerobes and Facultative wget roomate Themes Total count 15,000 38,000 ~ 160,000 1 1 650,000 750,000 33,500 0 1 1 1 2,600 | 5,600 0 15,000,000 {> sis. On the plates incubated at 20°, the total count per cc., acid formers and liquefiers rose more or less progressively to their highest values on the concluding day of the experiment; the anaerobes and facultatives rose progressively during the first fortnight of holding, then suffered an enormous decrease during the third and last week in storage. On the plates incubated at 0°, the total count per cc., 362 Changes in Milk and Cream at 0°C. anaerobes and facultatives and acid formers rose with more or less regularity to maximum values on the twenty-first and ies. day of the experiment. In the fresh formolized, raw cream, organisms were present, rae subsequent analyses demonstrated that the cream had become sterile. The sterile cream retained its sterility throughout the period of holding, as was demonstrated by platings made on the initial and final analyses. Since the initial counts for all groups of organisms differ widely in the raw untreated and reinfected sterile cream, comparison of the nts themselves at different periods of the experiment cannot be e. A comparison of the ratio of increase of the various groups of organisms in the two kinds of cream, however, will be made during the discussion of those ratios (see page 363). As a rule, the optimum temperature of incubation for all groups of bacteria remained at 20° throughout the entire Period of study. abe Ratio of increase of the organisms. at The maximum count obtained during the period of holds each group of organisms at each temperature of incubation is always compared with the count obtained with that group of organisms — at the same temperature of incubation at the beginning of the experiment, using the latter figure as unity in the ratio. Raw, UNTREATED Cream. Total count per cc. At 37° the maximum ra- tio of increase was attained at 14 days and was 9.8 times the count on the fresh cream. At 21 days the maximum ratio at 20° (1:19.7) and at 0° (1:54.5) was reached. Anaerobes and facultatives. The maximum increase at 37° (1:3.6) and at 20° (1:17.1) occurred at 7 days, while the maximum for 0° (1:214.3) was attained at 21 days. Acid formers. At 37° the maximum was reached at 14 days with the ratio 1:36.4, while at 0° the maximum was at 21 days with the ratio 1: 21.7; the maximum ratio at 20° was 1: 40,0 at 7 days, and was closely followed by the ratio 1:38.9 at 21 days. Liquefiers reached their highest value at 28 days with a ratio of increase 1:59.4. Reixrecrep, Srerine Cream. Total count per ce. The maximum in- crease at 37° (1:22.4) oceurred at 14 days, the maximum at 20° (1: 100.0) and at 0° (1:9280.8) took place at 21 days. Anaerobes and facultatives. The maximum ratio of increase at 37° (1:266.7) and at 20° (1;24.6) was attained at 14 days, while the maximum at 0° (1: 567.9) was at 21 days. Acid formers. At 37° the greatest increase (1:236.8) was attained at 14 days, while the maximum increase at 20° (1:74.7) and at 0° (1:5371.1) was found at 21 days. The greatest ratio of inerease for liquefiers wus 1;835.8 at 21 days. —— Pennington and Collaborators 363 The organisms which proliferate best at 37° always reached their highest numbers by the fourteenth day of holding, while the maxi- mum ratio of increase of the organisms which grow best at 20° and at 0° almost invariably occurred on the twenty-first day. Upon comparison of the ratio of increase in the raw untreated cream at each period of analysis and at each temperature of incu- bation for each group of organisms with the similar ratio in the reinfected sterile cream, it is seen that, in 25 of the 27 cases, 92.6 _ per cent, the ratio is higher in the reinfected cream than in the raw cream. Therefore, at 0°C., the reinfected cream has been a more suitable medium for buaterial reproduction and has given rise to a greater rate of growth. ae The bacterial flora of the raw untreated cream 2 of th2 reinfected senile cream, — The mode of tabulation of the organisms is the same as was used for the bacterial flora of the milk. In the fresh raw, untreated eream,on the plates incubated at 37° and at 20°, Micrococcus auran- —«tacus: (Cohn) predominated, while on the plates incubated at 0° Micrococcus ovalis (Escherich) occupied that position. During _ the period of holding, on the plates at 37°, M. ovalis predominated until the last analysis when it gave place to M. aurantiacus; on the plates at 20°, M. ovalis occupied the first position but was displaced by M. aurantiacus during the last two weeks of the ex- periment; on the plates at 0°, M. aurantiacus was the predominant _ organism throughout the entire period of holding. The first and _ second positions were almost invariably held by the two acid- forming micrococci, M. aurantiacus and M. ovalis; the only excep- tion occurred at 14 days on the plates incubated at 20° where Bacterium aerophilum occupied second place. On the whole the storage at a temperature of 0°C. exerted a favorable influence on M. aurantiacus and an unfavorable influence on M. ovalis, for the _ former gradually displaced the latter even on the plates grown at 37°. The disappearance of certain organisms during holding was illustrated by M. acidi lactis and Bacillus detrudens, which were isolated from plates sown with fresh cream and incubated at 37°; these organisms, however, were never found in the cream during the period of holding. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3 Changes in Milk and Cream at 0°C. TABLE VII. 364 Organisms isolated in pure culture from raw untreated and reinfected sterile cream, kept at 0°C. of anal ys PERIOD IN | pow THE woupawe Fete ews RAW UNTREATED CREAM | __—s—iREINFECTED STERILE CREAM CREAM ar 0°C. at °C. M. acidi lactis **M. aurantiacus (Cohn) **M. aurantiacus (Cohn) | *M. ovalis (Escherich) 37 *M. ovalis (Escherich) B. detrudens (Wright) B. flexuosum (Wright) S. Rosenbachii (Kruse) B. detrudens (Wright) S. Rosenbachii (Kruse) e Rech 1 M. acidi lactici | *M. aurantiacus (Cohn) ___ |**M. aurantiacus (Cohn) **M. ovalis (Escherich) 20 =| *M. ovalis (Escherich) B. aerophilum rophilum B. flexuosum (Wright) 4 : nt) | S. farcinica 8. flava S. Ros *M. aurantiacus (Cohn) |**M. auran (Co 0 |**M. ovalis (Escherich) *M. ovalis (Escherich B. aerophilum — B. aerophilum " B. flexuosum (Wright) B. flexuosum (Wright) M. alvi M. acidi lactici *M. aurantiacus (Cohn) | M. alvi 37 |**M. ovalis (Escherich) | *M. aurantiacus (Cohn) 8S. Rosenbachii (Kruse) “M. ovalis (Escherich) S. Rosenbachii (Kruse) *M. aurantiacus (Cohn) | M. aurantiacus (Cohn) 7 20 |**M. ovalis (Escherich) (**M. ovalis (Escherich) B. flexuosum (Wright) | *S. flava 8. flava **M. aurantiacus (Cohn) |**M. aurantiacus (Cohn) 0 *M. ovalis (Escherich) *M. ovalis (Escherich) B. aerophilum B. aerophilum 8. flava * * The predominant speciesof organism at each temperature of incubation and at each period * The second most numerous 4 “of organiam at each temperature of incubation and at each period of analysis, cama 4 ws i " a, Pennington and Collaborators TABLE ViI—Continued. B. flexuosum (Wright S. flava PERIOD IN FROM THE Resign cyl RAW UNTREATED CREAM REINFECTED STERILE CREAM at 0°C at °C. *M. aurantiacus (Cohn) |**M. aurantiacus (Cohn) **M. ovalis (Escherich) *M. ovalis (Escherich) 37 B. aerophilum B. aerophilum B. flexuosum (Wright) B. flexuosum (Wright) 8S? flava *M. aurantiacus (Cohn) | *M. aurantiacus (Cohn) **M. ovalis (Escherich) |**M. ovalis (Escherich) 14 20 B. aerophilum B. aerophilum **M. aurantiacus (Cohn 98 Burantiacus (Cohn) **M. ovalis (Escherich) B. aerophilum M. acidi lactici **M. aurantiacus (Cohn) M. ovalis (Escherich) "*B. aerophilum S. Rosenbachii (Kruse) **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum B. flexuosum (Wright) 8. flava **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum B. flexuosum (Wright) **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum B. flexuosum (Wright) 20 **M. aurantiacus (Cohn) *M. ovalis (Escherich) B. aerophilum of analysis. each period of ** The predominant species of organism at each temperature.of incubation and at each period * The pee jeorp most numerous species of orgavism at each temperature of incubation and at 365. r a 366 Changes in Milk and Cream at 0°C. In the reinfected, sterile cream, on the plates at 37°, Micrococcus aurantiacus (Cohn) predominated in the fresh sample and held that position throughout the entire experiment save. at 7 days, where it gave place to Micrococcus ovalis (Escherich). On the plates at 20°, M. ovalis was the predominating species save on the final analysis when it was displaced by M. aurantiacus. On the plates at 0°, M. aurantiacus held the first position throughout the . entire period of holding. At all three temperatures of incubation, the first and second positions were usually shared by M. ovalis and M. aurantiacus; however, the second position at 21 days on the lates at 37° was held by Bacterium aerophilum, and at 7 days on plates at 20° by Strepthothriz flava. The trend for some organ- isms to disappear during holding was illustrated by Bacillus de- trudens, which was isolated from the plates sown with the fresh sample and incubated at 37°; at mapecquent periods of analysis, the organism was never found. GENERAL CONCLUSIONS. In milk and cream which are held at a temperature Le on the following phenomena are noted: The proteolysis of the casein is, primarily, of bacterial origin. The proteolysis of the lactalbumin is due, primarily, to native enzymes of the milk. The bacterial flora and the native milk enzymes by their com- bined action give rise to more rapid proteolytic changes than are produced by either agent alone. The general trend of the proteolysis by bacteria, by enzymes, and by the combined action of these two agents, involves a breaking down of the true proteins and their passage through caseose and peptone to amino-acids. The fermentation of the lactose with the formation of lactic acid is largely, if not exclusively, due to bacterial action. The digestion of the protein, the fermentation of the lactose and the increase in acidity are progressive changes, and are ac- companied by more or less progressive lowerings of the freezing point of the milk. The depression of the freezing point of the cream is to be ascribed to chemical changes in its protein and lactose. ee 7 Pennington and Collaborators 367 The lecithin of the cream was not decomposed during the period of holding. The iodine number and the index of refraction of the butter-fat remained unchanged, while the Reichert-Meissl number under- went no marked increase or decrease in any of the samples. The hydrolysis of the fat and the increase in acid value, which is usually progressive, are due to the action of bacteria. _ The Hehner number always becomes greater; the saponification number usually increases, although it underwent progressive decreases in the reinfected sterile cream. The increase in Hehner number, accompanied by a decrease in saponification number, in | q the reinfected sterile cream is to be ascribed to bacterial action. — : : The simultaneous increase in the two constants in the ste _ cream is doubtless due to oxidation and the fine state of division of _ of the butter-fat in the cream, possibly aided by a thermostable inorganic catalyst; these same causes, plus the action of the native enzymes, must give rise to the simultaneous increase in the Hehner and saponification numbers of the formolized raw cream; while the simultaneous rise in these two constants in the raw untreated cream are the resultant of the action of bacteria, native enzymes, - oxidation, inorganic catalysts and fine state of division of the butter-fat. The simple reductases and ae trikresol oxidases of the milk may be native enzymes and may also be of bacterial origin, while the aldehyde reductase and the guaiac oxidases, apparently, are of bacterial origin. Both varieties of reductase and the trikresol oxidases are native enzymes of the cream and are also secreted by the microérganisms, while the guaiac oxidases probably have their origin in bacterial action. The catalase of both the milk and the cream is a native enzyme and is also due to the activity of microérganisms. The two varieties of oxidase, the two varieties of reductase and the catalase retain their activity in spite of the prolonged exposure to a temperature of 0°C. During the holding at 0°C., the organisms of the raw untreated and reinfected sterile milk and cream undergo an increase, which is most striking in the raw untreated milk. 368 Changes in Milk and Cream at 0°C. , In both milk and cream, the organisms which proliferate best at 37°C. reach a maximum growth during the earlier stages of the period of holding, while those growing best at 20°C. and at 0°C. continue to increase and attain their highest values during the later stages of the experiment. Some organisms disappear during holding, others become accli- mated to the lower temperatures of incubation. The total count per cc. at 37° was practically the same in the raw untreated and reinfected sterile milk; in over 90 per cent of the subsequent determinations the rate of increase of the various ups of organisms was greater in the raw than in the reinfected nple. In over 90 per cent of the determinations made during the period of holding, the rate of increase of the various groups of organisms was greater in the reinfected sterile cream than in the raw untreated cream. Oe _ Almost invariably Micrococcus aurantiacus (Cohn), and Micro- coccus ovalis (Escherich), which belong to the group of acid-formers, were the predominant organisms of both raw untreated ope rein- fected sterile milk and cream. ae e. THE REACTION OF SOME PURINE, PYRIMIDINE, AND HYDANTOIN DERIVATIVES WITH THE URIC ACID AND PHENOL REAGENTS OF FOLIN AND DENIS. By HOWARD B. LEWIS anp BEN H. NICOLET. _ (From the Sheffield Laboratory of Physiological Chemistry and the Sheffield _ Chemical Laboratory, Yale University, New Haven, Conn.) gm ; (Received for Sablisation, October 10, 1913.) * In the course of an investigation by one e of on the behavior of certain thioderiv of nydant aie in the animal organism, it was observed th ‘the substances studied reacted with the ‘ reagents of Folin and Denis? with the development of the blue ; color, bed as typical for uric acid. The possibility of using _ this reaction for the detection of the compounds under investi- " tion at once suggested itself, and an examination of various related substances in regard to their behavior toward the new reagents was undertaken. Recently Funk and Macallum* have studied the reaction with certain purines, pyrimidines, and related substances of -biochemical importance. Inasmuch as our results supplement and extend the work of these investigators, we offer them in the hope that they may prove of value in pointing out a line of attack for the solution of the problem of the chemical basis of the reactions. For many of the compounds studied we are indebted to Profs. Lafayette B. Mendel and T. B. Johnson, ~ who have placed them at our disposal. With the exception of thiourea no substance was observed to react typically with the phenol reagent, which does not contain a phenol group or react with the uric acid reagent. N-c-methyl- tyrosine reacts positively with the phenol reagent, but the related 1 Lewis: this Journal, xiv, pp. 245-56, 1913. 2 Folin and Denis: this Journal, xii, p. 239, 1912. § Funk and Macallum: Biochem. Journ., vii, pp. 356-58, 1913. 369 Color Reagents of Folin and Denis 370 a TABLE I. URIC ACID PHENOL REAGENT REAGENT a eeyaentoin. ....... ... sss eee es — - 2. Hydantoin-4-acetic acid..................... _ _ 3. Hydantoin-4-propionic acid.................. - - 4. p-Hydroxybenzylhydantoin........... J Rp _ choo 5. p-Aminobenzylhydantoin.................... _ _ G6. Parabanic acid... .... .....s:. sae sles ss - _ 7. Aminocarboxyhydantoin..................... +++ seopeb SB. 2-Thiohydantoin. ..3:..... sae ss +p te cee 9. 2-Thio-4-methylhydantoin........ TS he +++ en ee ACID ieee as 6... +++ oe eR 2-Thio-4-benzalhydantoin.................... - — 12. 2-Thio-4-benzylhydantoin Res, seb +44. 13. 1-Phenyl-2.- -- - 14. 1-Phenyl-2-thi rdantoin............ a Be +++ 15. 1-Phenyl-d-thio 4 oti i sine - - 16. 1-Phenyl-2-thio-4-furfuralhydantoin..........). — “ 17. 1-Phenyl-2-thiohydantoin.................... +-+- 18. 3-Phenyl-2-thiohydantoin.................... ++- 19. 1,3-Diphenyl-2-thiohydantoin...... og aie a ii 20. 1-Phenyl-2-thio-4-piperonalhydantoin........ _ Purine derivatives. 21. 2 3ibiypuriie Cie css 6s OES o's s SMEG | = 22. 2 ‘S:Dhoxy-Goanbibii juries ene, Age SSL = 23. 2,8-Dioxy-6,9-dimethylpurine................. = 24. 2,8-Dioxy-9-methylpurine.................... a 25. 2,8-Dioxy-1,9-dimethylpurine................. - 26. 2-Oxy-9-methylpurine...................00005 — 27. 2-Oxy-6,9-dimethylpurine.................... 0) _ 28. 2-Oxy-6,8,9-trimethylpurine.................. - 29. 2,6-Dioxy-8-methylpurine...................08 _ 30. 2,6-Dioxy-1,7-dimethylpurine................. _ 31. 2,6-Dioxy-1,3-dimethylpurine.. ....4.,........ $2. 2,6-Dioxy-1,3,7-trimethylpurine... fh Agnes « «| = 33. Sodium uroxanate.................... aan... ~ SH aT hio-G-Oxypurine. dy... .... seats. scales... be +44 35. 2,8-Dithio-G-oxypurine................0.000055 44 -f +++ 36. 2-Thio-6,8-dioxypurine...............6..0005. +++ +++ 37. 2-Thioglycollic acid-6-oxypurine............. - _- 38. 2,8-Dithioglycollic acid-6-oxypurine...... - ao 39. 2-Thioglycollic acid-6,8-dioxypurine., . .. te + Howard B. Lewis and Ben H. Nicolet 371 TABLE I.—Continued. Pyrimidine derivatives. URIC ACID PHENOL REAGENT REAGENT ES Ri ae - ES AO ete - MT MEOEDVICVUOSIDG,,...-.+.....-......0¢eennue — . ES SES LATS ba — 44, 1-Methylthymine.........................00% _ 45. 1,3-Dimethylthymine.. . «sa - 46. Diethylbarbituric acid Barons!) Ss ve ce - MPTeeeeUAUTINGG ADIG, (..,03..>--....-.5..-ce0 eee - a NUS SE 7 ap A _ . a0) eer mioviolurig acids... ............ 0. 0b age +++ (eet + 50. Cyanacetylguanidine..............°......... - 51. 2-Phenylamino-6-oxypyrimidine............ BS . = 52 ae 53 55. Melongiediiil a oS ae | _ 56. 5-Aminomalonylguanidine hydrochloride..... | eee) | 57 -2-Thiopseudouric Es Paras. . «2c 8 open Fe so is Ue a ota ait ES URES APPL SS Phe [so +++ Miscellaneous compounds. SS ES 0S a | Se oe tp es, WS; . . oelecas see -s case - t+ EE + | a B 62. N-methyldiiodtyrosine....................... | a e+ Di 63. N-i-methyltyrosine.......................65. — +++ j 64. a-Methylamino-$-p-methoxyphenylpropionic 4 acid. ..... |< IEE. « - RRs) «+ alg | - - H 85. Glycollic acid... ss... eee Le — - mee Gemmllantoin........0.......-...- Ns SOM es ss - — vy 67. Malonaminourethane...................... ine 4+4+ +44 68. Dithiodimethylpiperazine.................... | +++ ++ Mu temmoylithiourea. se... ...0......05.4...... a —_ 70. Benzoylphenylthiourea..................... | o ~ a-methylamin oparamethoxyphenylpropionic acid in which a methyl group has been substituted for the hydrogen of the hydroxy] group fails to give the reaction. N-methyldiiodtyrosine, which does not respond to Millon’s test, gives a positive reaction with the 372 Color Reagents of Folin and Denis phenol reagent, indicating that it is not a necessary condition for this reaction as for Millon’s that the position ortho to the hydroxy] group in the benzene ring shall be unsubstituted. Thiotyrosine reacts positively with the phenol reagent, but inasmuch as it also reacts positively with the uric acid reagent and as will be dis- _ cussed later sulphur seems to play a réle here, no conclusion as eto the effect of —SH groups replacing —OH groups can be drawn. Of the cyclic compounds examined which contained no sulphur, only three, 4-aminocarboxyhydantoin, 2,4,5-triamino-6-oxypyri- midine, and 5-aminomalonylguanidine react positively with the acid reagent. The striking feature possessed by these three mpounds in common, in contrast with the closely related hydan- toin, 2,4-diamino-6-oxypyrimidine, and malonylgyanidine, none of which gives a positive reaction, is the presence of an amino group. In 2,4,5-tri 0 imidine and 5-malonylguan- idine, the amino group which ddeibidines the reaction is in the 5-position as is evidenced by the inactivity of the 2,4-diamino compound and malonylguanidine respectively. It is int eresting to note that malonaminourethane of which 4-aminocarboxyhy- dantoin is the cyclic anhydride is the only acyclic compound - free from sulphur which reacted positively with the uric acid be reagent. The activity of amino groups is in harmony with the ~ observations of Folin and Denis and Funk and Macallum, that monohydrie phenols do not react with the uric acid reagent unless an amino group is present in the benzene ring (e. g., 2- or 3-amino- tyrosine). With the exception of the thiopurines, none of the purines studied react positively. Neither xanthine nor its isomer, nor any of their methyl derivatives react. Allantoin and uroxanic acid, both oxidation products of uric acid, fail to react. The presence of sulphur replacing oxygen in the 2-position of a purine, pyrimidine, or hydantoin, gives rise to a positive reac- tion with both reagents (cf. 8-20, 34-36, 49, 57, 58). In the case of the condensation products of thiohydantoins with aldehydes, derivatives in which the carbon in the 4-position is unsaturated (11, 13, 15, 16, 20) the reaction is negative. When reduced the resulting products react positively (ef. 11 and 12, 18 and 14). That sulphur replacing oxygen rather than mercapto sulphur is a condition for a positive test is shown by the negative reactions Pe wc ae eo a a - compounds. Howard B. Lewis and Ben H. Nicolet 373 of the mercapto derivatives of the purines and pyrimidines (37- 39,52). Dithiodimethylpiperazine, the anhydride of the thiopoly- peptide, thioalanyl-thioalanine, reacts positively. Although thio- urea gives the test with the phenol reagent, neither of the two substituted thioureas (69, 70) gives a positive reaction. The suggestion made by Funk and Macallum that in the pu- rines the substitution of the hydrogen atoms of the ring lessens or destroys the power to react with the uric acid reagent, does not hold for the hydantoin ring. Those thiohydantoins in which substitution has occurred on the 1- or 3-carbon atom or on both (17-19), react as readily with the reagents as do the unsubstitujeds | ax a = uM » COse. THE FORMATION OF GLUCOSE FROM PROPIONIC ACID IN DIABETES MELLITUS. . By ISIDOR GREENWALD. (From the Chemical Laboratory of the Montefiore Home, New York.) (Received for publication, October 15, 1913.) It has been shown by Ringer! that the administration of pro-). | ‘ pionic acid to phlorhizinized dogs is followed by the elimination — of “extra glucose” equal in amount to that capable of being formed from the propionic acid if all three carbon atop are used in the formation of glucose. . Shortly after the publication of this Work we had under obser- vation a patient who exhibited a G : N ratio of about 3.5. The case seemed to be eminently suitable for the comparison of glu- ation in phlorhizin glycosuria and in diabetes mellitus. ic acid being readily obtainable it seemed desirable to | kite ascertain if it would give rise to glucose. This was found to be the case. The experiment was subsequently repeated upon the . same patient and also upon two others with less severe forms of diabetes. In one patient, who had only a slight diabetes, the administration of propionic acid did not increase the excretion of glucose. Otherwise, a distinct rise was observed in every ex- ' periment. The excretion of acetone and §-hydroxybutyric acid was followed in some of the experiments. In only one of these was there an increase in the amount of these substances elimi- nated and in this case there was a further rise on the following day. Four typical experiments are summarized in the accom- - panying tables. The propionic acid was prepared by Kahlbaum. The analyti- cal methods employed were the Kjeldahl for nitrogen, the Bene- dict for glucose and the Shaffer for acetone and 6-hydroxybutyric acid. The figures for the carbohydrate in the diet were calcu- lated from the data given in Bulletin 28, U. 8. Department of Agriculture. 1 Ringer: This Journal, xii, p. 511, 1912. 375 376 Glucose Formation from Propionic Acid As is evident from the tables, in the first experiment the “extra glucose”? was almost exactly equivalent in amount to the pro- pionic acid ingested. Later when the G:N ratio was lower, indicating an increased capacity for the oxidation of carbohy- drates, the amount of glucose formed from the propionic acid administered was much diminished. The second patient, whose utilization of carbohydrates was much greater, excreted much less glucose as a result of the ingestion of propionic acid. TABLE I. Patient J. L. grams GLUCOSE” 22 | 16.0 glucose.) 23 «| «15.6 | 14.61)104.1 | 88.5 | 6.06 | 25.2 | 4.00*|22.21*| 20.0 gms. pro- pionie acid (equivalent to 24.3 gms. glucose.) | The experiment was discontinued because the patient ate other food. July | 20 | 17.2| 6.93 | 34.8 | 17.6 | 2.53 | | 21 | 22.9) 8.74) 44.5 | 21.6 | 2.47 | | 22 | 19.6 |10.37 | 56.7 | 27.1 | 3.58 | | 23 | 17.4| 9.23 | 43.8 | 26.4 | 2.86 1.10 | 7.62 | a | 13.0|7.72| 58.8| 45.8 | 5.92 | 23.9| 1.43 11.41 [36.65 grame propionic acid (equiv- alent to | | _ 44.57 grams | glucose). 25 | 19.0 | 9.85 | 45.3 | 26.3 | 2.60 | 2.19 14.44 * Analyses made of two-day composites, Midor Greets TABLE IL. ». Patient G. E. THE ACTION OF RADIUM EMANATION ON LIPASE. By E. K. MARSHALL, Jr. anp L. G. ROWNTREE. (From the Laboratories of Physiological Chemistry and Pharmacology of the is Johns Hopkins University.) (Received for publication, October 17, 1913.) ‘ It is claimed that both radium rays and radium emanation 7 _ possess the power of activating certain enzymes. A —— - —" . erating influence has been shown for pepsin by the tion (Bergell and Bickel’), for autolytic ferme by rays (Neu- . berg,’ Wohlgemuth*), and by the ema n (Lowenthal and Edel- _ stein‘), for tyrosing by the radium rays (Willcock®), for diastase from various origins ( Léwenthal and Wohlgemuth’), and the uric ~ acid forming ferments of the spleen by the emanation (Schultz’). O e other hand, a.slight inhibitory effect was observed by Sch dt-Nielsen® for rennin by exposure to strong radium prep- ‘arations. Trypsin, invertase, and emulsin are reported by Henri and Mayer® to be inactivated through long exposure to the rays. The possibility of the increased rate of growth in plants'® (ger- minating oats, for instance) being due in part to enzymatic acti- vation has been suggested. The idea has been entertained that the therapeutic effect of radium treatment in gout is dependent on this cause. On account of the ease and accuracy of determin- _ ing quantitatively lipolytic activity, the effect of the radium ema- nation in this connection has been investigated. ia i ¢ 1 Verhandl. der Kongr. fiir inn. Med., Wiesbaden, 22d Kongress, 1905, p. 157. 2 Verhandl. d. deutsch. path. Gesellsch., vii, p. 157, 1904. 8 Ibid., vii, p. 158, 1904. 4 Biochem. Zeitschr., xiv, p. 484, 1908. 5 Journ. of Physiol., xxxiv, p. 207, 1906. ® Biochem. Zeitschr., xxi, p. 476, 1909. ? [bid., xlviii, p. 86, 1913. 8 Mitt. a. Finsen’s Med. Lysinst. in Kopenh., Jena, 1906, 10 Heft, p. 107. ® Compt. rend. de UV’ Acad. des Sci., exl, p. 521, 1904. 10 Falta and Schwartz: Berl. klin. Wochenschr., xlviii, p. 605, 1911. & ; 379 é THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3. 380 Action of Radium Emanation on Lipase The experiments have been carried out with the lipase of pig’s | liver and the lipase of the castor oil bean. In the case of liver, a 10 per cent “brei’’ was made, allowed to digest over night, and portions of the clear filtrate used in the experiments either diluted | or undiluted. Different liver preparations were used. With the | castor bean lipase, the beans were finely ground, extracted with ether in a Soxhlet and weighed portions of this powder used. In- some of the experiments with the liver lipase, a saturated aqueous solution of ethyl butyrate was used as substrate, in all other cases 0.5 cc. of ethyl butyrate and 25 cc. of water as the solvent ail a been used. The mixtures were allowed to digest for a defi- 4 ‘ite time and then titrated with 0.116 N barium hydroxide using 5 or 6 drops of a 1 per cent alcoholic solution of phenolphthalein | as indicator. Toluene was used to prevent bacterial growth. The radioactive solutions w: B prepared as follows: Radium emanation was collected in Way over mercury in a test tube, the amount present being estimated by y-ray electroscope. This was transferred to a flask containing water, thoroughly shaken for a period of 20 minutes and the amount of emanation per cc. calculated. Varying amounts of this water were added to the lipase-ethyl butyrate mixtures in the amounts indicated." The final volume of the mixture was made up to that of the control. The tubes or flasks used were at least half filled with liquid and tightly stoppered with well-fitting corks. The results of the experiments can be found in the following | tables. The activity of the enzyme is expressed in ce. of 0.116. N barium hydroxide necessary to neutralize the butyric acid formed. | The values have been in all cases corrected for the acidity devel- oped for the ethyl butyrate and lipase extract alone. The data of the following tables show conclusively that no accel-_ erating influence is exerted upon the lipase of the pig’s liver or castor oil bean by radium emanation in the amounts used. On the contrary, inhibition of the enzymatic activity is suggested. We acknowledge with pleasure our indebtedness to Dr. H. H. Young, through whose generosity the radium was placed at our disposal. , “Inasmuch as the flasks were not completely filled with liquid, a large proportion of the radium emanation would eventually be found in the air_ space above the solution. E. K. Marshall, Jr. and L. G. Rowntree 381 TABLE 1. Ten cc. saturated ethyl butyrate solution, 3 cc. undiluted liver extract, with either 2 cc. water or 2 cc. emanation water. (room temperature). cc. 0.116 wn BaA(OH)2 REQUIRED MICROCURIES TIME IN EMANATION MINUTES Without R. E.| With R. E. 46 20 1.96 1.90 4 30 2.35 2.36 18 30 a 2.40 92 60 2.75,-" 2.80 18 70 3 28 3.30 — ; 18 75 3.45 3.35 gis 4 85 3.60 3.50 = : TABLE 2 a * Ten cc. saturated with either 1 cc. wate butyrate solution, 2 cc. undiluted liver extract, *1 cc. emanation water (room temperature). rs ps : co, 0.116 w BA(OH): REQUIRED i! ci TIME IN MINUTES Without R.E.| With R. E. = 32 1.80 1.76 45 2.30 2.26 83 3.72 3.52 83 3.81 3.50 TABLE 3. | Ten cc. saturated ethyl butyrate solution, 2 cc. undiluted liver extract, 2 ce. water or 2 cc. emanation water. ik cc. 0.116 n Ba(OH): REQUIRED * TIME IN * MINUTES Fe Without R.E.| With R. E. ‘ 25 0.36 0.82 102 1.98 1.65 161 2.20 1.70 161 2.30 2.50 230 2.80 2.07 & 230 3.00 284 3.20 2.62 382 Action of Radium Emanation on Lipase TABLE 4. Ten cc. saturated ethyl butyrate solution, 2 cc. undiluted liver extract, and 5 cc. water or 5 cc. emanation water allowed to digest 5 hours at room temperature. cc. 0.116 n Ba(OH)2 REQUIRED Without R.E.| With R. E. 3.98 4.22 4.31 4.21 4.16. 3.42 4.05 4.13 : 2 | | TABLE 6. 0.5 ee. ethyl butyrate, 1 cc. liver extract (diluted with three volumes of - water) and 25 ce. water or 25 cc. emanation water digested 44 hours at 38°. co. 0.116 N PHOS REQUIRED MICROCURIES | _ a fr? EMANATION Without R.E.| With R. EB. er peel . 3 6.70 5 00. 3 6.75 5.10 3 6.60 5.10 3 6.60 5.00 0.17 6.75 5.90 017 5.80 0.02 6.50 0.02 6.70 0,002 6.70 0,002 6.65 E. K. Marshall, Jr. and L. G. Rowntree 383 TABLE 7. 0.5 cc. ethyl butyrate, 1 cc. liver extract (diluted three times), 25 cc. water or 25 cc. emanation water. Each flask contained 150 microcuries. cc. 0.116 nN BaA(OH)2 REQUIRED TIME IN rie to ot ars HOURS Without R. E.| With R. E. 22 4.65 4.30. 22 4.80 4.20 48 6.25 5.70 48 6.10 5.80 TABLE 8. 0.5 cc. ethyl butyrate, 0.2 gm. castor bean, and 28 emanation water. 25 cc. contained 150 microcuri 5 0.116 N Ba(Ol OH): 5 Without R. E.| With R. E. 1.70 1.65 1.65 1.65 2.80 2.70 2.80 2.70 TABLE 9. 0.5 cc. ethyl butyrate, 0.5 gm. castor bean, and 25 ce. water or 25 ec. emanation water, digested 18 hours at 38°. y cc. 0.116 nN Ba(OH)2 REQUIRED 8 MICROCURIES ec : EMANATION i= : Without R. E.| With R. E. "i 3 3.75 3.70 Hi 3 3.75 3.65 ay 150 3.70 iP 150 3.65 a eS pecs on aaa and 25 ce. water or 25 cc. Fises baad ec. contained 80 microcuries. NOTE ON THE DETERMINATION OF AMINO-ACID NITROGEN IN URINE. By S. R. BENEDICT anp J. R. MURLIN. (From the Laboratories of Physiology and of Physiological Chemistry of t Cornell University Medical College, New York City.) (Received for publication, October 18, 1913.) _ Something over a year ago the writers! reported a modified technique for the preparation of urines for the formalin titration method of Henriques and Sérensen. This consisted entially in the removal of ammonia and other bodies eans of phos- i _ photungstic acid, the clearing-out of the ple hotungstic acid _ by means of tribasi¢: acetate and litharge and the final pre- ; cipitation of the lead by a stream of hydrogen sulphide. The _ water-clear filtrate was then neutralized to litmus and titrated to the third stage end-point of phenolphthalein as directed by m original authors. The results by this method in comparison with those of the “new” method of Henriques and Sérensen? were very much lower. - With certain pure substances added to urines values very close _ to the theoretical were found. Subsequently the list of pure substances was extended and it was found that a number of amino-acids were removed from the solution at times in whole or in part by the lead. This was no- tably true of aspartic acid and tyrosine, less so of glutamic acid and leucine. Much depended on the amount of basic lead ace- tate employed. After a time, however, it became apparent that it would be impracticable to control the quantity of lead and the conditions of temperature, etc., accurately enough to make the method of much value, and a different method of removing the phosphotungstic acid was sought. van Leersum* had employed KCl for this purpose in the amino-acid method which he devised. ! Benedict and Murlin: Proc. Soc. of Exp. Biol. and Med., ix, p. 109, 1912. * Zeitschr. f. physiol. Chem., lxiv, p. 120, 1909. § van Leersum: Biochem. Zettschr., xi, p. 121. 4 385 y 386 Determination of Amino-Acid N in Urine as a modification of the Pfaundler method. The first trials with the method as prescribed by van Leersum were not very suc- cessful. Traces of phosphotungstic acid (Merck) could always be found by the zinc test. Later it was found that the 5 per cent KCI solution would remove all of the acid provided a considerable excess of acid was left unprecipitated, and when this was not the case the addition of a little 10 per cent phosphotungstic acid in 2 per cent HCl would bring about complete precipitation. Under these circumstances also KCl could be used in substance and a water-clear filtrate could be readily obtained. Potassium salts owever do not remove the phosphates and sulphates; conse- ntly the attainment of an exact neutral point to litmus is very ‘difficult. Going back to barium hydrate as a means of removing the phosphates and sulphates it was found that the barium would suffice also to precipitate the phosphotungstic acid. The whole procedure thus became very simple and as used in these labora- tories now is as follows: eet PROCEDURE. ee 1. Measure into a 500-ce. Erlenmeyer flask 200 ce. of a Mt hour human urine diluted to 2000 ce. 4 2. Add an equal quantity of 10 per cent phosphotungstic acid (Merck*) in 2 per cent HCl. Let stand at least three hours; better over night. , 3. Pour off 250 ce. of the clear fluid; add 1 cc. of a 0.5 per cent solution of phenolphthalein, and barium hydrate in substance until the whole fluid turns decidedly pink. The barium hydrate should be added a very little at a time. Let stand one hour. 4. Filter off two 100-ce. samples (= 50 ce. urine). 5. Neutralize to litmus (Squibb’s papers answer for all prac- tical purposes) with # HCl. 6. Add 10-20 ce. ateal formalin and titrate cautiously to deep red color, z.e., until the drop produces no additional color with #; NaOH. ~ 7. Correct by deducting the amount of #; NaOH necessary to produce the same depth of color in an equal quantity of CO,- free water with the same quantity of neutral formalin added. Some control tests are given below. * Kahlbaum’s preparation is a very different substance. S. R. Benedict and J. R. Murlin 387 I. Removal of ammonia by means of phosphotungstic acid. a. Pure substances. 1. 40 cc. 75 solution of aspartic acid + 20 cc. 5 (NH,)280, solution. 100 cc. 10 per cént phosphotungstic acid solution in 2 per cent HCl added in equal quantity. Stood over night. 5-10 grams KCl added to remove phosphotungstic acid. 100 cc. filtrate titrated 10.1; theory, 10.0. 2. 20 cc. 7h glycocoll + 20 cc. 7 (NH4)2SO4. 100cc. Same procedure. 100 cc. filtrate titrated 9.5; theory, 10.0. b. Urines. 1. 200 cc. urine containing 4.6 per cent NH;-N + 200 cc. 10 per cent phosphotungstic acid in 4 per cent HCl stood « over night. 20 cc. filtrate aerated by Folin method three hours into 10 ce. 7y H:SO.. Titrated 10.0. Therefore all ammonia out. 2. 100 cc. urine containing 9.5 per cent NHy-N + 100 cc. 10 per cent — phosphotungstic acid in 4 per cent HCl. Stood over night. filtrate by Folin method three hours against 10 cc. 7) HS, titrated I 3.°200 cc. urine containing 9.5 per cent NH;-N + 50 ce. 10 per cent phosphotungstic acid in 4 per cent HCl. Stood Sallie. 20 ce. filtrate by Folin method four hours against 10 ce. oa ‘H.SO, titrated 6.0. Therefore ammonia not all » 2 4. 100 cc. urin taining 9.4 per cent NH;-N + 50 cc. 10 per cent phosphot stic acid in 4 per cent HCl. Stood over night. 20 ce. filtrate Folin method four hours titrated 5.0. : 5. 100 ce. urine containing 13 per cent NH;-N + 100 ce. 10 per cent phosphotungstic acid in 4 per cent HCl. Stood one and one-half hours. 200 ce. filtrate by Folin method eighteen hours titrated 9.9. These results agree with those of Gumlich® in proving that at least an equal quantity of the 10 per cent phosphotungstic acid solution must be added to the urine and that the mixture should stand at least three hours. II. Removal of phosphotungstic acid by KCl. 1. 100 ce. filtrate from urine (3) above + 100 ce. 5 per cent KCl. Stood two hours. Filtrate gives no blue color with zine. 2. 100 ce. filtrate from same urine + 10 grams KCl in substance. Stood two hours. Filtrate clear of phosphotungstic acid. 3. a. 200 cc. urine + 200 ce. 10 per cent phosphotungstic acid in 2 per cent H,SO,. b. 200 cc. same urine + 0.362 gram tyrosine + 200 cc. 10 per cent phos- photungstic acid. c. 200 cc. same urine + 0.266 gram aspartic acid + 200 cc. 10 per cent phosphotungstic acid. All stood for one week. Many crystals found on side of flask b and c. - Warmed in water bath adding 200 cc. distilled water. Crystals dissolved. Stood four hours. Decanted: 400 ce. clear fluid from each flask * 10 ans Ket. Btood two errs, 5 Gumlich: Zeitschr. f. Gees. Brom. xvii, p. 13, 1893. 388 Determination of Amino-Acid N in Urine 100 ee. filtrate from b titrated 8.5 c¢.; 100 ce. filtrate from c titrated 8.5 ec. 100 ce. filtrate from a titrated 5.3 cc., 5.3 cc. Difference b—a = 3.2 cc.: ¢ — a = 3.2 cc. Theoretical difference, 3.33 cc. III. Removal of phosphotungstic acid by Ba(OH):. 1. a. 200 ce. urine-(case of pernicious vomiting) + 200 cc. 10 per cent phosphotungstic acid in 2 per cent HCl. b. Same containing 0.262 gram leucine. Stood 4 hours, 200 ce. filtrate 4+ 50 cc. saturated solution Ba(OH)>s. Stood one hour. _ 100 ce. filtrate titrated 6.4 and 6.1 cc. ty NaOH. 100 ce. filtrate titrated 2.4 and 2.4 ce. 75 NaOH. Difference, 4.0 and 3.7. Theoretical differenee, 4.0. 2. a. 200 ce. urine + 200 cc. phosphotungstic acid. b. Same containing 0.326 gram tyrosine. Stood three days. Crystals of tyrosine found on sides of flask. Added few drops concentrated HCl and warmed in water bath until crystals dis- eC solved. : Phosphotungstic removed with Ba(OH). in substance while keeping flask warm. ; te b. 100 ce. filtrate titrated 9.9 and 10.0. ae a. 100 ce. filtrate titrated 4.8 and 4.9. et Difference, 5.1 and 5.1; theory, 5.0. "4 * ; Aspartic acid added in similar quantity was partially removed by the phosphotungstic acid or possibly by the barium hydrate used in too great concentration. Levene and Beatty® have shown that while amino-acids in general are not precipitated by phosphotungstic acid unless they are present in great concentration, there is considerable variation in this respect among the individual amino-acids. Because of the great insolubility of tyrosine, leucine, aspartic acid, etc., in neutral medium, it is important not to let the filtrate stand after the neutral point is reached. The neutral formalin should be added at once. 6 Levene and Beatty: Zeitschr. f. physiol. Chem., xlvii, p. 149, 1906. METABOLISM STUDIES ON COLD-BLOODED ANIMALS. II. THE BLOOD AND URINE OF FISH. By W. DENIS. - _ (From the Biochemical Laboratory of the Harvard Medical School, Boston, _ and the Laboratory of the U. S. Bureau of Fisheries, Woods Hole, Mass.) (Received for publication, October 20, 1913.) ¥ In a recent paper Folin and Denis! have published data regard- | ing the non-protein nitrogen, urea and uric a ‘in the blood of a number of ma s and a few birds. More or less as a con- - tinuation of the work I have collected the blood of a number of the more common. fish of the North Atlantic coast and in this blood have determined the total non-protein nitrogen, urea, am- monia, uric acid and creatine. For the determination of the first four constituents the methods _ recently published by Folin and Denis? have been used; creatine __ was determined colorimetrically by the Folin method, the follow- ing procedure being employed. Twenty grams of oxalated blood were poured slowly into 100 cc. of boil- ing ;}p acetic acid solution and the mixture heated for two or three min- utes until coagulation was complete; the solution was then filtered, the coagulum returned to the vessel in which the coagulation had been made and washed with about 200 cc. of boiling water. The original filtrate and the wash water were then combined, strongly acidified with acetic acid and rapidly evaporated down to a volume of about 5cc. This residue was transferred to a small flask, the evaporating dish being first rinsed with 10 cc. of normal hydrochloric acid and then with a few cubic centimeters of water. The mouth of the flask was then loosely closed and the mixture heated for four hours on the boiling water bath. The color was developed in the usual way, but as a standard I used a solution of pure creatinine’ in place of the customary half-normal potassium bichromate. 1 This Journal, xiv, p. 29, 1913. 2 Ibid., xi, p. 527, 1912; xiii, p. 469, 1913. 5 Ibid., xii, p. 149, 1912. 389 390 Blood and Urine of Fish The blood was in every case taken from fish, just brought in from the traps and therefore still alive; in the case of the larger specimens the blood was removed from the heart by means of a — needle and syringe, while in the case of the smaller animals it — was collected from the caudal artery and vein. All figures pre- sented are for whole blood in which coagulation had been pre- vented by the use of a little solid potassium oxalate. In the case of the larger fish (shark, goosefish, squeteague, etc.) not more — than six animals were employed to secure the composite sample of blood, while in the case of the smaller ones (butterfish, mack- ie 3 eel) from fifty to a hundred fish were used in order to obtain a requisite quantity of blood. | / { * Results off the examination of the non-protein nitrogen toaaein of the blood es of fish. (The figures rep: ¢ milli, per 100 grams of blood.) ais il |e Plghl | i tl i. \ a8 ee ee. B | ee | ge eee Dogfish (Mustelis COMAS) oi... Scbepaees ss ss 1000 800; 1.4) 0 . 4, Sand shark (Carcharias littoralis).........| 1160 | 1000} 2.5| 0 40 | Skate (Raia erniacea)..............06.5055 1100 | 868; 1.6) 0 cP a Alewife (Ponnolobus pseudoharengus)..... 54 10; 5.5} 1.1} 110 Butterfish (Poronotus triacanthus)......... 50 9| 5.1| 14) 160 Mackerel (Scomber scombrus).............| 86 10| 3.8) Lit tea Squeteague (Cynoscion regalis)...........| 66 20; 1.0| 0.7| 6.0 | Menhaden (Brevoortia tyrannus).......... 47 10| 3.3} Lot eee Summer flounder (Paralichthys dentalus)..; 46 8| 1.14 0:8 s.0c Shad (Alosa sapidessima)................ 90 16 1.1) Bonito (Sarda sarda) ............-.s00005 | 90 17:| 3.81 2am Goosefish (Lophius piscatorius)........... | 40} 8| 36] 0.9) 50 % Eel (Anguilla crysypa)........06.00scnee. 50 9; 28) 06;106 | ee = _ — ee —— 2 td From the figures given in the - table i it will be seen that by the new analytical method used I have in the case of the three elasmobranch fishes examined (shark, dogfish and skate) — given quantitative confirmation of the earlier observations‘ regard- — ing the large amount of urea contained in the blood of these animals. In the case of the teleosts, however, the percentage of *Schréder: Zeitachr. J. physiol. Chem,, xiv, p. 576, 1890; Baglioni: Cen- tralbl, f. Physiol,, xix, 1905. W. Denis 391 non-protein nitrogen accounted for by the urea fraction is much smaller than in the blood of man or of any mammal so far exam- ined. Im the series of determinations of urea in the blood of seven different kinds of animals (rabbit, sheep, pig, horse, monkey and beef) recently made with the same method by Folin and Denis; and in the large number of urea determinations on nor- mal human blood made by the same investigators it was found that the urea nitrogen fraction accounted for about 50 to 60 per cent of the non-protein nitrogen of the blood. In the case of bird blood (chicken, duck and goose) it appears, however, that the urea-nitrogen fraction accounted for only 25 to 30 per cent ta. the non-protein nitrogen. pe, The low urea content of bird blood agrees, well with what i is _ known regarding the small percentage of the | urinary nitro- gen of these animals which is accou d for b a. Regarding the urinary urea of.fish but little is known. ‘In a recent paper® I have reported es of the urine of the dogfish in which urea nitrogen amoun ‘to from 80 to 89 per cent of the total nitrogen. The of the elasmobranchs differs, however, so markedly fom ht the teleosts that it is to be expected that the urine _ of the two classes would also show marked dissimilarity. Below is given the result of an examination of a composite _ sample of the urine of the goosefish (Lophius piscatorius). This urine was secured from the bladders of six fish about one hour after death, and was examined on the day of collection. In - general it may be said that this urine was a clear, pale yellow fluid, with an acid reaction and a markedly fishy smell. On heat- ing to boiling a heavy coagulum of earthy phosphates was pro- duced which dissolved on the addition of dilute acetic acid. Analysis of a composite sample of urine from six goosefish. Specific gravity 1.016. Albumen and reducing sugar absent. Mgm. per Per cent of : liter total N Mgm. per liter Total N........... 830 Phosphates (as POs)... 440 SE aes, ..... 120 14.4 Chlorides (as NaCl)..... 10800 Ammonia N....... 12 2.7 Total sulphur........... 108 Uric acid N........ 1 0.1 Inorganic sulphates (as 8S) 92 Creatinine N...... 7 0.8 Creatine N........ 140 16.6 ; 5 This Journal, xiv, p. 29, 1913. ‘ ® Tbid., xiii, p. 225, 1912. 392 Blood and Urine of Fish From this examination of the urine of the goosefish, a repre- — sentative teleost, it is apparent that the small percentage of urea in the blood is also coincident with a small urea excretion in the | urine. The large percentage of undetermined nitrogen in this urine is also noteworthy. It occurred to me that this might be due to the presence in large amounts of amines. Qualitative tests, how- ever, have not confirmed this hypothesis. Another interesting point brought out is the fact that apparently in the urine of a bony fish creatinine is almost entirely replaced by creatine. A mewhat similar condition has been shown to exist in the bird.’ | should be remembered, however, that the goosefish is a voracious Teac, as much as five pounds of food being frequently found in its stomach and as this food consists of small fish the goosefish must undoubtedly consume a considerable quantity of creatine, a fact which may account in part at least for the large amount of creatine contained in the urine. The high dilution of this urine is also not without interest.as it would seem to support the theory that in the fish nitrogenous waste products may be Penmpicd in part by some organ other than the kidney. The figures obtained for the ammonia nitrogen fraction of the blood are surprisingly large. A number of ammonia determina-— tions made by the same method by Folin and Denis* showed that the quantity of ammonia in the systemic blood of cats amounts to not more than 0.1 or 0.2 mgm. per 100 grams of blood. Further observations (unpublished) on the ammonia content of normal and pathological human blood have shown that here too_ ammonia is present to the extent of only a fraction of a milligram per 100 grams of blood. In connection with the high ammonia content of fish blood it is not inappropriate to mention the well-known experiments of Cohnheim on the deaminizing power of the intestinal mucosa,° in which work, although able to demonstrate the deaminizing power of the surviving intestines of fish, he met with small success when the intestines of cats and dogs were employed. 7 Paton: Journ. of Physiol., xxxix, p. 485, 1910, * This Journal, xi, p. 161, 1912. * Zeitachr. f. physiol. Chem., lix, p. 239, 1909; also Ixi, p. 189, 1909. W. Denis 393 My results on the uric acid content of the blood and urine of fish are somewhat difficult to explain. As will be noted I have been unable to find more than a minute trace of uric acid in the blood of any of the elasmobranchs examined, while in the blood of all the teleosts it exists in moderate amounts. Uric acid was found in small amounts in the urine of both classes of fish. These findings are contrary to those of Baglioni!® who states that in the blood of the dogfish there is to be found a larger quantity of uric acid than in the urine. 10 Centralbl. f. Physiol., xx, p. 105, 1906. NOTE ON THE TOLERANCE SHOWN BY ELASMOBRANCH FISH TOWARDS CERTAIN NEPHROTOXIC AGENTS. By W. DENIS. _ (From the Biochemical Laboratory of the Harvard Medical School, Boston, . and the Laboratory of the U. S. Bureau of Fisheries, Woods Hole, Mass.) (Received for publication, October 20, 1913.) a Since the initial publications by Schlayer and his : iates in _ 1907 a large number of investigations have é t deal- _ ing with different phases of the experimental hritis Paced _ by the injection of various organie and inc - agents. In these iments rabbits, dogs and cats have been _ the animals iny y used. - In a recent publications! it has ee, shown that in the and rabbits in whom nephritis had been experimen- y produced the non-protein nitrogen of the blood may be te “increased to many times the normal amount. In these experi- iq ments our experience has been that in cases in which the non- protein nitrogen content of the blood was greatly elevated the i prognosis was bad. _ The elasmobranch fish occupy a unique position with regard _ to the non-protein nitrogen content of the blood. In these ani- mals this fraction amounts to about 1000 mgm. per 100 grams ‘blood, of which about 80 per cent is urea nitrogen; the tissues likewise contain large amounts of urea, which substance has also been shown to be present in considerable quantities in the intesti- _ nal contents, and in the bile.2 Elimination by way of the kidneys _ is small, only 20 to 50 mgm. of urea nitrogen per kilo being ex- _ creted in the urine by the starving dogfish in twenty-four hours. - In view of the above facts it occurred to me that it might be ‘interesting to see whether the elasmobranch fishes might not be ‘Folin, Karsner and Denis: Journ. of Exp. Med., xvi, p. 789, 1912; Frothingham, Fitz, Folin and Denis: Arch. of Int. Med., xii, p. 245, 1913. * Van Slyke and White: this Journal, ix, p. 209, 1911. ® Denis: ibid., xiii, p. 225, 1912. 395 | THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3 396 Resistance of Fish to Nephrotoxic Agents able to withstand the administration of relatively enormous doses of renal poisons. For all experiments the smooth dogfish (Mustelis canis) was used. During the experiment the animals were kept in large tanks supplied with running sea water and were fed every second day with fish; in many cases, however, food was refused. As the number of tanks available was somewhat limited I have confined myself to the study of two well-known nephrotoxic agents, ' 7.€., uranium nitrate and potassium chromate. These substances were administered by means of subcutaneous injections into the will be seen by inspection of Table I, uranium nitrate: may be administered to the dogfish in doses as large as 80 mgm.4 TABLE I. Cn whee, experiments. "a in each dose being injected into several different places. URANIUM NITRATE wm | ee ne mwas, KILO Se 1 1010 64 Killed 5 days after inj odtious te 2 1015 64 Killed 5 days after injection. * 3 2050 32 Killed 5 days after injection. 4 1012 64 Killed 5 days after injection. 5 1500 53 Killed 5 days after injection. 6 1500 53 Killed 5 days after injection. 7 1020 80 Killed 5 days after injection. 8 800 100 Killed 5 days after injection. 2% | 1350 60 Killed 5 days after injection. 31 700 80 Killed 6 days after injection. 32 1550 51 Killed 7 days after injection. 62 1000 80 Killed 6 days after injection. 63 | 1200 66 Killed 6 days after injection. 64 1350 60 Killed 8 days after injection. 65 | 1650 50 Killed 8 days after injection. 23 ' 750 106 Died 68 hours after injection. 27 | 630 126 Died 26 hours after injection. 2 570 140 | Died 43 hours after injection, 3 | 510 156 | Died 37 hours after injection. } * Rabbits are frequently rendered anurie by doses of 1 to 3 mgm. of ura- nium nitrate per kilo of body weight; dogs and cats are somewhat less sen- sitive, but even with these animals serious symptoms are obtained by the administration of 5 to 10 mgm. per kilo. W. Denis 397 per kilo of body weight without ill effects. All animals killed remained in excellent condition during the entire experimental period. A similar series of experiments was undertaken in which potas- sium chromate was administered. The results are given in the following table, and show that here again the dogfish appears to be very resistant towards this class of poisons. TABLE II. Potassium chromate experiments. POTASSIUM poe | wanes, | emcee cor newanes PER KILO 37 1350 48 Killed 5 days after injection. 38 1150 55 Killed 6 after injection. 39 1080 59 | Killed 6 days after injection. 40 1090. 59 Killed 6 days after injection. 19 1020, 04 Killed 5 days after injection. 20 1080 88 | Killed 5 days after injection. 18 600 | Sie Died 40 hours after injection. eee ran attempt was also made in a few cases to determine whether - the accumulation of nitrogenous waste products could be demon- strated in dogfish in whom nephritis had been experimentally induced. In order to obtain an idea of the average amounts of total non-protein nitrogen and urea nitrogen present in the blood of the dogfish, samples of blood were secured from twenty dif- ferent animals, care being taken to choose fish of varying sex, weight and age. In these samples urea and total non-protein nitrogen were determined by the methods of Folin and Denis.® The maximum, minimum and average values found were as follows _ (results are expressed as milligrams per 100 grams of blood): Maximum Minimum Average » Non-protein nitrogen...................... — 1240 900 1000 ee a 960 713 800 As will, be seen by the results presented in Table III no accu- mulation could be demonstrated in dogfish to whom large doses of uranium nitrate and potassium chromate had been given, a 5 This Journal, xi, p. 527, 1912. 398 Resistance of Fish to Nephrotoxic Agents TABLE III. Non-protein nitrogen and urea in the blood of nephritic dogfish. (Milligrams per 100 grams blood.) Lanosaronr DRBA non Peoranil Stuaes « ae van noineaOebiit FISH | ied aie NITROGEN FISH acre acy NITROGEN par ae 757 1000 32 765 990 3 | 800 1050 28 713 900 4 840 1120 19 687 875 8 | 880 1250 39 713 925 31 800 1000 40 687 920 , t not surprising if we take into consideration the small elimi- nation by the kidneys, and the apparent ability of these animals to utilize the liver and peta the intestine as an excretory organ. - # An attempt was made withiiae na of animals to collect samples of urine by means of a cannula tied in the urinary papilla; in every case, however, anuria had apparently ry the third or fourth day so that I am unable to report the Its. of any urine examinations. », UREA FORMATION IN THE LIVER.! A STUDY OF THE UREA-FORMING FUNCTION BY PERFUSION WITH FLUIDS CONTAINING (a) AMMONIUM CARBONATE AND (b) GLYCOCOLL. By CYRUS H. FISKE ann HOWARD T. KARSNER. (From the Laboratory of Pathology (Phillips Fund), Harvard Medical School.) (Received for publication, October 21, 1913.) ay INTRODUCTION. It is a desirable thing to be able to localize Kia functions to definite sites. Attem ts of this sort are frequently made, and in a large number of it has been the liver to which attention _ has been directed. “There seem to be a variety of reasons for this. - One is that the location of this organ is such as to suggest that i the function of protecting the organism in general from toxic Bistances entering the circulation from the digestive tract; another, that a considerable amount of evidence of such a pro- _ tective function exists; still another, that its size is taken, and _ probably rightly, to be more or less of an index of its importance. _ Undoubtedly it has often served as the scapegoat where ignorance has existed, especially from a clinical standpoint. Perhaps, also, the fact that the liver is a comparatively easy organ to perfuse by itself has something to do with its popularity. That its impor- tance as a specific site of metabolic processes has been overesti- mated is shown by the decrease in the réle ascribed to it in carbo- hydrate metabolism in past years. Fischler and Bardach? have apparently demonstrated that normal utilization of sugar can take place when it is severely injured. Among the processes at various times attributed to the liver, few have been so prominent and so much discussed as that of the formation of urea, especially from ammonium salts and from amino-acids. 1 Aided by a grant from the Rockefeller Institute for Medical Research. * Zettschr. f. physiol. Chem., \xxviii, p. 435, 1912. 399 pm: 4 ss 400 Urea Formation in the Liver Concerning the ability of the liver to synthesize urea from ammonium salts of organic acids which are oxidized in the body to carbon dioxide and water, there can be no doubt. The first demonstration of this by v. Schroeder,’ including the isolation and identification of urea nitrate, took place and was confirmed long ago. The purpose of such a function appears definite in the light of recent investigations by Folin and Denis,‘ which demonstrate that the ammonia of the portal blood originates largely in the large intestine and is therefore chiefly a bacterial product. It is easy to understand why it should be rendered ocuous before being distributed to the tissues. That the liver e sole site of this process is far from being a proved fact. The old assumption that the increase in the excretion of ammonia in disturbances of the liver is due directly to hepatic msufficiency need not be considered, since poe evidence against it is too great (Muenzer® and others). 3 Quite otherwise, however, stand the facts concerning urea for- mation from amino-acids. According to the view which appeared most probable until recently, the amino-acids are deaminized some- where between the lumen of the intestine and the liver, the nitro- gen being carried in the form of ammonia to the liver, there to be converted into urea, or else being immediately resynthesized into protein in the intestinal wall. Various discoveries by Folin and Denis® have greatly modified the status of this question: (1) the demonstration of the rapid distribution of amino-acids as such to the tissues, without immediate alteration, the storage of the same in a “nitrogen reservoir” (at least partly in the muscles) and their later conversion into urea, and (2) the finding of ammonia in the portal (and systemic) blood in concentrations much below those previously supposed to exist, with the major part of it coming from the large intestine. These findings make it quite unnecessary to assume that either deaminization or resynthesis occurs in the intestinal wall. The only direct evidence now existing of the occurrence of urea * Arch. f. exp. Path, u. Pharm., xv, p. 864, 1882. . * This Journal, xi, p. 161, 1912. ® Deutsch. Arch. f. klin. Med., lii, p. 199, 1894; Arch. f. exp. Path. u. Pharm., xxxiii, p. 164, 1894. * This Journal, xi, pp. 87, 161, 1912; xii, p. 141, 1912. 2 =e ELEN NCLD. Oe IT a a C. H. Fiske and H. T. Karsner 401 | formation directly from amino-acids in the liver is that of Salas- kin,’ which has stood for fifteen years with no great amount of criticism except to the effect that he did not isolate urea. His results, obtained by the perfusion of the liver with amino-acids, which showed an increase in urea by the method of Schéndorff, can hardly be accepted. In the first place he gives only the per- centage of urea (amid-nitrogen) in the perfusing fluid, without stating the amount of fluid recovered. It is therefore impossible to know how much urea (total amount) there really was in the fluid at the end of any experiment, although with the amounts used (about 1000 cc.) it is not likely that any large proportion of the volume disappeared by concentration. The point in his re- sults that stands out most prominently, however, is the faet that in all but one of four experiments in which he made analyses in the middle of the perfusion (including both of the two with gly- cocoll) he obtained a greater increase in urea in the second than in the first half of the experiment (lasting about three or four hours). _ It is quite inconceivable that such a result could be obtained with an organ removed from the body in an experiment begun fifteen to thirty-five minutes after the death of the animal and lasting more than three hours. If he did actually find such an increase in urea as his results indicate, a certain amount of it at least can be explained only as a result of abnormal post-mortem change. The liver certainly could not be more active in the second half of the experiment than in the first. The possibility suggests itself that ammonia was formed from amino-acids by autolysis or bac- teria, or both. Any decisive information for or against such a possibility need hardly be attempted now, inasmuch as no one knows the nature, the abundance or the activity of the flora inhabiting his material, or the relationships of factors governing autolysis therein. If the above suggestion be the correct one, _ the increase in urea is easily explained. Certain it is that a liver, after a three-hour perfusion, is far from normal in appearance (edema, hemorrhages, excessive fragility, etc.), so that degener- ative changes are quite conceivable. That ammonia is formed in the course of a comparatively short perfusion is evident from our results below. The assumption of such a formation of ammo- 7 Zeitschr. f. physiol. Chem., xxv, p. 128, 1898. 402 Urea Formation in the Liver | . nia in the experiments of Salaskin might possibly explain his — failure to get any evidence of urea formation in the perfusion of — muscle, inasmuch as urea formation from ammonia has never been demonstrated there, although that it may occur is still quite — possible. Since Salaskin’s figures cannot be interpreted as mean- — ing that all the increase in the urea values found represents urea — (amid-nitrogen) formation from amino-acids, it is impossible to prove that any of it does. Another possible explanation of the discrepancy between the results of Salaskin and the quite different ones obtained by us — oe lies in a consideration of the methods employed. It © is interesting to note that analyses of samples of the same mate- rial (intestinal mucosa) by Salaskin and Kowalewsky,® using the — Schéndorff and the Morner-Folin methods (applied to tissue anal- — ysis) showed 32 mgms. per 100 grams by the former, and only 14 | mgms. by the latter, whereas on the other hand, urea deter- | minations by the Schéndorff method for urine tend to give lower — results (Folin®). Furthermore, the normal urea-co t of human — blood (expressed as nitrogen) by the Schéndorff method has been — found to be 23 to 28 mgms. per 100 ee. by v.. Jaksch!® (Sché dorff" earlier, in one case, found 28.5 mgms.), while by the metho of Folin and Denis" it is uniformly 11 to 13 mgms. per 100 ce. The figures of v. Jaksch and of Schéndorff for the urea nitrogen — are practically the same as those obtained by Folin and Denis for the total non-protein nitrogen (22 to 26 mgms. per 100 cc.). There- fore, it is extremely probable that the Schéndorff figures include — something not hydrolyzed by the method of Folin and Denis. — It would appear that this substance, whatever it may be, is present in relatively greater concentration in the blood than in the urine. — Leaving out of consideration experiments with organ extracts and the like in vitro (Jacoby, Gottlieb, Lang and others), which, as suggested by Folin and Denis, are quite inacceptable (since 8 Zeitschr. f. physiol, Chem., xlii, p. 410, 1904. * Ibid., xxxii, p. 504, 1901. 10 # Interna, Beitr, z. inn. Med., i, p. 197, 1902. Quoted by Maly, xxxii, p. 265, 1902. i Pitiiger’ & Archiv, Ixxiv, p. 307, 1899. " This Journal, xiv, p. 29, 1913, 4 Thid., xi, p. 527, 1912. C. H. Fiske and H. T. Karsner 403 the abnormal production of ammonia first is not excluded), the only other evidence of urea formation from amino-acids in the liver has been derived from the results of operations upon the liver (Eck fistula, extirpation, etc.) and from liver disease (in human beings, and experimentally in animals). In the former case the evidence is based on urine analyses made either shortly after the operation, or later during a period of acute intoxication as a result of feeding meat, etc.; in either event the animal is distinctly abnormal (not solely with respect to its liver) and we believe that conclusions ‘drawn from such results have no importance what- soever as bearing upon this question, for it is certainly —- to confine a general acute intoxication to a single organ. For instance, it has frequently been entirely disregarded that many of the procedures used in such experiments uce a disturbance of renal activity, and no one knows 4 part the nephritic element has played in such results, or it is now well known that the absence of en and casts is far from being an absolute proof of | normal renal function. As far as urine analyses in liver disease are concerned, Fawitzky™ _ demonstrated nearly a quarter of a century ago that the chief cause of the earlier obtained low urea values was purely a result of the low protein intake. When the true significance of the increased excretion of ammonia in such conditions was brought to light (Muenzer™), the interest attached to this product was trans- _ ferred to the amino-acids. The evidence from the examination of the urine in cases of liver disease, in recent times, has consisted chiefly in demonstrations, by methods of very varying degrees of accuracy, of the existence of an increased excretion of amino-acid nitrogen, and of the recov- ery of amino-acids from the urine after they have been fed in amounts supposed to be largely destroyed by the normal individual (the latter primarily by Glaessner"). In the first place it is quite impossible to say, in any case of liver disease, that the liver is the sole site of functional disturbance, and therefore, no matter how important such findings may prove to be for the clinical diagnosis of such affections, it is not permissible to draw definite conclusions “4 Deutsch. Arch. f. klin. Med., xlv, p. 429, 1889. % Loc. cit. 16 Zeitschr. f. exp. Path. u. Ther., iv, p. 336, 1907. 404 Urea Formation in the Liver as to the normal functions of the liver from urinary findings in disease-complexes in which the liver merely dominates the clini- eal and the anatomical pictures. The assumption that the in- creased excretion of amino-acids in cirrhosis, for example, is due in part at least to autolysis is supported by Samuely’s!’ finding of a similar phenomenon in lobar pneumonia at the time of the crisis (isolation of abnormally large amounts of 8-naphthalinsul- phoglycine), although before this result appeared it was stated that the absence of such an increase in pneumonia was against the view that autolysis was concerned. As a matter of fact, the ts of investigations of this nature in liver disease so far have a I scccinaty variable. For example, Bergell and Blumenthal!® found a normal nitrogen partition and normal behavior of 20 grams of alanine in acute yellow atrophy. Masuda!® also has obtained results quite different from those of Glaessner. Certainly there is no functional liver test that gives results sufficiently constant for definite conclusions to be based on them, especially since the conditions in which they are the most constant are acute toxic states (when organs other than the liver are similarly damaged) and advanced stages of more or less localized, extensively destrue- tive diseases of the liver (in which the patient is far from being — functionally normal otherwise than as to his liver). Even in the most extensive of these, viz., acute yellow atrophy and phosphorus poisoning, a number of cases have been reported in which normal amounts of urea and other nitrogenous constituents (allowing when necessary for “neutralization ammonia’) were excreted, and in some of these instances amino-acids administered were destroyed to a normal extent (Rosenheim,?° Muenzer,” Badt,” Richter,® Neuberg and Richter,“ Bergell and Blumenthal,” Ishihara” and others). It is evident, therefore, from the numerous inconsisten- 17 Zeitachr. f. physiol. Chem., xlvii, p. 377, 1906. 18 Charité-Annalen, xxx, p. 19, 1906. 1° Zeitschr. f. exp. Path, u. Ther., viii, p. 629, 1911. 20 Zeitschr. f. klin. Med., xv, p. 441, 1888. %! Loc. cit. % Centralbl. f. klin. Med., xiii, p. 251, 1891. *% Berl. klin. Wochenschr., 1896, p. 454. * Deutach. med. Wochenschr., 1904, p. 499. % Loc, cil. % Biochem, Zeilachr., xli, p. 315, 1912. C. H..Fiske and H. T. Karsner 405 cies in the literature, that the time is not yet when the nature of normal processes occurring in the liver can be definitely settled by the examination of the urine in cases of liver disease. Such investigations will be of more value when they can be carried on from another view-point, for when the physiology of the liver as learned by more direct methods shall become better understood, their results may be of importance in determining more nearly what actually occurs in such disturbances. On the other side, an important piece of evidence in favor of the assumption that the liver is not a special site of urea forma- tion from amino-acids has appeared in the work of Folin and. Denis,”’ in which they have failed to find, by a method which as they say could hardly fail to show it if it existed, any difference in the urea-content of the hepatic venous blood and of blood from other parts of the body while non-protein nitrogen from amino- acids or Witte’s peptone was being absorbed from the small intes- tine and the urea eontent of the blood increasing. EXPERIMENTAL, vad rie ~ Our experiments consist in the perfusion of the livers of rabbits and cats with defibrinated blood, with or without the addition of serum and Ringer’s solution or of the latter alone. The methods used have permitted the employment of comparatively small _ quantities of fluid (usually about 100 ce.). In all the cat experi- ments the blood has been that of the animal whose liver was used; in all, the fluid has been from the same species. We have per- formed experiments of two kinds; with the addition of (1) com- mercial ammonium carbonate, and (2) glycocoll, to the perfusing fluid. : Method. The analyses were made in duplicate. Total non-protein nitro- gen, urea and ammonia were all determined by the recently de- vised methods of Folin and Denis.** The figures given as urea nitrogen represent, of course, everything hydrolyzed in a constant boiling mixture consisting chiefly of potassium acetate at about 27 This Journal, xii, p. 141, 1912, 28 Ibid., xi, p. 527, 1912. 406 Urea Formation in the Liver 150°C. in ten minutes, minus the ammonia nitrogen separately determined. Technically, the experiments were arranged so that small quan- tities of perfusing fluid could be used, passing continuously and repeatedly in the same direction at an approximately constant temperature. The fluid was accurately measured at the begin- ning and end of the experiment and was usually about 100 ce. The temperature was usually at 38°C., but occasionally momen- tarily rose to as much as 40° and very rarely sank to 35°. The accompanying diagram shows the scheme adopted for the use of quantities of fluid, the bottles being of 125 ce. capacity with ride necks. Perforated rubber stoppers were used and rubber tubing of 2 mm. caliber. The fluid passed out of one bottle through the liver and into the other bottle, then by throwing over the rocker valve, the filled bottle was made the supply bottle, the fluid continuing to flow in the same direction without notice- able interruption. The tube conducting fluid away from the bottles was connected with a coil of glass tubing in a leaden box containing water heated by an electric stove. From the coil the fluid passed to a T-tube in which the bulb of a thermometer was > placed so that one arm connected with the heating coil, one arm with the thermometer, and one arm with a rubber tube 10 em. in length which led to a cannula in the portal vein. The tubing from the T-tube to the portal vein, as well as the liver, trunk of the animal. and several centimeters of tubing leading away from the inferior vena cava were contained in a double walled tin box, in which a temperature of about 40°C. was maintained by means of an electric light bulb. All animals were bled to death before using the liver for per- fusion. In the case of the cat, the animal was etherized and bled from a cannula placed in the carotid artery. In this way sufficient blood could be collected from one animal to serve for the experi- ment with its own liver. The blood was rapidly defibrinated in a flask with glass beads and the blood mixed with the other materials as described above. As soon as bled, the abdomen and thorax were opened, glass cannulae placed in the portal vein and in the inferior vena cava immediately above the diaphragm. The aorta and vena cava were ligated beneath the diaphragm and the C. H. Fiske and H. T. Karsner 407 lower part of the trunk and upper part of the thorax severed. The remaining part of the trunk including liver and intestines was wrapped in towels soaked in hot salt solution (0.85 per cent NaCl), placed in the tin box and connected with the tubing so that the perfusing fluids entered the portal vein. This procedure occupied from ten to twenty minutes. In the case of the rabbit, the same technique was used except that usually the blood of two animals had to be mixed in order to obtain the proper amounts. These animals were bled by a rapid severing of the femoral artery so that a minimum of ether was used. The animals were dead before being opened for expo. sure of the liver. I The length of time occupied in the perfusion was im most in- stances about one hour. The definite res »btained with am- monium carbonate in this length of time, combined with the distinctly abnormal appearance of the liver after several hours’ perfusion and abate: results obtained by Salaskin appear to justify the employment of short experiments. In every case the blood was removed from the liver by previously perfusing with Ringer’s solution, therefore it was not considered necessary _ to run the fluid through before making the first analysis. The substance used in each case was added in aqueous solution to the fluid and the two thoroughly mixed by shaking. Samples were then taken for analysis (5 cc. for precipitation with acetone-free methyl alcohol, 4 to 10 ec. for duplicate ammonia analyses) and the fluid poured into one of the bottles. Samples were again taken at the end of the experiment (and in one case during its course). The total quantities of the various substances analyzed were calculated from the amounts of fluid put in at the beginning and removed at the end of each experiment in those cases in which obvious loss, as a result of leakage, etc., did not occur. The remarkable constancy of the total non-protein nitrogen in most of the experiments in which there was no such loss (Experiments 10 and 11 are exceptions) indicates how slight must have been any washing out from or absorption by the liver during their course. Naturally the fluid left in the liver could not be added to that recovered, nor was its amount determined, but there can be no doubt that calculations based upon the total amount recov- ‘a. ae ey) ona ered are of more value than mere percentages. Slowtzoff and Ssobolew?® furthermore found only 2.0 to 5.6 per cent of blood in the livers of human cadavers by colorimetric determination. 408 Urea Formation in the. Liver I. Ammonium carbonate. EXPERIMENT 1. Normal rabbit. Fluid: 23 ec. rabbit serum, 36 cc. de- fibrinated rabbit blood, 31 cc. Ringer’s solution and 101 mgms. of com- mercial ammonium carbonate. Sixty-five minutes, nineteen times. In, 100 ec. Out, 115 ce. MILLIGRAMS PER 100 cc. TOTAL AMOUNT IN MILLIGRAMS Total non-pro-| Ammonia Total non-pro-| Ammonia tein nitrogen nitrogen tein nitrogen nitrogen 62 22 62 22 54 7 62 8 ExpertMent 2. Normal rabbit. Fluid: 23 ce. rabbit serum, 20 cc. de- fibrinated rabbit blood, 57 cc. Ringer’s solution and 103 mgms. ammonium carbonate. One hour, forty times. In, 100 ce. Out, 98 ee. (diluted to 155 cc.). MILLIGRAMS PER 100 cc. TOTAL AMOUNT 1N MILLIGRAMS Total non-pro-| Ammonia Total non-pro-| Ammonia tein nitrogen nitrogen tein nitrogen nitrogen Belore 33530 66 24 66 24 miter... sta... ee 43 5 67 8 In the following three experiments (3, 4 and 5) accidents caused the loss of considerable amounts of fluid, therefore the total recovery is unknown. Experiment 3. Normal rabbit. Fluid: 18 ce. rabbit serum, 538 ce. de- fibrinated rabbit blood, 9 ec. Ringer’s solution, 20 cc. normal saline solu- tion, 71 mgms. glyeocoll and 45 mgms. ammonium carbonate. In, 100 ec. | a area PER waeciil ae “Toul non pale nie | ! jie nitrous Bae... sven TEs o-« clad oy 62 a 14 ASR, ss. Vee ah 60 4 % Biochem, Zeitschr., xxxi, p. 234, 1911, ae C. H. Fiske and H. T. Karsner 409 Experiment 4. Liver of rabbit in which biliary cirrhosis had been produced by ligation of the common bile-duct five weeks before (Richard- son*®), Fluid: 63 cc. defibrinated rabbit blood, 37 cc. Ringer’s solution and 104 mgms. ammonium carbonate. Thirty-two minutes, five times. In, 100 ce, MILLIGRAMS PER 100 cc. Total non-pro- | Urea Ammonia ‘aif Wass : tein nitrogen | nitrogen nitrogen nitrogen NET Pear a earn esa | 43 | 18 Trace*! 25 Fluid befors........... eg 1182 rc} -36 Pitig Miter... ores | 62 26 7 29 The ammonia fell from 47 to 11 per cent of the total non-pro- tein nitrogen (decrease of 77 per cent.) The urea increased from 22 to 42 per cent of the total non-protein : (an increase corresponding to 56 per cent of ammonia lost). Although a per- fectly definite inerease in urea (amid nitrogen) is seen, 44 per cent of the lost ammonia was not recovered as urea. The two possibilities are: (1) that the remainder of the ammonia was re- tained in the liver, and (2) that it was converted into something ‘other than urea. The results of Experiment 6 throw some light upon this. Expertment 5. Normal rabbit. Starved forty-eight hours. Fluid: 65 cc. defibrinated rabbit blood, 35 cc. Ringer’s solution and 105 mgms. ammonium carbonate. Twenty-eight minutes, eleven times. In, 100 cc. MILLIGRAMS PER 100 cc. * Total non-protein nitrogen | Ammonia nitrogen PIES, 2 ta ee, 54 21 Re oT 60 7 80 Journ. of Exp. Med., xiv, p. 401, 1911. 31 Quantitative ammonia determinations were not made when 10 cc. of the Nesslerized solution could not be read in the ordinary colorimeter. Where a “‘trace’’ is reported, however, it is certain that there was con- siderably less than 0.5 mgm. in 100 cc. of the blood. This and other anal- yses of normal blood from both rabbits and cats confirm the findings of Folin and Denis, viz., that the ammonia in the systemic blood of normal animals is present only in traces. * Calculated from the amount found in the defibrinated blood before anything was added to it, on the basis of the dilution. 410 Urea Formation in the Liver EXPERIMENT 6. Normal cat. Starved twenty-four hours. Fluid: 40 ec. defibrinated cat blood, 35 cc. Ringer’s solution and 107 mgms. ammo- nium carbonate dissolved in 4 ce. water. One hour, forty-five times. In, 80 ec. Out, 92 ce. MILLIGRAMS PER 100 cc. TOTAL AMOUNTIN MILLIGRAMS) - Total ey | Urea | Ammonia | Rest | Total n.p. | Urea | Ammonia} Rest nitrogen | N N N nitrogen N* N N Blood..... 51. |) Trace 31 20 9 Trace 1% Fluid be- fore..... 52 28 42 23 Fluid after 48 15 7 26 44 14 6 24 last experiment, in which a complete recovery was made, shows the same thing as Experiment 4. Seventy-two per cent of the ammonia Gisappeared, but only 26 per cent of the lost’‘ammonia was recovered as nstancy of the total non-protein nitrogen here, as in other experime ats in which practically every- thing was recovered, is greatly against the assum rtion that any appreciable amount af nitrogen, in the form of mia or other- wise, has been retained in the liver. The conclusion of Kowalew and Markewicz* from their perfusion experiments, that the am nia is deposited in the perfused organ, is based upon two experi-- ments with muscle, in which increases of 4.8 and 7.8 mgms. of © ammonia per 100 grams, respectively, were found, after perfusion with blood containing about 14 mgms. per 100 grams; the method (distillation with magnesia) is certainly not accurate for blood, and is therefore even less likely to be so for tissues. If the in- ** The determination of urea in the presencé of comparatively large amounts of ammonia by the method used has not given satisfactory re- sults, the loss occurring, as far as we now know, in evaporating off the methyl alcohol. ments 4 and 6 in the blood before adding the ammonium carbonate, and the urea content of the fluid calculated from the result. For the same reason, the figures for the urea at the end of the experiment are in all probability minimal, for it is very likely that some ammonia is lost even there. Even assuming, however, that the total amount of nitrogen ob- tained after hydrolysis in determining the urea and ammonia together represents urea nitrogen alone, which cannot be so, for with the larger amounts the loss was only about 20 per cent, there is still not enough increase to equalize the loss in ammonia during the perfusion, 4 Loc. cit, For this reason the urea has been determined in Experi- — ee a C. H. Fiske and H. T. Karsher 4ll crease in “rest nitrogen” in our experiments were due to washing out from the liver, it is hard to see why it should so nearly equal the ammonia unaccounted for. In another experiment with ammonium carbonate, all but the ammonia determinations miscarried. ExprertmmMent 7. Normal cat. Fluid: 87 cc. defibrinated cat blood and 103 mgms. ammonium carbonate in 5 cc. Ringer’s solution. Perfused one hour. | MILLIGRAMS PER 100 cc. TOTAL AMOUNT IN MILLIGRAMS | Ammonia nitrogen Ammonia nitrogen > | ; Lane LORIE ete e NG cs 4's / 23 p> é UGE. cls ede Sak ees 7 a 7 4 * es ; The results of all the experiments with ammonium carbonate are given below in tabular form: MILLIGRAMS PER 100 cc. Before After Total n.p.| Urea | Ammonia | Rest | Totaln.p.| Urea | Ammonia | Rest nitrogen N N N nitrogen N N N | 62 22 54 | 7 2 | 66 24 43 COC 5 3 62 | 1433 60 4 4 | 51 11 24 16 62 | 26 7 29 5 | BA 21 60 | 7 Of Cael: 11 28 | 18 4815 7 26 1 23 7 AMMONIA N IN PER weumenor | EGET: | Mor zomuaa, N | ramomee op | ZR CENT Or, j , — - —— Secor N RECOVERED Before After | Before After ee oaes 1 35 | 63 2 36 2 | 67 3 23 | 70 4 47 i. | 22 42 | 77 56 5 39 12 | 69 6 54 15 21 31 72 =. 26 THE J NAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 3 412 Urea Formation in the Liver Therefore, from 63 to 77 per cent of the added ammonia dis- appeared, and of this only a part was recovered as urea. II. Glycogoll.% ExpEerRIMENT 8. Normal rabbit. Fluid: 46 cc. rabbit serum, 36 cc. defibrinated rabbit blood, 23 ec. Ringer’s solution and 183.8 mgms. glyco- coll (34.3 mgms. nitrogen). Fifty minutes, twenty times. In, 105 cc. Out, 102 cc. MILLIGRAMS PER 100 cc. TOTAL AMOUNT IN MILLIGRAMS Total non-pro-| Urea and am- | Total non-pro-| Urea and tein nitrogen |monia nitrogen | ‘tein nitrogen ammonia N. 76 20 80 21 76 mee 77 19 EXPERIMENT 9. Nort bt | L: 31 ce. rabbit serum, Bac. defi- brinated rabbit blood, 15 ec. ed ion and 122.3 mgms. glycocoll (22.8 mgms. nitrogen). Sixty-one minutes. Thirty times. a. 100 cc. Out, 115 ce. MILLIGRAMS PER 100 co. TOTAL AMOUNT IN urbe \ Urea and ammonia N. | Total non-pro-| Urea and am- | Total non-pro- tein nitrogen fmonia nitrogen) tein nitrogen Belore..xavi,.. »s see 70 22 70. 22 SCOR ci ees.» sae 60 17 69 — ne - nenetine alanis —— ns a ny In the next two experiments, with cats, there appears to have been a certain amount of nitrogen washed out of the liver. EXpEeRIMENTIO. Normaleat. Liversomewhatfatty. Starved twenty- four hours after meat régime. Fluid: 44.5 ec. defibrinated cat blood, 58 cc. Ringer’s solution and 188.5 mgms. glycocoll (35.2 mgms. nitrogen). One mies mighteen tienes. In, 102 cc. Out, 122 ce. MILLIGRAMS PBR 100 ce, TOTAL AMOUNT IN MILLIGRAMS Total non- pro-| Urea andiam- Total nonpro- Urea and tein nitrogen a nitrogen | tein nitrogen ammonia N. Bilee...... Gm... 62 | o | 6 10 BE. side eee. 65 | |e 1 * Kahlbaum’s glycocoll was used in all cases. C. H. Fiske and H. T. Karsner 413 EXPERIMENT 11. Normal ‘cat. Fluid: 56 ce. defibrinated cat blood, 19 cc. Ringer’s solution and 188.7 mgms. glycocoll (35.2 mgms. nitrogen) in 5 ce. water. One hour. Thirty-one times. In, 80 cc. Out, 105 ce. MILLIGRAMS PER 100 cc. TOTAL AMOUNT IN MILLIGRAMS Urea and ammonia N. Total non-pro- | Urea and am- | Total non-pro- tein nitrogen |monia nitrogen| tein nitrogen IRON oes te a 72 17 58 14 PRBS STV s civie oes 66 15 69 15 | of ammonia nitrogen, making the total urea nitrogen only 13 In the fluid at the end of the perfusion were found 1.9 milligrams ° ’ _ A comparison of this result with the normal amount of onia _ nitrogen in cat’s blood found by Folin and Denis (about 0. ‘mgm. _ per 100 cc.) demonstrated that there is f i _ during the course of an hour’s perfusion. The next experiment - (12) shows the . thing. As stated earlier, in no case have we been able to get enough ammonia from 5 cc. of normal blood to be anywhere nearly readable in the Duboseq colorimeter as | ordinarily used, 7.e., without the polariscope tube and iris dia- phragm used by Folin and Denis. ExpPrerRIMentT 12 Normal cat. Starved twenty-four hours. Fluid: 76 "ee. blood and 179.5 mgms. glycocoll (33.5 mgms. nitrogen) in 4 ce. water. | Total time, three hours. Analysis of fluid after one hour as well as at | end. In, 80ce. Out, in one hour, 105 cc.; in, three hours, 84 cc. MILLIGRAMS PER 100 Cc. * TOTAL AMOUNT IN MILLIGRAMS a ae aeip. rom OZ, § gH s aaa | Mm | 8 ee | | § a, $ $ g Total n.p. N |g E LB AM eS _ Before........ 75 | 12.0 60 | 96) | fi t a removed 61 | 10.5) eet. 58) Ep iar 58 | 10.0 _ After 3hours.) 69 | 13.8/11.6| 2.2 58 | 11.6; 9.8| 1.8 5 cc. of the 105 removed at the end of one hour were used for analysis, the other 100 ce. returned to the apparatus. In the following experiment the outflow of fluid from the liver was obstructed early in the perfusion, causing the liver to swell permanently. Therefore, the recovery was only partial. Ringer’s 414 Urea Formation in the Liver solution had to be added to make the volume large enough to continue the experiment. EXPERIMENT 13. Normal cat. Starved twenty-four hours after meat diet. Fluid: 81 cc. defibrinated cat blood and 173 mgms. glycocoll (32.3 mgms. nitrogen) in 9 cc. water. Forty minutes, thirty times. In, 90 ce. Out, 97 cc. MILLIGRAMS PER 100 cc. TOTAL AMOUNT IN MILLIGRAMS : pre UREA+NHs-N Epes ia oo: and) am- arr Urea ang d oo TOTAL N. P. nitrogen nitrogen nitrogen in per cent “a Before... Se Cleave 78 20 26 | Alver... 50 11 48 11 23 u It is seen that in none of the e any suggestion of urea formatior ents with glycocoll is there : not even in Experiments 10 and 11, in which a considerable amount of nitrogen was washed out of the liver), although precisely similar experiments with ammonium carbonate demonstrate the ability of the surviving liver under these conditions to metabolize considerable amo of ammonia, and, in the two experiments in which data are mat oo able (4 and 6), to form urea, or some nitrogenous substance not blown over by the air-current, but hydrolyzed at 150°C. under — the conditions of the determination. The concentrations of gly- | cocoll added have been about those used by Salaskin. In the introductory part of this paper we have offered a number — It is not at — of possible explanations of this difference in results. all likely that the difference in the animals used would lead to so — different results. We by no means wish to argue, from the results obtained, that urea formation from amino-acids does not occur in the liver at all, for such a conclusion would be quite unjustifiable on the basis We do believe, however, that it is extremely doubtful that such a process has ever been demonstrated. The only amino-acid used by us has been glycocoll, but inas- much as there is no evidence for urea formation from other amino- ° acids in the liver that is not subject to the same criticism as in the case of glycocoll, we feel justified in making the statement of the above data. more general, . ; i 4 | -C. H. Fiske and H. T. Karsner 415 SUMMARY AND CONCLUSIONS. 1. The surviving liver is capable of destroying ammonia per- fused through it in the form of ammonium carbonate, and of converting it partially into urea. The entire amount of ammonia changed, however, does not have this fate. How much, if any, _ of it undergoes synthesis to amino-acids has not been determined. It is doubtful whether the binding of ammonia as such by the liver cells is of much significance in the protective influence of the organ, as indicated by the lack of variation in total non-protein nitrogen content of the fluid during the experiment. 2. The perfusion of the liver of the cat or the rabbit with homol- ogous defibrinated blood containing as much as 44 mgm, of nitro- gen as glycocoll per 100 cc. does not lead to any increase in the amount of urea in the fluid used. - 4 3. The formation of urea from amino-acids by the liver is not conclusively demonstrated. There is no incontestable ground for the assumption that the liver is a special site for such a process. We are greatly indebted to Professor Otto Folin for valuable advice received at various times during the course of this work. MH o = —] Vv a ~ q — a ° ‘Ss (01 = M ° fy 4) v Le es 416 ss Wail § nt € : ; wy i fh =3V — J N Ay S 27 . J i2 \\ g a 8 Th are et at Si J ii iu \ g . a. | , | 5 * as ¥ ee Wd ( } z i ——— ) herr Walch Ty YA s Ii EZ ee rae SS ii Tube A is for inlet of air pressure, passing through A‘ or A? to the bottles. Fluid passes through B' or B? to B. Fluid returns through C to C! or C*, and thence to bottles. Air is rel bot through tubes D'or D?. When the rocker valve is e, the air pressure reaches bottle J, tube D being closed side of throttle; fluid passes through tube BB! to ae of the Y-tube conmeling with C! not per- mitting passage of i. of closure by Z side of rocker. Fluid passes thence through B, through water coil, liver and returns to bottle II by w of C and C2, the latter being open on X side of throttle; fluid 4s into bottle IJ because air is allowed to escape through D?. and D? were frequently connected with a Y-tube and the exhaust air passes through 74 HCl, but not enough ammonia was detected with _ Nessler’s reagent to affect the figures. Tube EZ was kept closed by means of a clamp, but: when bottle IJ was filling, air freed from ammonia was forced into the bottle so as to aerate _ the fluid. With reversal of the rocker, the process was reversed but without notice- id able interruption to the outflow through B and inflow through C, although extremely careful attention had to be given water coil and thermometer at this time. e THE SATURATED FATTY ACID OF KEPHALIN. By P. A. LEVENE anp C. J. WEST. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, October 22, 1913.) The work on the composition of kephalin has made little prog- > _ ress from 1909 until the last few months. Last winter we had in our possession a considerable quantity of theether-soluble frac- - tion of brain lipoids and hence employed it for preparation of _kephalin. The preparation was carried out following the process employed by previous workers.' The purification was accom- _ plished by repeatedly dissolving the crude substance in petrolic ether (boiling point, 40—50°C.) and precipitating it with alcohol. This operation was repeated about ten times, until the final product had on drying a light straw-yellow color. The ethereal solution _ was allowed to flow into the alcohol in a very slow stream, me- chanical stirring being employed during the precipitation and at _ least about one-half hour after the completion of the precipitation. _ The product was then suspended in acetone, which was allowed to act on it under constant stirring. The final product was a dry, light powder. At the close of last season the work was completed only so far as it concerned the saturated fatty acids obtainable on hydrol- ysis of kephalin. There existed conflicting statements regarding their nature, Cousin? having isolated only stearic acid, while Frankel and Neubauer claimed the presence in kephalin of both stearic and palmitic acids. Four samples of kephalin were analyzed and in all of them was found on hydrolysis only one saturated fatty acid, namely, stearic 1 Falk: Biochem. Zeitschr., xiii, p. 153, 1908; xvi, p. 187, 1909; Frankel and Neubauer: ibid., xxi, p. 321, 1909; Frankel and Dimitz: ibid., xxi, p. 327, 1909; Parnas: ibid., xxii, p. 411, 1909. ~ ? Cousin: Jour. d. pharm. et d. chim., xxiv, p. 101, 1906; xxv, p. 177, 1907. 419 420 The Saturated Fatty Acid of Kephalin acid. The publication of this result was postponed as it was planned to resume the further study of the composition of kephalin in course of the present season. However, there has appeared in course of the summer a very important work by Parnas and his co-workers* on the nature. of the nitrogenous component of kephalin, and in the last number of the Biochemische Zeitschrift! another publication, also by Parnas,® on the saturated fatty acids of kephalin, in which the writer arrives at the conclusion that stearic acid is the only saturated fatty acid present in the molecule of kephalin. We therefore concluded to present the results of our last year’s york not only because they strengthen the conclusions of Parnas, but also for the reason that in four out of five experiments the separation of the saturated from unsaturated fatty acids was carried out by a process different from the one employed by pre- vious workers. ORG The separation was based on the Giiferbhive i in the solubility of | the ethyl esters of the saturated and unsaturated acids. At 0°C. the saturated acid precipitates out of the alcoholic solution, while the unsaturated still remains in solution. The separation of tl saturated acid is perhaps not absolutely quantitative, but ve neat and convenient. The results obtained on the acid hydrolysis of kephalin were corroborated by the result obtained on alkaline — hydrolysis by means of a barium hydrate solution. EXPERIMENTAL PART. The kephalin used in this investigation was purified by repeat- edly dissolving it in 40-50° petroleum ether and pouring the solu- tion with stirring into aleohol. Two different samples were ana- lyzed for nitrogen and phosphorus. * Baumann: Biochem. Zeitschr., liv, p. 30, 1913; Renall; ibid., lv, p. 296, 1913, * Parnas: Ibid., lvi, p. 17, 1913. * Parnas, referring to the article by one of us (Levene) in this Journal, xv, p. 153, attributes to the writer the statement that the fatty acid was obtained on hydrolysis of a lipoid by means of alcohol acidulated with sul- phuric acid. This was an error on the part of Parnas, for the reason that it is clearly stated in the article referred to that the ethyl ester was the first product obtained on the hydrolysis of the lipoid with acidulated alcohol. ‘ 3 ee | ie f . P. A. Levene and C. J. West 421 I. 0.2988 gram substance gave 0.0430 gram Mg,P20;. 0.2004 gram substance required 2.15 cc. 7 HCl (Kjeldahl). II. 0.3058 gram substance gave 0.0440 gram Mg2P207. 0.4152 gram substance required 4.8 cc. fy HCl (Kjeldahl). Found: I Sy BO (Re RR eS Sa 4.01 4.01 PMS ei, ate REA Glin ai so as oe 3.0 ee 1.50 1.62 Hydrolysis with alcoholic HCl. Fifty grams of kephalin were heated with 600 ec. methyl! alco- hol and 20 ee. concentrated sulphuric acid ten hours under a reflaiie and in an atmosphere of carbon dioxide. At the end of the heat- ing there was a slight mineral residue. * The colored solution was decanted from this residue and cooled in the dee box over night. The precipitate which formed was red off and recrystallized from acetone. It melted at 37-38°. Methyl stearate melts at 38°. The ester was hydrolyzed by heating with alcoholic potas- _ sium hydroxide, the soaps decomposed by warm hydrochloric acid, washed free from acid and recrystallized from acetone. The first on was analyzed (I), then further purified by changing into _ the lead salt and decomposing with hydrogen sulphide in toluene; - this was then recrystallized from acetone and the last trace of solvent removed by melting. The mother liquor from (I): was concentrated and the acid which separated analyzed (II). Four different hydrolyses were carried out with similar results. The four samples showed m.p. of 69°, 69°, 68-69°, 69°, and gave no _ depression when mixed with a sample of Kahlbaum’s stearic acid, purified through the lead salt. Analysis of the acid. I. 0.1280 gram of the substance gave 0.3557 gram CO, and 0.1492 gram H,0. II. 0.1388 gram of the substance gave 0.3860 gram CO, and 0.1576 i gram H,O. ‘Calculated for Found: 18Hs¢O2: I Thy. Xt ec... 5. a... 76 .00 75.79 75 .84 SES SR Sa 12.70 13 .05 12.71 Molecular weight estimations. Samples from four different hydrolyses were used, all of which were _ purified through the lead salt. he 422 The Saturated Fatty Acid of Kephalin I. 1 gram of theacid dissolved in amixture of absolute methyl alcohol and benzene, when titrated with -} alkali, using phenolphthalein as an indi- cator, required 34.8 cc. 7 NaOH for neutralization. II. 1 gram of the acid, as above, required 34.9 cc. 75 NaOH. III. 1 gram of the acid, as above, required 34.9 cc. 7 NaOH. IV. 1 gram of the acid, as above, required 34.7 cc. #4; NaOH. Calculated for Found: CisH 302: Aqueous barium hydrolysis. A fifth sample of the same kephalin was hydrolyzed by boiling with aqueous barium hydrate in an autoclave for several hours. The mixture of barium salts was filtered off, the unsaturated acid removed by repeated extraction with ether, the saturated acid liberated with warm dilute hydrochloric acid, changed into the lead salt, decomposed with hydrogen sulphide and recrystallized from acetone. The acid thus obtained melted at 68-69°. Mixed with pure stearic acid, it melted at 68-69°. No trace of any other saturated fatty acid could be found. be l gram of the acid, as above, required 34.6 ec. 4 NaOH for neutralizati Calculated for CisHssO2:; Found; M: Wisasess. cRiRERts :Ah cen MM 5 286 288 THE INFLUENCE OF BUTTER-FAT ON GROWTH.:' By THOMAS B. OSBORNE anp LAFAYETTE B. MENDEL, With the Coédperation of Epna L. Ferry and Atrrep J. WAKEMAN. (From the Laboratory of the Connecticut Agricultural Experiment Station and the Sheffield Laboratory of Physiological Chemistry in Yale University, New Haven, Connecticut.) (Received for publication, November 4, 1913.) We have recently pointed out? that while young rats grow for a time at a normal rate on the ‘protein-free milk” diet used in our i they sooner or later cease to grow, so that more than two-thirds of the weight normal rats receiving a diet chiefly composed of milk and . ermore, we showed that rats which had ceased to TOW and were declining on a “protein-free milk’ diet, at once ered and resumed a normal rate of growth when a part of . their food was replaced by a quantity of unsalted _ butter cor ng to that in the milk-food. The striking way in which setae thus supplied, influenced the growth of these young rats made it evident that it furnishes some substance which ~ exerts a marked influence on growth. _ These observations have since been verified by numerous addi- _ tional experiments and an attempt has been made to determine _ with which of the components of the butter this growth-promoting _ power is associated. As is well known, butter consists of about _ 82-83 per cent of the glycerides of numerous fatty acids, about _ 15 per cent of water containing each of the soluble constituents of milk, and from 1 to 2 per cent of solid, matter, consisting chiefly of cellular débris from the mammary glands, bacteria, calcium 1 The expenses of this investigation were shared by the Connecticut Agricultural Experiment Cation and the Carnegie Institution of Wash- ington, D. C. * Osborne and Mcidel:-ihe Relation ‘ot Growth to the Chemical Con- stituents of the Diet, this Journal, xv, pp. 311-326, 1913. oes a Wie Pe. 424 Influence of Butter-fat on Growth phosphate, particles of casein, and accidental impurities intro- duced during the process of making the butter. In view of the possibility that even an extremely minute quan- .tity of some substance might exert the favorable influence on growth observed in all of our experiments, we separated the buv_cr into three parts, namely the fatty substances, the insoluble solid elements, and the aqueous solution containing lactose, soluble inorganic salts and other soluble components of the milk; and by feeding trials found that the growth-promoting factor was contained in the fat fraction. Further consideration of the two other fractions is therefore unnecessary. . The butter-fat used in the analytical, as well as the feodinig experiments here described, was prepared as follows. The butter was melted by heating in a flask immersed in a bath of water not exceeding 45°, and centrifugated for about an hour at a high speed. The melted butter was thus separated into * layer of perfectly clear fat, an opalescent aqueous layer, a “posit of white solid matter. The clear fat was ageced off y . nd- thus separated from all of the other parts of the butter. his~ method the use of all solvents was avoided, and any subst¢ which might have been dissolved thereby from the other y part of the butter were excluded. nail” So much has been written about the significance of phosphatides (lecithin, ete.) in various biological phenomena, and in growth among others, that a careful analysis of large quantities of the butter-fat was instituted to detect the presence of members of this group of so-called lipoids. The butter-fat prepared as described was found to be ,entirely free from nitrogen and phosphorus and was devoid of any ash-yielding, or water-soluble, components. The absence of phosphatides from the product corresponds with the recent statement of Njegovan* who con- cludes that milk contains no lecithin whatever, and suggests that the contradictory claims of other investigators are attributable to inadequate methods of analysis. Butter-fat, thus prepared, has proved to be quite as effective as butter or milk in promoting the recovery and renewed growth of animals which have ceased, or failed, to grow on the natural * Njegovan: Biochem, Zeitschr., liv, p. 78, 1913. T. B. Osborne and L. B. Mendel 425 “protein-free milk’ dietaries in which lard furnished the fat | component. Our previously published-experiments* in which sim- ilar recoveries were made when butter replaced a part of the lard of diets containing the “artificial protein-free milk’ are compli- eated by the fact that butter contains about 15 per cent of butter- ‘milk, and hence the improvement shown by such experiments might be attributed to some constituent of the buttermilk. Ex- periments with “artificial protein-free milk” and butter-fat are in progress and it is hoped by these to learn whether or not ‘accessory substances other than those contained in the butter- fat are necessary for growth. These experiments are not yet completed. Charts showing the body-weights of growing rats that had begun to decline on our “protein-free milk” mixtures, and were supplied with foods having butteror butter-fat introduced,° are appended (see Charts I hd, and III). The composition of the mixtures Was as = Butter foods Butter-fat foods ; per cent per cent SU see eee 18 18 7. Coo WO cc... ce cae 26 26 il ee, wk... es cca 28 28 | SS ani 10 10 Pen. Wee, yo cte c's wena e vais 15.3 18 es 2 4 butter 4 ® The efficiency of the butter-fat, or some component thereof, n specifically promoting growth is further shown in another way. ‘on a mixture of purified protein, lard, starch, and “protein-free milk,” prepared in imitation of the gross composition of the highly , they show a varying capacity to grow. In cases gro} bas stopped after sixty days; other animals grow for one hundred days or more. But in periments there has always been an inevitable ulti- *Phis Journal, xv, p. 326, 1913. ‘ee Bor ‘similar recoveries induced by milk-food, see Charts II and TH, this Journal, xv, pp. 321-322, 1913. s ° For details regarding the milk-foods used in our experiments, and other recs, see Osborne and Mendel: this Journal, XV, p. 318, 1913. Br ee has already been pointed out, when very young rats are placed. ~N 426 Influence of Butter-fat on Growth mate inhibition of growth, and nutritive decline, connected with some diet factor. Very few of our rats grew after they were 140 days old, at which age two-thirds of the normal growth of the male, and three-fourths of that of the female is usually made. We have accordingly attempted to learn whether the continued exhibition of butter-fat from very early in the growth period would enable the animals to attain their normal maximum weight and thereby avert the invariable failure which hitherto was met with sooner or later. If it be assumed that in these earlier experiences the substance, or substances, essential for growth, and supplied inadequately, or not at all, in the artificially prepared diets, is furnished by a reserve stored in the cells of the young animals, this must be exhausted after a time, and lead to failure of growth. If, however, a growth-promoting substance is present in the butter- fat, and the latter is supplied in abundance, growth ought to continue to its logical conclusion, in the absence of other inhibi- — __ tory factors. | Experiments to test this were begun with youn ng | diet from an early period consisted of food of the fo following com-_ | position: wy, Purified proteit. cag. «e's dy SE ck a... ss sie a ee ms Staroh. ..:. sce eee. sss kG es ss SRI.» ss ces 26 Protein-free milk............ ee... SSRs... « » SAG one acai 28 Lard ids oc. cc ictsbce « added butter-fat may have simply supplied something analogous to the so-called vitamines, which Funk considers to be essential — for life, and thereby enabled the animals to resume growth on a food thus made adequate for maintenance. Until it is shown that rats can be maintained for long periods on such diets as McCollum and Davis used no final conclusion can be draw respecting the above question. q By numerous experiments we have shown that mature rats can be maintained on our “protein-free milk’ diets for more than a year, and that young rats on similar diets containing proteins inadequate for growth can be maintained nearly as long. Such / / foods consequently supply all that is essential for mazntenance 4 alone. Since growth ceases on these foods after a comparatively short time, and is at once resumed and continued throughout the entire period of normal growth when a part of the lard is replaced by butter-fat, it is almost certain that butter-fat contains some- thing essential for growth in addition to what may be required for maintenance. This recovery and renewed growth must be attributed to something which distinguishes butter from the ordi- nary fats, for not only do lard and olive oil® lack this growth- promoting power, but young rats grow on our “protein-free milk”’ foods when all of the lard is replaced by carbohydrate!’ and no ether-soluble substances are present in the food. *McCollum and Davis; loc. cit, ' Cf, Osborne and Mendel; Feeding Experiments with Fat-free Food Mixtures, this Journal, xii, pp. 81-89, 1912, \ T. B. Osborne and L. B. Mendel 429 It thus appears improbable that glycerides of the fatty acids ordinarily present in foods are responsible for the promotion of the growth observed when butter-fat replaces lard in the diet of rats which have ceased to grow. Lecithin and other phosphorus- or nitrogen-containing substances are excluded by the absence of phosphorus and nitrogen from our butter-fat; and cholesterol by the fact *hat even more of this substance has been obtained from lard than from butter." So far as our experience has shown, the addition of butter-fat to our natural “protein-free milk” foods gives them an efficiency quite comparable with that of our milk-food in promoting recovery TY sie and the completion of growth. The exact chemical differences between the adequate butter-fat and the inadequate lard (which determine success and failure respectively i in the food-mixtures employed) are far from being satisfactorily wn. Chemical examination of the butter-fat indicates that "ite effective compo- nent is not a phosphatide or any inorganic substance, inasmuch as nitrogen, phosphorus and ash are lacking in the product em- ployed. It is suggestive to note that in the one case (lard) we are dealing essentially with a fat-mixture deposited in storage depots of the animal organism; in the other, the butter-fat repre- sents the product of metabolic activity and synthesis on the part of the cells of the mammary gland. What, if anything, this dis- tinction between cellular product and reserve fat may mean physiologically, remains to be investigated. The researches which have been devoted in recent years to certain diseases, notably beri-beri, have made it more than prob- able that there are conditions of nutrition during which certain essential, but, as yet, unknown substances must be supplied in the diet if nutritive disaster is to be avoided. These substances apparently do not belong to the category of the ordinary nutrients, and do not fulfil their physiological mission because of the energy which they supply. Funk has proposed the name vitamine for the type of substance thus represented.” Without minimizing the importance of the new field of research _ and the new viewpoints in nutrition which are presented by these 1 Cf. McCollum and Davis: loc. cit., p. 174, who fed cholesterol to rats. 2 The literature on the subject has beeu reviewed by Funk: Ueber die physiologische Bedeutung gewisser bisher unbekannter Nahrungsbestand- ) tte der Vitamine, Ergeb. d. Physiol., xili, p. 125, 1913. 430 Influence of Butter-fat on Growth recent findings, we may nevertheless hesitate to accept the extreme generalizations which have already been proposed on the basis of the evidence obtained largely from the investigation of patho- logical conditions. The statement, for example, that a “tadellose Nahrung” may prove entirely inadequate unless “vitamines”’ are present, at once suggests a series of questions bearing on what is included in the new term. It is still rather early €0 generalize on the réle of accessory “vitamines” when the ideal conditions in respect to the familiar fundamental nutrients and inorganic salts adequate for prolonged maintenance are not completely solved. Speculation is quite justifiable in so far as it directs ail attention to a new phase that needs to be taken into account. p Funk has expressed the belief that the substance which pro- motes growth and must be present in order to avert the cessation of growth, which we have described to occur after a certain period of successful growth on our earlier dietaries, is either identical with, or analogous to, the ‘vitamine”’ which plays the rdéle of an antiscorbutic substance. For this we can as yet find no compelling evidence. Certainly the nitrogen-free butter-fat, so successful in remedying our growth failures, contains no substance chemically _ related to the nitrogenous products which have lately been ered- , ited with this unique physiological efficiency.% Furthermore it 18 Cf. Funk: Ueber die physiologische Bedeutung gewisser bisher unbe- kannter Nahrungsbestandteile der Vitamine, Ergeb. d. Physiol., xiii, p. 130 et seq., 1913. In reviewing our earlier published experiments Funk has erroneously assumed that we secured completed growth with the diets in which the butter component was not yet employed. It is true that the increments in weight were in some cases very noteworthy; butin every instance cessation ultimately ensued before the completion of the normal | progress of the growth or subsequent maintenance. We have never denied the necessity of a growth-promoting food accessory in accord with the claim of Hopkins; and recently we pointed out that the successful partial completion of growth, such as has been obtained in our experiments, may well have been due to a store of the essential compound in the body of the experimental animals at the beginning of the trials. It is by no means necessary to assume with Funk that small quantities of these accessory substances were inadvertently left in our food preparations owing to insuf- ficient extraction with aleohol. Furthermore we cannot agree w/th Funk that the rat is not well adapted to experiments on the physiology of growth. The superiority of this ani- mal has been pointed out by us elsewhere (cf. this Journal, xiii, p, 233, j — T. B. Osborne and L. B. Mendel 431 is well to bear in mind that it is not improbable that the anti- neuritic and antiscorbutic constituents of foods are not identical with the substances alleged to assist in maintaining body-weight.™ Funk" has lately asserted that the simultaneous administration of at least two substances is necessary to produce the curative effect obtained in his previous experiments with the “vitamine”’ fraction from rice-bran or yeast. Voegtlin and Towles” have noted that extracts.of autolyzed spinal cord may be antineuritic, yet be unable to reestablish normal metabolism, #.e., restore body- weight. Butter-fat has shown a further interesting nutritive SUPETIO‘ ie ity over lard. At certain periods of the year, particularly in summer months, we have frequently failed to secure satisfactory growth on the dietaries which proved adeq during the usual period of sixty to one hundred days at other seasons. Occasionally young rats in the stock colony have exhibited a similar “epidemic” of poor growth at the same season. The failures are, however, not common to rats fed on the milk-food; and we have lately observed that the seasonal failure is also averted by the addition of butter-fat to the usual “protein-free milk” food-mixtures."’ i Again, another type of nutritive deficiency exemplified in a form y (of infectious eye disease prevalent in animals inappropriately fed'* is speedily alleviated by the introduction of butter-fat into the experimental rations. The chemical character of the unique “accessory substance” | in butter-fat must be investigated in detail and its possible pres- 1912) and is also apparently recognized by both Donaldson and McCollum and their coworkers. We have found rats to be responsive to changes in diet; and we count it no disadvantage that the experiments must be con- ____ tinued over sufficient time to exclude minor incidental fluctuations. i) 14 Cf. Cooper: Journ. of Hygiene, xii, p. 433, 1912. 1 Funk: Ergeb. d. Physiol., xiii, p. 547, 1913; Brit. Med. Journ., April 19, 1913; Journ. of Physiol., xlvi, p. 178, 1913. 16 Voegtlin and Towles: Journ. of Exp. Pharmacol., v, p. 67, 1913. 17 These summer failures in growth have been reported to us by colleagues to occur likewise in other institutes. 18 Cf. Knapp: Experimenteller Beitrag gur Erniihrung von Ratten mit kiinstlicher Nahrung and zum Zusammenhang von Ernihrungsstérungen mit Erkrankungen der Conjunctiva, Zeitsehr. f. exp. Path., v, pp. 147-170, 1908. < Stee = 432 Influence of Butter-fat on Growth ence elsewhere determined. Experiments are already under way with varying proportions of butter-fat in the ration; but we have not thus far determined the necessary allowance. On the other hand, no amount of butter-fat will induce growth on certain die- taries in which the proportions and nature of the inorganic salts are inappropriate (as in our Salt mixture I),!° or the quantity and character of the protein is inadequate. The “Bausteine”’ must not be overlooked in our enthusiasm for these newer features, - AppENDUM. An investigation now under way to determine the possible efficiency of fats other than butter-fat in preventing ecline on our protein-free milk-food and promoting growth in Way that butter does, has already indicated marked differ- ences in fats from different sources. Egg yolk-fat, for example, appears to behave like butter-fat; some other oils have thus far proved no more efficient than lard. Such considerations make it evident that the comparative value of the natural fats em- ployed in nutrition must be determined, as a as the individual role of the different proteins, carbohydrates, anc" iineral nutrients. See Osborne and Mendel: Carnegie Institution of Washington, Pull. cation 156, pt. ii, p. 80. ie T. B. Osborne and L. B. Mendel 433 ea Teg ee ee hy ‘ 2 ‘ , - / / Pd 7 Recovery with Neturp! Protein — j Free Milk + Butter / 20 Days Days u ; Cuart I. Curves of body-weight of rats which have ceased to grow _ and have declined on foods containing the natura ‘protein-free milk,’’ L. und have recovered when 18 per cent of unsalted butter replaced the same ; a 1204, 1268, 1276, 1281, 1292; ovalbumin, Rats 1268, 1276. __ The ordinates represent grams of body-weight, as indicated. The divi- ee - sions of the abscissa represent 20-day periods. Re a a b> ale REN, para tie 434 Influence of Butter-fat on Growth / with Natural tein / F, ik + Butter -faet Ps ’ eA 20 Days ye ae | | eys Cuarr II. Curves of body-weight of male rats which have ceased to grow and have declined on foods containing the natural “protein-free milk;’’ and have recovered when 18 per cent of butler-fat replaced the same quantity of lard in the diet, as indicated by the interrupted lines (-0-0-0-0). The proteins furnished in the different experiments were as follows: casein, if Rats 1224, 1235; edestin, Rat 1391; zein + casein, Rat 1616, \ & The ordinates represent grams of body-weight, as indicated. The divi- sions of the abscissa represent 20-day periods, ; T. B. Osborne and L. B. Mendel 435 “ ee ee 4 4 . - ~ 2 4 ae spa | oo with Natural a Free + a ide 20Dys - Cuart III. Curves of body-weight of female rats which have ceased to grow and have declined on foods containing the natural ‘protein-free : amilk, ”? and have recovered when 18 per cent of butter-fat replaced the same quantity of lard in the diet, as indicated by the interrupted lines (-0-0-0-0). Pa _ The proteins furnished in the different experiments were as follows: casein, “Rat 1217; zein + lactalbumin, 1186. The ordinates represent grams of body-weight, as indicated. The divi- sions of the abscissa represent 20-day periods. 436 Influence of Butter-fat on Growth a ‘ GS Neturdl Protein — “7 WN YY of’ free Mike « [Butter - fot s Cuart IV. Typical curves showing prolonged normal growth of white rats on foods containing 18 per cent of butler-fat. The proteins furnished in the different experiments were as follows: casein, Rats 1592, 1599, 1619, 1636, 1652, 1655, 1657; edestin, Rats 1650, 1654. The ordinates represent grams of body-weight, as indicated. The divi- sions of the abscissa represent 20-day periods. i i i ea T. B. Osborne and L. B. Mendel 437 “@ Tae I owth on Ce ntrifugated Milk 20 Days Days Cuart V. Curves of the body-weight of rats which have ceased to grow _ and have declined on the centrifugated-milk-food, and have recovered when 18 per cent of butter-fat replaced the same quantity of lard in the diet. __The ordinates represent grams of body-weight, as indicated. The divi- ions of the abscissa represent 20-day periods. ee 434, THE EFFECT OF FERMENTS AND OTHER SUBSTANCES ON THE GROWTH OF BURLEY TOBACCO. By J. Du P. OOSTHUIZEN anp O. M. SHEDD. (From the Kentucky Agricultural Experiment Station, Lexington, Ky.) (Received for publication, October 9, 1913.) The growth of the tobacco plant during the first stage is very — slow compared with some of the plants comprising the general farm crops. ‘This is shown by the fact that it, t requires about two months or more in the seed bed to become of sufficient size for transplanting. In other words, it takes the young plant from the time the seed is sown until it reaches this stage, from one-half to three-fifths as long as*it requires after transplanting to grow to maturity. Therefore, it can readily be understood that if some way could be found of shortening the first period of growth until the transplanting stage is reached, this would save the grower considerable time in looking after the seed beds as well as decreas- ing the chances of insects, fungi and bad weather from injuring and destroying the plants. Furthermore, in sections of the coun- try where tobacco might be grown but for the short season, this would be a benefit in assisting to produce a crop. The grower realizes these difficulties and takes advantage of all the best con- ditions for obtaining the largest plants in the seed bed in the shortest time by locating it in virgin soil with the best exposure and applies fertilizers if necessary. A great deal of work has been done in the past few years on the changes taking place in the germination and growth of seeds. It is now generally recognized that these changes depend prima- rily upon what are commonly called ferments or enzymes. The seed, as is well known, contains a certain amount of reserve food material which nourishes the seedling and enables it to grow until it can draw on the soil for its future supply. This stock of food in the seed may be in the form of carbohydrates, oils, fats, pro- tein, ete., and the functioris of the different enzymes are to act 439 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 440 Effect of Ferments on Growth of Tobacco upon their particular part of this food supply and convert it into soluble form which can be utilized by the seedling. F. A. Waugh! has shown that by soaking old seeds of some plants in different enzyme solutions, higher percentages of germi- nation could be obtained than in the same seeds soaked in water. He worked with tomato, cucumber, radish and watermelon seeds of different ages, from five to twelve years old, and obtained in- creased germination in some which had been soaked in diastase solution, given in the following table? Influence of enzymes upon germination. Per cent iption of seeds Solution employed germination me om: 4 years ale . sas 6h Saas 5.6 so son bat 12 | tase GEREN. =... ss os cles 85 Sk tae a Sess 34 ey CO 5 RN tie. 5 ee? 70 ” phate rer 2s See a oa 14 cet. —— RARER i ed Teas see Se WYROMES SO cen ss oe rated 36 ee 5 ee ‘¢ Lsgeges . ce Diastase.......... Sau 46 Cucumber 5 years old................ Water....:........ 50 Mey 44 2 sil 8 6 aes ae SIMStaSOc. | ...).. cee 54. opp. Badish, 6 years old............. gabe NA SR eae 46 ee 6 ee. 9 bene, ene ES Ee ok 66 The tomato seeds above were of different varieties and the plan was followed of soaking the seed in water and the ferment solution for the same time, then draining and transferring them to the germination apparatus. The tentative conclusions reached by Waugh were: (1) In some eases the percentage of germination in seeds is greatly increased by soaking for several hours in a so con- taining some active enzyme or enzymes. (2) The vigor of the young plantlets is often enhanced at the same time. (3) Within limits these beneficial effects increase with the strength of the enzyme solution. (4) Diastase, either from malt or from various commercial prep- arations, seems to be most useful. Tenth Annual Report, Vermont Agric. Exp. Station, 1896, p. 106. * Taken from the Research Bulletin No, 22, p. 105, of the Wisconsin Agricultural Experiment Station. / J. Du P. Oosthuizen and O. M. Shedd = ga (5) Tomato seeds seem to respond especially well to the action of enzymes, particularly to the action of diastase. S. M. Babcock*® has found in a similar experiment that corn, less than 50 per cent of which germinated when the seeds were soaked in water, all germinated when they were soaked in com- mercial diastase solution. The maximum growth was about the same in each lot, but the growth of the seeds soaked in diastase was very uniform, while that of the water lot varied greatly. The increased vitality of the diastase lot was very noticeable. He also found that seeds from the same lot that were soaked for fifteen hours in a 3 per cent glucose solution, instead of water, all germinated, thus confirming the view that lack of suitable — food was the chief reason why the untreated seed germinated poorly. In this case, there was probably a lack of a starch-in- verting enzyme in the seed since equally good results were obtained when either diastase or glucose was supplied. Among other investigators who have obtained good results working in the same manner may be mentioned A. Thomson,‘ who obtained excellent results on the seeds of barley, oats, corn, peas, white and yellow clover, using 5 per cent pepsin and Glestase Fiilitions separately. Others might still be mentioned who have obtained increased germination and vigor of growth of the seed- lings by employing these and solutions of other enzymes, but it is not necessary to mention them here. From the literature at hand, it was found that very few, if any, experiments have been made by supplying ferments or substance which ferments act upon to the growing plant, but the work seems to have been confined chiefly to the germination of seeds and the initial growth of the young seedling. No reference has been found where any work of this kind has been done on tobacco. In view of the fact that in some recent work® the writers have found several enzymes in the growing tobacco and also in the seed, it was thought it might, prove of interest to try the effect of supplying dilute solutions of some of the ferments, or of the materials which they act upon, to the young plants to see if they * Wisconsin Agricultural Experiment Station, Research Bulletin No. 22, p. 106. * Gartenflora, Berlin, xlv, p. 344, July, 1896. ° Journ. Amer. Chem. Soc., xxxv, No 9, September, 1913. 442 Effect of Ferments on Growth of Tobacco would promote their growth. It was thought that the seed might be lacking either in a part of the necessary food reserve material or that one or more of the enzymes might not have sufficient activ- ity to promote the desired changes required by the growing plant. A brief outline of the plan of the experiments was as follows: Before sowing, the seeds were soaked in the solutions of the different substances.and after germinating in the soil, the young plants were supplied with fresh dilute solutions of the same until they had reached the transplanting stage in the greenhouse. In the meantime, observations were frequently made as to the uni- formity, thickness and growth of the plants in the different boxes; and at the end the plants in each were cut close to the soil and weighed so that more definite results could be obtained in regard to what the experiments had demonstrated. ‘ For the work, a sufficient quantity of virgin bluegrass soil was sterilized by heating at 90°-100°C. for thirty minutes to destroy the weed seed. The soil was then thoroughly mixed and equal quantities of about ten pounds were measured into boxes 12X12 X 3 inches in size and supplied with good drainage. Two hundred Burley tobacco seeds that had been cleaned were used for each box and each lot was soaked in the different solutions for twenty- . four hours, filtered and allowed to dry over night before sowing. The amount of the different substances used in the above solu- tions, the manufacturer and the box number in the series were as follows: No. 1. Burley seed + 2.5 cc. of hydrant water used as a check. No. 2. Same as No. 1. No. 3, Burley seed + 2.5 ce. of 5 per cent peptone solution (Witte). No. 4, Burley seed + 2.5 ce. of 5 per cent diastase of malt solution (Eimer & Amend). No. 5. Burley seed + 2.5 ce. of 5 per cent Taka-diastase solution (Parke, Davis & Co.). No. 6. Burley seed + 2.5 cc. of a nutritive solution containing 1500 ec. hydrant water +1 gram KNO, + 0.5 gram MgSO,°7H,0 + 0.5 gram CaSO, 2H,0 + 0.5 gram CasP.0,. No. 7. Burley seed + 2.5 ec. of 5 per cent glucose solution (Kimer & Amend). No. 8. Burley seed + 2.5 cc. of 5 per cent trypsin solution (Fairchild Bros. & Foster), No, 9. Burley seed + 2.5 ec. of 5 per cent pancreatin solution (Merck, pure), ~~ J. Du P. Oosthuizen and O. M. Shedd = 443 No. 10. Burley seed + 2.5 cc. of 5 per cent papain solution (Eimer & Amend). 11. Burley seed + 2.5 cc. of 5 per cent casein solution (acc. to Hammar- sten). No. 12. Burley seed that had been treated with H.SO,. No. 13. Burley seed + 2.5 cc. of a mixture of equal parts of 5 per cent Taka-diastase and 5 per cent pepsin solutions. No. 14. Burley seed + 2.5 cc. of 5 per cent pepsin solution (Merck, U. 8. e., Viil). No. 15. Burley seed + 2.5 cc. of 5 per cent emulsin solution (Kahlbaum). No. 16. Burley seed without any previous soaking with water. This was used as a check to compare the unsoaked seed. No. 6 was carried on so that a comparison might be made of plants that had been supplied with a solution containing all the necessary plant food. No. 12 was included in order that the plan recommended by Love and Leighty,® of acid treatment of some hard coated seeds? to insure a quicker and better germination, might be tried. This treatment was as follows: A quantity of seeds were soaked with five or six times their volume of H2SO, (sp. gr. 1.84) for fifteen minutes. Water was then added and decanted quickly, so as to prevent the seed from heating. The seeds were then washed with water until free of acid by litmus paper test and dried. On January 15, the seeds were mixed with sand and sown in their respective boxes and to each were added 800*cc. of water. After this date, the boxes were watered frequently, adding each time 500 ce. of a fresh 0.01 per cent solution of the substances given above to the respective boxes, except No. 13 to which were added 500 cc. of a solution containing 0.005 per cent of each fer- ment. Of course No. 6 was watered only with the nutritive solution and Nos, 1, 2, 12 and 16 with water. To each box, however, the same volume of 500 cc. was added every time. The nutritive solution was not used constantly but alternated with water. Hydrant water was used throughout the experiments. From the beginning Nos. 14, 13 and 11 gave the best growth and Nos. 14 and 11 maintained this lead until the plants were weighed. Nos. 10, 15, 9 and 7 also showed up well from the first. Nos. 6 and 3 made good gains in the last few weeks of the ® Bulletin 312, Cornell Agric. Exp. Station, p. 335. 7 Tobacco seeds were not used in Love and Leighty’s experiments. 444 Effect of Ferments on Growth of Tobacco experiment. After January 25, the strength of each solution ex- cept No. 6, was reduced to one-half of that formerly used because it was thought that this would lessen the chances of fungi devel- oping in the boxes. The boxes were watered on an average two and one-half to three times a week using the same amount of water or solution during the first part of the experiment. Afterwards it was found that the boxes which were making the better growth appeared to retain the water longer than the others. The plan was then followed of adding extra water to the dry boxes so as to keep all at about the same degree of wetness. Consequently the checks and those » boxes which gave the poorest growth had more water added. It is interesting to note that the boxes which gave the better growth of plants were the ones which seemed to retain the mois- ture longer and this was true from the beginning. This was very noticeable during the first part of the experiment on comparing these boxes with the checks. It ld be expected that as the result of the larger growth of the plants and therefore of their larger water requirement that these boxes would have dried out quicker than those containing the smaller plants, but the- Teverse... seemed to take place. There might be a possible reason for this when the plants attained sufficient size to shade the soil and enable it to hold the water, but even then it would more than likely be counterbalanced by the increased size of the plants. Certainly this explanation would not: hold true when all the plants were very small. While the foregoing from frequent observations appeared to be true, further work is required to prove this inter- esting point. About the first of March, approximately the same number of plants in each were transplanted and arranged in their respective boxes so that they would be more evenly distributed in the soil. On March 15, the plants in several of the boxes were of sufficient size to be transplanted. At this time, No. 14 had the largest plants. These appeared to be more uniform in size and thicker in the box. Next in order were Nos. 10, 11,and 15. Then’ came Nos. 3, 6,7, 8 and 13 with the plants as evenly distributed and of uniform size as in the boxes mentioned above.but somewhat smaller. Next in order was No. 16, and finally Nos. 1, 2, 4 and a J. Du P. Oosthuizen and O. M. Shedd 445 5 had about the same appearance. In Nos. 1, 2, 4, 5 and 16 the plants did not come up with an even stand and were as a rule much smaller than the rest. No. 16, however, made a very good growth towards the end and gained on some of the other boxes. This may have been due to the fact that the plants in this box, not being as thick as those in some of the others, had more room in which to grow. On March 21, the plants were cut off close to the ground and those in each box weighed separately in order to see what differ- ences would be shown in the individual weights. The results were as follows: No. Grams No. Grams No. ceil ; 1 (check).........: 55 60s. Ae ti) — 139 2teneck):.).. 1.923 92 Bs, vd 116 | | oe —* EY 2 | ota te 113 ae 0d 116 1 es 91 a QGetEEL.........;.... 174 a 89. 10.8) en ea 138 me 16 (check)....... 120 * Not weighed: only two seeds germinated. The above weights of the plants in the different boxes are in fairly close agreement with the conclusions reached as to their _ general appearance before cutting. The differences between the _ checks Nos. 1 and 2 are larger than desirable but differences like these will occur in work of this kind and are difficult to explain. The differences between Nos. 1, 2 and 16, all check boxes, are -again large and in favor of the unsoaked seed (No. 16), which is contrary to what would be expected. One explanation, although hardly a plausible one, as to why this may be true, is that soaking the seed before planting may have extracted some soluble reserve material from it which the plant needs for its growth. The dif- ferences between Nos. 1 and 2, according to this supposition, may have been caused by the water extracting more of this material from one lot of seeds than from the other. The growth of the plants in Nos. 1, 2 and 16 was in harmony with this theory since No. 16 maintained the lead over the others almost from the be- The proper checks on this series are Nos. 1 and 2 and, taking the better or No. 2, the differences between this and some. of the best, for example, Nos. 11, 14 and 15 are very large. In only 446 Effect of Ferments on Growth of Tobacco three, Nos. 4, 5 and 9 were the results about the same as this check while the remainder show consistent gains over it. The above results may be due to some fertilizer constituent that was added in the solutions. On the other hand, they may be due to some deficient substance or ferment that was supplied to the plants. As it is almost impossible to obtain the active ferment free from protein material, it was thought that the total nitrogen in the substances used would probably explain the differences ob- tained in the growth of the plants. Accordingly, nitrogen deter- minations were made on all the materials by the modified Kjeldahl method so that any nitrate nitrogen, if present, would be included. For convenience, the materials used, the weight of the green plants, the increase over check No. 2, the amount of substance supplied and of nitrogen in the same, are given in Table I. ee a TABLE I. 4 5 &Z |e lee ne | 88 §-} &8 | weteur or sussTance z ae SUBSTANCE | gE ro ADDED g a a ERE) ESE bb) o82 grams \per cent grams per cent) gram Mo. 1. Chtek, ey... | 85 | me. 2. Cepek ised ss «92 mo. 3. Pentone.:) ices... a: | 113 | 22.8] 0.550 No. 4. Malt diastase......... 91 —1.1* 0.550 No. 5. Taka-diastase........ 89 |-—3.3*| 0.550 No. 6. Nutritive solution... 128 | 39.2 5.000 of KNOs No. 7. Glucose..............) 116 | 26.1] 0.550 No, S:arypein.tos... sae eka? 38.0) 0.550 No. 9. Pancreatin........... | 96; 4.3} 0.550 No, 10; -Papain :tya5. .:e | 126 | 37.0 0.550 No. 11. Casein...............| 189} 51.1] 0.550 No. 12. Seed treated with | | { Tee icocee | Taka-diastase . | (0.275 No. 13. \ Pepsin } ‘ 116 26.1 os No. 14. Pepsin............+..| 174 | 89.1 | 0.550 No. 15. Emulsin.............| 188 60.0) 0.550 “ Nopeao. Checkiccs....sasae:. | 120 | ; J. Du P. Oosthuizen and O. M. Shedd = 447 CONCLUSIONS. After a period of two months, it appears from the above table that marked differences can be observed in regard to the growth’ made by some of the plants. For instance, pepsin, casein and emulsin made larger and more uniform plants than therrest. It is interesting to note the good results obtained from the use of emulsin, and in this connection it might be mentioned that good tests were obtained for this enzyme in the Burley seed and grow- ing plant in some previous work. In every case, with the exception of two, all the boxes were better than the two checks, Nos. 1 and 2, which properly belong to the series and one-half of the boxes were better than the third _ check, No. 16. As arule, the proteolytic ferments and the protein substances gave the best growth. The results obtained with _ trypsin and pepsin are contrary to what would be expected since it is doubtful if the latter is present in plants. Pepsin is regarded as an animal ferment and works best in an acid medium. No posi- tive results were obtained with the diastase ferments. The results in all cases cannot be explained from the total nitrogen added, since pepsin, which gave the best growth, has only 5.14 per cent of nitrogen, whereas some of the others which did not have so great an effect, as pancreatin, peptone and trypsin, - contain respectively 12.58, 15.88 and 12.26 per cent of nitrogen. Again, on comparing pepsin which gave 89.1 per cent increase of growth with malt diastase which showed a decrease of 1.1 per cent, we find that the total nitrogen in the two samples is 5.14 per cent and 6.04 per cent respectively with the advantage in favor of the latter. To No. 6 was supplied in the nutritive solution a total of 5 grams KNO;; 2.5 grams MgSO,;°7H2O; 2.5 grams CaSO,:2H2O and 2.5 grams CasP2,Os. This box was certainly supplied with abundant plant food, and while it made a steady growth through- out the experiment, nevertheless pepsin, emulsin and casein made better growths with much smaller amounts of nitrogen supplied. Papain, trypsin and the nutritive solution gave about the same results yet the nitrogen added to each was 0.0086, 0.0674 and 0.6930 gram respectively. The question of the availability of the nitrogen in the various 448 Effect of Ferments on Growth of Tobacco substances might explain some of the differences. In this case, the nitrogen supplied in the form of potassium nitrate in the nutritive solution should be readily available. Again, nitrogen in peptone would likely be regarded as more available than in casein yet casein gave a much better growth with less nitrogen added. * Lack of an opportunity prevented any further work from a done on the samples to find if any other fertilizer constituent had been added in appreciable amounts. Granting that some potassium and phosphorus were added in these substances, on the other hand considerably larger amounts of these elements in ~~ form were added in the nutritive solution. acid treatment of tobacco seed according to the matted used was not satisfactory since only two germinated. The acid was either too strong or the time allowed for soaking the seed too long and this permitted the outer coats to be broken up and partly dissolved, thus allowing the e strong acid to come in con- tact with and kill the embryo. The substances used in this work were recently obtained from the manufacturer but no previous tests were made on the differ- ent ferments to determine their degree of activity. stem The plants did not grow as rapidly in the greenhouse as cig generally do in the plant bed. When they were cut, the plants in some of the best boxes mentioned above which had reached the transplanting stage were two to three weeks in advance of checks Nos. 1 and 2. While these experiments are only preliminary, nevertheless some of the results are interesting inasmuch as they indicate benefits that might be acquired in practice. It is intended to continue this work when an opportunity affords and compare these and other materials i in, which active enzymes are present and the same substances in which they have been destroyed by heat. Parr II. Many experiments have been made by different investigators on the effect of adding certain substances known as catalytic fertilizers to plants to promote their growth. These substances are generally spoken of under the above title inasmuch as growth ore J. Du P. Oosthuizen and O. M. Shedd = 449 is promoted by their presence and since, as is assumed, they do not necessarily enter into the chemical composition of the plant, they cannot be regarded as plant food materials. They are also referred to as plant stimulants. Among the substances that have been commonly used are salts _of certain metals, such as iron, aluminium, manganese, etc., and while it is only necessary to refer briefly to the literature in regard to the work which has been done along this line, the results of numerous experiments might be mentioned where marked differ- ences in the growth of different plants were obtained on the addi- tion of certain substances to the soil. This is especially true in regard to the use of manganese salts. The work is in the experimental stage but it opens up an in- teresting field of investigation in view of the fact that these metals are commonly found in plants and in much larger amounts in the soil. No importance has ever been attached to them other . than the fact that one or two are included in the list of the essen- tial elements required by plants, but since they are found in con- siderable amounts in the soil, it has been assumed that it is not necessary to add them in practice. Since Euler’ mentions the fact that certain substances (for example iron or manganese salts) accelerate the action of some enzymes and as several ferments have been found in the tobacco, it. occurred to one of us (Shedd) that it might prove of interest | to try the effect of some of these substances on the growth of Burley tobacco plants. It is possible that the good effects of the catalytic fertilizers may be partly due to an acceleration of the enzyme activity in the plant. For these experiments, iron and manganese salts of citric, malic | and oxalic acids were used because it was thought that since these acids have been found in tobacco, their salts would not have a : tendency to retard the growth of the plant. BeSides the above Mictancee, weak solutions of hydrocyanic acid and potassium | cyanide were used in the series in two boxes to determine their | effect on the plant. As emulsin has been found in tobacco, it _ was thought that the addition of these substances might prove Interesting since it is generally recognized that smal] amounts of | 8 Euler-Pope: General Chemistry of the Enzymes, 1912, pp. 108-10. 450 Effect of Ferments on Growth of Tobacco strong poisons sometimes act as powerful stimulants. Amygdalin was also tried alone in this series. As a continuation of the pre- ceding series, mixtures of the substrate and the ferments were also included and finally another trial was made of the sulphuric acid treatment of the seed, except in this case the seed was soaked for only three minutes instead of fifteen as before. The experiments were carried out in the same manner as those in the preceding series, except the seed was not soaked before planting. The same kind of soil prepared as previously described was used. Two hundred seeds, from the same lot used before, were sown in each box. » The materials used in the series were as follows: No. 1. Check. No. 2. Casein + trypsin + pepsin. No. 3. PAancreatin + peptone + glucose. No. 4. Sulphuric acid treatment of seeds. No. 5. Iron and manganese carbonate (Merck). No. 6. Iron and manganese peptonate (Eimer & Amend). No. 7. Manganese citrate (Merck). No. 8. Iron malate (Eimer & Amend). No. 9. Iron citrate, U. S. P. (Eimer & Amend). No. 10. Iron (ous) oxalate, pure (Eimer & Amend). No. 11. Potassium cyanide (Eimer & Amend). No. 12. Hydrocyanic acid (Eimer & Amend). No. 13. Iron and manganese lactate (Eimer & Amend). No. 14. Amygdalin. The seeds were planted February 5, 1913, and the boxes were watered each time with 500 ec. of the following solutions made with hydrant water from the above substances. The substances given above which were used in the preceding series, were taken from the same samples used before. No. 1. 500 ceghydrant water. No. 2. 500 ce. of a solution containing 0.0025 per cent of each substance. No. 3. 500 cc. of a solution containing 0.0025 per cent of each substance. No. 4. 500 ec. hydrant water. No. 5. 500 ce. of a 0.01 per cent solution. No. 6. 500 ce. of a 0.01 per cent solution. No. 7. 500 ec. of a 0.01 per cent solution. No. 8. 500 cc. of a 0.01 per cent solution No. 9. 500 ce. of a 0.01 per cent solution. No. 10. 500 cc. of a 0.01 per cent solution. J. Du P. Oosthuizen and O. M. Shedd = 451 No. 11. 500 ce. of a HCN solution containing 1.03 parts HCN per million, No. 12. 500 ce. of a KCN solution containing 2.47 parts KCN, 7.e., 1.03 parts HCN per million. No. 13. 500 cc. of a 0.01 per cent solution. No. 14. 500 cc. of a 0.005 per cent solution. In Nos. 2, 3, 11, 12 and 14 the solutions were made fresh each time, while in the others stock solutidns of 2500 cc. were made as needed. From the beginning, Nos. 3, 10, 11 and 12 were much better in appearance than the others and No. 10 had considerably more seeds to germinate in less time than the rest. Only a few seeds, not over eight or ten, germinated in No. 4 and this box was dis- carded. : _ On February 28, Nos. 3, 9, 10, 11, 12, 13 amd 14 were better than the check but unfortunately some mice destroyed most of the large plants in these boxes, especially No. 10 in which the tops were gnawed close to the ground. On March 6, the strengths of the HCN and KCN solutions were increased to 2.06 parts HCN per million and from this time about every three days, the strength was increased about 2 parts per million each time until on March 18, the solutions contained 10.3 parts HCN per million and these were used until the end of the experiments. _ On March 24, the plants were transplanted and arranged in the boxes so as to evenly distribute them in the soil. At this _ time, in general appearance, No. 11 was in the lead; next in order were Nos. 6 and 12; then Nos. 3 and 13 were about the same; Nos. 2, 5, 7 and 10 were about like the check and Nos. 8, 9 and 14 appeared to be behind it. | On April 11, the plants in the best boxes were of sufficient size _ for transplanting. At this time they ranked in general appear- ance as follows: Nos. 6 and 11 were the best. The plants in No. 6 were moré uniform and probably were a little better than No. 11, although the latter had some larger plants. Next in appearance were Nos. 3, 12 and 13. The plants in these boxes were not so very uniform especially in No. 12 which had some | large plants and some very small ones. Nos. 6 and 13 had made _ good growth during the preceding few days. Next in order were | 452 Effect of Ferments on Growth of Tobacco 7 | Nos. 7 and 10 which were fairly uniform and looked slightly better — than No. 1. Nos. 2 and 8 appeared about the same as the check. There were fewer plants in No. 8 but they were larger than those | in No. 1. In No. 2 the plants were uniform while in the check they were not, which made the comparison very difficult. Nos. 5, 9 and 14 did not show up as well as the check. | On this date, the plants were cut close to the ground and weighed. — For purposes of comparison, the weights of the green plants, the - gain or loss in weight compared with the check, and the amounts of the different substances and nitrogen added in the same are given in Table II. | TABLE II. Ba i2.-| 8, | ae tee me |eas| ge pet NUMBER AND SUBSTANCE oe ze cfs g58 oes Weg Geb; RES | BG9) eee " _ | grams |per grams per cent} gram 1. Cheok.3:... .ccteses.. 3 Se 95 Caseini.:. ... Sects... > cee 0.2625 14.64 (0.0384 2} Trypan... 2. 8.) } 92 | —3.2| { 0.2625 /12.26 0.0322 Porsiti::.. 270 ooo Se 0.2625 | 5.14 ).0135 [Fete odes s >». ae (0.2625 12.58 (0.0330. 3. | Pepboie.. ::si¢ dep:

C =O with a minute concentration of the form > —O NG O-H which, with H and OH, gives the hydrate x . and this, in a H 1 455 i oe THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 456 Glycid and Acetole in the Organism manner analogous to that shown by methyl alcohol, dissociates with the loss of water into hydroxy methylene and water, thus: H H H H mK = »e +H,0. The hydroxy methylene molecule with H H an excess of available H becomes methyl alcohol; with an excess of OH (or O), formic acid; with H and OH in equal quantity it regenerates the aldehyde, etc. In short, formaldehyde first dissociates and the fate of the dissociated particles depends upon the character of the reaction mixture in which they. find themselves. Alkali, heat, light, certain enzymes, etc., serve to increase dissociation.! i tiplication of experiments makes it possible in many cases to ize with considerable definiteness the actual configura- tion of the unsaturated residues concerned. Yet in the interpre- tation of metabolic phenomena this system of chemical thought has found little application. Ri ‘holds for incorrect Neubauer’s conception® of oxidative deaminization of alanine and his belief that when alanine is broken down in the body, pyruvic acid is the main first ‘‘fassbares”’ product. This dissent is based on the fact that alanine in a fully phlorhizinized dog goes over almost quan- titatively into sugar while pyruvic acid does so in some cases only we to a limited extent if at all. But pyruvic acid, according to — Embden, will yield lactic acid in a perfused liver and lactic acid will go over quantitatively into sugar in a phlorhizinized dog. The point which looms clearly out of these experiments would seem to us to be that alanine is first dissociated and that the dis- sociated residue may make lactic acid under conditions in which there is an equal concentration of OH and H, 7.e., when free O is deficient (Embden’s perfusions), or form pyruvic acid where O is in excess and the whole bodily equilibrium is not upset in the direc- tion of glucose (Neubauer’s oxidative deaminization) or form glucose (via methyl glyoxal or lactic acid?) when the rapid with- drawal of glucose from the body upsets the entire chemical equi- librium of the metabolites in this particular direction (phlorhizin diabetes). Of. Liebig’s Annalen, coexxxv, p. 191 et seq. *This Journal, xv, pp. 145-52, 1913. * Deutsch, Arch. f. klin. Med., xev, p. 211. J. R. Greer, E. J. Witzemannand R. T. Woodyatt 457 One might assume, with a mass of chemical data to support the view, that alanine is dissociated in at least two ways, as follows: egg 7 R\ OH | kf Sy | CH, — C —- COOH CH, — CHNH; — oy 0: (a) (b) Molecule a with H and OH gives lactic acid ;‘ with 20H or O, pyruvic acid. Molecule b represents the phase which with an excess of H permits the reduction of the carboxyl group by addition of 2 H’s and subsequent loss of H.O, ete. Again Embden assumes that because the trioses (e.g., glyceric aldehyde) are capable of forming lactic acid by the action of sur- viving muscle, leucocytes, blood plasma, etc., that ‘ Milchsiiure ein auf dem Hauptwege des Traubenzuckerabbaus gelegenes Pro- duct sei.’”®> But all of Embden’s experiments were carried out under what may be termed asphyxial conditions, and prove only that when there is a lack of free oxygen lactic acid is a chief product of the _ breakdown of the glucose. An asphysiated alkaline glucose solu- tion in vitro also forms lactic acid, but the latter is certainly no intermediate in the fully oxygenized alkaline glucose solution since no lactic acid appears in a fully aerated alkaline sugar solution, whereas preformed lactic acid is not destroyed when added to such amixture. This experiment, performed by Meisenheimer and by Nef, we have amply confirmed. Glyceric aldehyde is rather first dissociated, and whether the residue burns or goes into lactic acid via methyl glyoxal will depend upon the conditions of the experiment. The following experiments with acetole and glycid were per- formed for the purpose of studying their behavior in the body apart from any theory but with special reference to the appa bility of the above-mentioned views. *The simultaneous occurrence of types @ and 6 with intramolecular ' rearrangement can yield methyl glyoxal—which by a benzilic acid rear- rangement may give lactic acid. 5 Biochem. Zeitschr., xlv, p. 108, 1912. 458 Glycid and Acetole in the Organism Glycid. O ae Glycid, CH,OH —-CH—CHhp, is an internal ether, representing a class of substances which, so far as we know, has not received attention from the biological standpoint, owing perhaps to the difficulties attending their preparation. Since the purely chemical behavior of these substances resembles that of the ordinary ethers, a parallelism might be anticipated in the body. Glycid does not reduce Fehling’s or other sugar test solutions’ although it is gen- erally stated to the contrary in the literature, and when boiled o parts of water for eight hours, passes over into glycerol. sa ~ Sie implies a preliminary rupture of the ring with the intermediate exigience of the unsaturated residue , O | i (A) CH.OH—CH—CH, or (B) CHLOH—CH—CH2 or both. Consequently if the ring in glycid were opened in the body, the biological effect of these residues might be determined. When glycid (or glycidic acid) is treated with halogen acids, the halogen is added chiefly to an end C atom, giving " CH,OH—CHOH—CH.Cl (or COOH—CHOH—CH,Cl) which — is interpreted most simply by assuming that under these condi- tions the dissociation which yields the molecule B is the pre- dominating one. Furthermore when glycid is heated (450°) it forms acetole Ch Sa CH; which is also eee to the 1 B residue, thus: CHLAOH—-CH—CH.— CH.OH—C—CH; (Nef). If in the body the ring remained intact, the biological effects should resemble those of ethers in general. As a matter of fact the bio- logical behavior of glycid is in harmony with the idea that the ring is but little attacked in the body. Material. The glycid used was prepared by treating a-chlor- hydrin with alcoholic KOH in the cold, and subsequent fraction- ation. It was a clear glycerol-like fluid of sp. gr. 1.10 (Westphal); b. p., 62°C, at 15 mm. pressure, and corresponding otherwise * Nef: Liebig’s Annalen, cccxxxv, p. 282. J. R. Greer, E. J. Witzemannand R. T. Woodyatt 459 entirely with the product described by Nef.’ It had a slight aromatic odor and mild taste. Animal experiments with glycid. Glycid was given subcutane- ously to guinea pigs and rabbits. When 1 gram of glycid was given to a pig of 519 grams body weight (1.9 grams per kilo) no effects were at first apparent. After fifteen minutes the animal ceased feeding and became sluggish. In forty-five minutes it was very dull and could scarcely be aroused. There was slight twitch- ing of the legs at this time, although it lay soporose in the natural position. In two hours it was dead without having changed its position except for the settling as the tone left the muscles. Post mortem examination showed an engorged right heart and lungs, 4 a few subserous ecchymoses, the liver pale, with some fatty infil- tration and parenchymatous degeneration, kidneys congested. Otherwise no change of note. * Dosages of 1 gram per kilo of weight were usually lethal. With 0.4 gram per kilo the animals became dull, but later recovered. In some fatal cases the parenchymatous changes in the liver and kidneys were more definite and the tendency to small hemorrhages ter so that the picture resembled chloroform poisoning. Death appeared to result from failure of respiration. Owing to the tox- icity of glycid, no satisfactory results with phlorhizinized dogs were obtained. The molecules A or B if formed in the body by opening of the ring, would be closely related to glycerol or acetole, neither of which is so highly toxic as glycid. These poisonous effects are accordingly ascribed to the preservation of the ring, 7.e., to the maintenance of the molecular form of an internal ether whose biological effects resemble those of other ethers. Acetole. Acetole (hydroxy acetone) CH,OH—CO—CHs; stands _be- tween acetone on one side and the ketotriose, dihydroxy acetone, on the other. According to Emmerling and Loges,* it is formed when hexoses are fused with caustic alkali and it has accordingly been held by certain writers to be an intermediate in the break- 7 Loc. cit., p. 232. 8 Ber. d. deutsch. chem. Gesellsch., xvi, p. 837. 460 Glycid and Acetole in the Organism down of sugars into lactic acid under the influence of alkali, and a similar réle has been suggested for it in the body. It is the alcohol corresponding to pyruvic acid and methyl glyoxal (pyruvic aldehyde), concerning the former of which discussion is now in progress, and is isomeric with lactic aldehyde CH; -CHOH—COH O and glycid CH,;OH—CH—CHg:. Acetole has not heretofore been directly tested in biological experiments. Its many known reac- tions are conveniently referable to two types of dissociation.’® Vv meatal A ) CH; — CO — CH,OH @ CH;COH + H — COH ‘* acetaldehyde, hydroxymethylene (B) = CH; — CO — CH,0OH = CH; — CO — CH < +H:0 acetyl methylene The occurrence of the former type, A, is supported by the formation of acetaldehyde and metaformaldehyde when acetole is passed through heated tubes (at 450°) and by the formation of acetic and formic acids when acetole is oxidized with mercuric or silver oxide or with chromic and dilute sulphuric acids. The type B is suggested by Nef as the basis for the condensation of acetole and of analogous substances with strong alkali. Moreover when acetole is treated with copper acetate at 130° the copper salt is reduced and cuprous oxide is formed, the acetole passing over mainly into lactic acid. This lactic acid formation Nef ascribes to the formation from molecule B of acetyl formaldehyde (methyl glyoxal), CH,CO—CH<-+ O = CH;—CO—COH, which then by a benzilie acid rearrangement becomes lactic acid CH;—CHOH—COOH. But when acetole is treated with alkali alone, no lactic acid, or only traces, occur. Consequently acetole cannot be an inter- mediate step in the formation of lactic acid by the action of alkali alone on hexoses. This lactic acid comes rather from methyl glyoxal by the process just mentioned—the latter substance being formed when alkali acts on hexoses in the absence of oxygen. It follows from the foregoing that if a fundamental parallelism exists between the behavior of sugars in alkaline solution and in the body, as we have found it practical to assume, then acetole would not be an intermediate in the catabolism of glucose. See- ondly, if the behavior of acetole itself in the body is analogous to * Liebig's Annalen, ccexxxy, p. 250. J. R. Greer, E. J. Witzemann and R. T. Woodyatt 461 that observed in the presence of metallic oxides in weakly alkaline solution, its fate in the body should be chiefly that of acetaldehyde and hydroxy-methylene in accordance with the scheme A. Now if acetole were convertible into lactic acid in-the body by any process at all, the lactic acid so formed would be capable of yielding glucose in a fully phlorhizinized dog. This is equiva- lent to saying that if acetole by its dissociation could yield B particles which were convertible into lactic acid, acetole would be convertible into glucose. A failure on the part of acetole to form sugar would also speak against the dissociation B since this dissociation would yield methyl glyoxal (pyruvic aldehyde) which has been shown by Dakin and Dudley to be a sugar former. — al The experiments with acetole warrant the conclusion that this substance is notasugar former when given either by mouth or sub- cutaneously to phlorhizinized dogs, and not an intermediate between CsH,.0,5 and C3H,O; (because all CsHsO; compounds are sugar formers regardless of their configuration; dihydroxy acetone [Mostkowski], glycerie aldehyde [Woodyatt], lactic acid [Lusk], hydracrylic acid). The results are most simply explained by as- suming that acetole dissociates in the body in accordance with the scheme A, viz., CH;CO—CH.OH = a sh + CHO, correspond- A H ing to that seen in the presence of weakly alkaline metallic oxides in vitro. ; In the experiments in which acetole was first given to phlor- hizinized dogs it was noticed that following its administration the urine gave the characteristic Gerhardt reaction with ferric chlo- ride in dilutions twice as high as before. A definite increase was also noted in the difference between the polariscopic and titration estimations of sugar. This suggested an increase of the acetone bodies and in another experiment (II)—in which these were fol- lowed—the suspicion appears to have been confirmed, although the rising acidosis may have been a mere incident in the diabetes. An increase of acetone bodies may be attributed to aldole formed from acetaldehyde.’ Material. Acetole was prepared from bromacetone and sodium formiate in accordance with the method of Nef." The product. © Cf. A. Magnus-Levy: Arch. f. exp. Path., xlv, p. 433, 1901. 1 Liebig’s Annalen, cccxxxv, p. 260, et seq. % « 462 Glycid and Acetole in the Organism has a sweet taste and was colorless when freshly made but ‘devel- oped a straw color on standing. That used for the experiments was freshly distilled, and the portion used which passed over between 53° and 56° at 20 mm. pressure. For its detection in the arine we made use of the following properties: (a) volatility; (b) power to reduce metallic oxides in alkaline solution in the cold; (ce) formation of a hydrazone (with phenylhydrazine); (d) lack of optical activity; (e) failure to give a color with Schiff’s reagent for aldehydes. Methods. Glucose in the urine was determined by polariscope and by the method of Bang and Bohmannsson, nitrogen by Kjel- etone and acetoacetic acid, Messinger; 8-hydroxybutyric affer’s method applied to the ether extract. Phlorhizin was given as described in a previous paper. Preliminary experiments with acetole. Healthy guinea pigs, rab- bits and dogs received as high as 2 grams per kilo of body weight without fatalities. Nevertheless even 1 gram per kilo often pro- duced symptoms. Following the ingestion or stibeutaneous ad- ministration of acetole the urine becomes dark and contains then a little albumin and gives the characteristic absorption spectrum of haemoglobin. No reducing substance was found in the urine of a 10 kgm. dog after the administration cf 20 grams of acetole by mouth, nor any other product of its decomposition. Acetole causes definite injury to the kidneys and this feature is also evi- dent in the two following experiments. Experiment [. Fully phlorhizinized fox terrier. Acetole 10 grams in 20 cc. of water made faintly alkaline with Na.CO; administered subcuta- memes at Siem of fourth 6-hour period. DEXTROSE | | PERIOD | _ ae N | D:N® | REMARKS \Polariscope | Titration | Difference | 1 | 7.56 | 8.44 | 0.88 | 1.60 |-5.62 | Tien. 6:20 7.20 1.00 |} 1.95 | 3:77 If | 5.62 7.2 1.60 | 2.14 | 8.36 | IV 5.60 8 27 2.08 | 4.07. | Urine dark, smoky with (7.27) | (1.67) | | trace albumin. ” 1.00 | 1.41} 3.76 | Urine still darker, con- tains haemoglobin. * DIN based on titration figures for dextrose, Haas ot ae J. R. Greer, E. J. Witzemannand R. T. Woodyatt 463 The urine of period IV reduces Haines’ solution in the cold. The cold reducing power measured by Bang’s solution (one-half hour of contact) corresponds to 1 gram of dextrose. The ability of the urine to reduce in the cold is removed by boiling. The distillate contains a substance which reduces Haines’ solution and silvers the walls of tubes containing ammo- niacal silver oxide,at room temperature. There is no reaction for alde- hyde in this distillate with Schiff’s reagent. With phenylhydrazine and paranitrophenylhydrazine, hydrazones were obtained from the distillate corresponding in appearance with those prepared for purpose of com- parison from pure acetole. The extra reducing substance is therefore attributed to acetole. ExreRiMENT II. Fully phlorhizinized fox terrier, weight 13 kg., which received 20 grams acetole dissolved in water, by mouth, at beginni fifth 6-hour period. " | DEXTROSE S-ny- PERIOD |_ ose ee e's N ; DROX YBU- ‘Polariseope Titration | Difference | TYRIC 11.78 | 43.75 1.97 | 1.86 | ‘7.39 | 9.55 \UI88) 2.33 | 1.86 | 6.39 | 0.053 | 0.226 8.03 7 2.30 | 2.22 | 4.65 | 0.079 0.422 6.00 8.55 2.55 | 2.28 | 3:75 | 0.132 | 0.951 4.65 7.50 2.85 | 1.96 | 3.83 0.275 0.997 During the sixth period in this experiment the dog died. The urine after the administration of the acetole was smoky as usual and the distillate con- tained reducing substance which answered the reaction for acetole as in the previous experiment. Aldehyde was not demonstrable in the urine. SUMMARY. Glycid and acetole have been prepared in pure form and ad- ministered to healthy animals and to fully phlorhizinized dogs. Glycid is toxic. Doses of 0.3 to 0.4 gram per kilo of body weight cause narcosis, accompanied at times by muscular twitch- ing. Larger doses cause death. The effects are ascribed to the ring, which is opened in the body with difficulty. Acetole is relatively non-toxic. Doses of 2 grams per kilo of body weight do not kill, but even moderate doses cause haema- turia and haemoglobinuria. When given to phlorhizinized dogs either subcutaneously or by mouth, acetole causes no output of extra sugar. Some unchanged acetole may appear in the urine 464 Glycid and Acetole in the Organism and so raise its total reducing power. There is an apparent rise also of the acetone bodies. The behavior of acetole in the body is explained on the basis that it dissociates into acetaldehyde and hydroxy-methylene. Acetole is not a normal intermediate between substances of the formula CsH».O, and those of the formula C3H,QOs. THE IODINE CONTENT OF THE THYROID AND OF SOME BRANCHIAL CLEFT ORGANS. By A. T. CAMERON. (From the Department of Physiology and Physiological Chemistry, University of Manitoba, Winnipeg.) (Received for publication, November 10, 1913.) It is still a matter of uncertainty whether iodine is an invariable constituent of the thyroid gland, though all fresh evidence in- creases the probability that this is the case. Iodine is unques- tionably present in the glands of most individuals and of most species. Numerous observers have confirmed its presence in the thyroids of man, cattle, sheep, swine, dogs, cats, and rabbits, and it has been observed further in the glands of stags, deer, goats, foxes, the stone-marten and pine-marten, guinea-pigs, hares, fowls, the African tortoise, the dogfish and skate. In some instances negative results have been obtained. Bau- ‘mann, who first observed its presence in the thyroid, obtained in man one negative case in ninety-one, in children twelve negative cases in thirty-nine, and in dogs two negative cases in nine.! Miwa and Stéltzner? stated that iodine is absent from the thyroids of normal new-born children. Roos* obtained negative results in the case of three foxes, three out of four stone-martens, one of two pine-martens, one pole-cat, four of nine domestic cats, four of fourteen dogs, and three of seven pigs. Charrin and Bourcet* obtained negative results with thyroids from eighteen of thirty- two children, and attributed the absence of iodine to their patho- logical condition. Mendel’ obtained four negative results from 1 Baumann: Zeitschr. f. physiol. Chem., xxii, pp. 1-17, 1896. 2 Miwa and Stéltzner: Jahrb. f. Kinderheilk., xlv, pp. 83-8, 1897. 8 Roos: Zeitschr. f. physiol. Chem., xxviii, pp. 40-59, 1899. 4Charrin and Bourcet: Compt. rend. de Acad. des Sci., cxxx, pp. 945- 8, 1900. > Mendel: Amer. Journ. of Physiol., iii, pp. 285-90, 1900. 465 466 Iodine Content of Thyroid six children. Reid Hunt and Seidell® obtained iodine-free thyroid from children, maltese kids, an Alaskan bear, and an aoudad. In considering these negative results the analytical method must be taken into account. In all these cases the iodine was estimated by some variation of the method devised by Baumann,’ in which after fusion with sodium hydroxide, oxidation with nitrate, and subsequent dissolution, the iodine was extracted with chloroform, or carbon disulphide, and estimated colorimetrically. It has been shown by Seidell® that this method frequently yields too low results, and it can therefore be inferred that minute quantities of iodine will frequently escape detection when it is employed. aii ~~ + Seidell® in referring to the “negative’’ amounts in their mat state explicitly that minute quantities may be present, but not detectable by their analytical method. No stress can be laid on any negative results so far published until they have been con- firmed by an accurate method at of Hunter.’ ‘In connection with the negati* es for children’s thyroids, the results of Fenger,'' who has elapipyed Hunter’s method, show clearly that iodine is normally present in the thyroids of foetuses and young of cattle, sheep, and swine, and in amount relatively comparable with that of thyroids in adult animals. It is probable that similar results will be obtained with children’s thyroids when the same method is employed. It is well recognized that there is marked variation in the iodine content of the thyroids of individuals of all species for which data have been obtained, but that in spite of this variation of individ- uals there is a definite relationship to diet exhibited. Thus the negative results quoted, even if they really indicate presence of slight amount of iodine, support Roos’ assumption that the thy- roids of carnivorous animals contain much less iodine than those of herbivorous species. Baumann’s negative results with dogs * Hunt and Seidell: Bulletin No. 47, U. 8. Hygienic Laboratory, p. 33, 1908. ? Baumann and Roos: Zeilschr. f. physiol. Chem., xxi, p. 489, 1896. * Seidell: this Journal, x, p. 95, 1911. * Hunt and Seidell: Bulletin No, 47, U. 8. Hygienic Laboratory, p. 33, note (a). ” Hunter: this Journal, vii, pp. 321-49, 1910, " Fenger: ibid., xi, pp. 489-92; xii, pp. 55-60, 1912; xiv, pp. 397-405, 1913. A. T. Cameron 467 were obtained after lean meet had been fed for some time, while it is now recognized that if any assimilable iodine compound be fed the iodine content of the thyroid is increased. There is strong presumptive evidence that the diet of all forms of sea-life is unu- sually rich in iodine, and I have recently shown that the thyroid of the dogfish (Scyllium canicula) contains an amount of iodine relatively greater than any previously recorded (1.16 per cent for dry material from female fish). | It seemed desirable to extend the observations on iodine content to as many different classes of animals as possible, with the object of increasing the evidence both as to the invariable concomitance . of iodine with thyroid tissue, and as to its variation in diffe classes consequent on their different diets. I have now obtained positive results with the thyroids of the pigeon, the alligator, the leopard frog (Rana pipiens) and a second speeies of dogfish (A can- thias vulgaris). The results for the pigeon are distinctly high, those for the alligator and frog distinctly low, in agreement with their respective herbivorous and carnivorous diets. In addition some data are included bearing on the presence of, iodine in parathyroid tissue. The sole results suggesting the presence of iodine in parathyroids in amounts comparable with that in thyroids are those of Gley."* He found that in rabbits the absolute amount of iodine was greater in parathyroids; in dogs the relative amount was greater. He employed the Baumann method and any error in his results must apparently be attributed to the small quantities of materials employed. Chenu and Morel'* investigated dogs, rabbits, and fowls, comparing equal quantities of thyroid and parathyroid from the same animal. They con- cluded from their results that the parathyroid contains iodine, but to a less extent than the thyroid. Doyon and Chenu” found that the parathyroids of the African tortoise contain little or no iodine. Estes and Cecil'® obtained negative results with the glands of the cow, horse, sheep, and man. Infinitesimal amounts present in % Cameron: Biochem. Journ., vii, pp. 466-70, 1913. 8 Gley: Compt. rend. de l’ Acad. des Sci., exxv, pp. 312-5, 1897. 4 Chenu and Morel: ibid., exxxviii, pp. 1004-7, 1904. © Doyon and Chenu: ibid., cxxxix, pp. 157-8, 1904. © Estes and Cecil: Johns Hopkins Hospital Bulletin, xviii, pp. 331-2, 1907. , : 468 Iodine Content of Thyroid two experiments with dogs’ and one with horses’ parathyroids were attributed to accidental thyroid contamination. I have obtained absolutely negative results with the ventral branchial body of the frog—an organ whose function is still un- known. The presence of one or more parathyroids in close juxta- position probably resulted in the removal of these organs with the ventral branchial body, so that the results may bear also on the iodine content of these organs.’ I have also carried out a com- parison between the parathyroids and thyroids in a series of dogs. The results show at least a marked differentiation, while the actual amount found in the parathyroids may be completely attributable to the almost unavoidable contamination with thyroid tissue incident on the removal of the internal parathyroids in the dog. The results as far as they go support Estes and Cecil’s conclusions that the parathyroids do not contain iodine. I have used Hunter’s method throughout. It has been tested and found accurate for ordinary amounts of iodine by numerous observers. I found some difficulty at first in obtaining perfectly negative results in known tests where iodine was absent, but found that, where the quantities of material analyzed were not greater than 0.5 gram, and after combustion, solution, and chlo- rination, the not-too-acid solution was boiled vigorously for at least one and one-half hours, the quantity of liquid being kept throughout between 150 and 300 cc., blank tests invariably gave perfectly negative results. Numerous tests with known quantities of iodine proved satisfactory. Hunter claims that his method will detect and approximately measure 0.01 mgm. of iodine (0.002 per cent of 0.5 gram). An absolutely negative result probably indicates a much lower iodine content than this figure. The pigeon. I have found only a single observation on the iodine content of the thyroid of birds. Chenu and Morel'* compared the thyroids 'T A full account of the anatomical relationships of these bodies and of the thyroid, for which the ventral branchial body must frequently have been mistaken, has been published recently by Mrs. F. D. Thompson: Phil. Trans. (B), eci, pp. 91-132, 1910. # Loc, cit. on ———— . A.T. Cameron — 469 and parathyroids of the domestic fowl. Since they took weights of thyroid tissue equal to those of the parathyroids present, only very small quantities were employed, and the results cannot be other than inaccurate. They are furthermore expressed for fresh tissue. This does not allow easy comparison with other pub- lished data. I have found that such small amounts of tissue can be weighed accurately only with difficulty on account of rapid drying. An approximation to the dry-tissue value may perhaps be obtained by dividing their results by four. WEIGHT PER CENT IODINE " OF THYROID TAKEN AMOUNT OF | IODINE FOUND | f ; | Fresh tissue | Dry tissue 1 year old cock........ 0.019 1 year old cock........ 0.026 (0.014) (0.013) It is doubtful whether these results show even the order of the amount of iodine present. Ihave carried out a series of analyses with the domestic pigeon. Material was obtained from a large number of no certain type. In most cases the pigeons were less than six months old, the thymus being still well developed. The thyroid could be dissected com- pletely from surrounding tissue, and the results are therefore prob- ably correct to within 1 or 2 per cent (allowing for.a trace of un- removed fibrous capsule). The material was dried in vacuo over sulphuric acid in this and all the succeeding analyses. ess on eae | acre, gan cane Moist Dry “| FOUND | DRY GLAND hs OX ie in gram gram : | gram | — oe 23 (from 12 birds)..... 0.236 0.075 0.000412 0.550 47 (from 24 birds)... 0.562 0.135 | 0.000644 0.477 36 (from 18 birds)..... 0.504 0.117 | 0.000530 0.453 Total, 0.327 0.001586 Average, 0.485 470 Iodine Content of Thyreid The alligator. Three thyroids were obtained from three young alligators. each about twelve inches long. The dissection was clean, and the degree of error is determined only by the small amount of tissue examined. . PER CENT IODINE IN DRY TISSUE gram gram 0.0402 0 .0000239 | 0.059 WEIGHT OF DRIED THYROIDS AMOUNT OF IODINE FOUND » results agree with the low figures found generally for us animals. The frog (Rana pipiens). Treupel!® injected Baumann’s iodothyrin subcutaneously into frogs, and in two cases removed the thyroids.(under Gaupp’s direction), and, employing Baumann’s method, considered that he obtained unmistakable evidence of the presence of iodine in the tissue examined. This affords no evidence as to the presence of iodine under normal conditions, although Gaupp claims that it indicates that the thyroids of the frog function as in other verte- — brates. Gaupp himself®® states that he has confirmed the presence of thyroid tissue in the frog (R. esculenta, var. Hungarica) by a chemical test (‘Die chemische Diagnose bestiitigte, dass nicht irgend etwas Anderes, Muskelfasern und dergl. filschlich fiir die Schilddriise genommen war’’), and since he immediately refers to Treupel’s work he presumably indicates the presence of iodine, though I have found no further details of his examination. I have examined the thyroids and ventral branchial bodies ob-_ tained from a large number of frogs (Rana pipiens) bought from — Chieago and Minneapolis dealers during the period September to December, 1912, so that these frogs varied from well nourished to partially nourished individuals. On account of the minute size of the thyroid in the frog it is probable that some surrounding muscular tissue was frequently removed with it. ” Treupel: Miinch, med. Wochenschr., xliii, pp. 885-6, 1896. ™Gaupp: ef. Beker-Wiedersheim'’s Anatomic des Frosches, 111, i, p. 206. UOT ANT A. T. Cameron 471 oe AMOUNT ORGAN | werent | WEIGHT TAKEN | OF IODINE 7 cava % ovidae tae nian Bers gram eae .. 185 thyroids........... 0.2089 | 0.0987 00000727 0.073 0.1102 }0.0000592 0.054 | | Total, 0.2089 |0.0001319|Average, 0.063 en oe eee | 0.2655 | 0.1169; O | | 0.1371| 0 | The dry thyroid material was greasy as though some fatty tissue was present. In consequence sampling was difficult; this probably explains the non-agreement in the two results. They indicate definitely that iodine is present im the frog’s thyroid under normal conditions. As has been Mitooed, surrounding tissue was probably present, so that the figure must be regarded as a minimum one, to the extent perhaps of a from 20.to 40 per cent error. Even with this correction the amount present is distinctly small, corresponding with the known earnivorous habits of the frog. The dogfish (Acanthias vulgaris). As far as I am aware the only data for fish thyroids hitherto published are those I obtained this year for Raza clavata and Scyllium canicula.2! The samples of Raia gave figures varying from 0.283-0.438 per cent. A sample of male Scyllium thyroids gave the figure 0.719 per cent, another of thyroids from female Scyllium the very high figure 1.16 per cent. Through the kindness of Professor E. E. Prince, Dominion Com- missioner of Fisheries, a consignment of Acanthias was obtained ‘last winter from the Atlantic Coast. They were preserved in for- malin during transit, and it was found difficult to dissect the thyroid in the preserved fish, since the tissues had become hardened and discolored. In order to be certain that thyroid tissue was present much of the surrounding tissue was frequently removed, and the figure obtained consequently only indicates the order of the amount present. It was much smaller than that found for Scyllium. In a1 Loc. cit. . #NAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 ~ ay 472 Iodine Content of Thyroid all, 0.752 gram of material (dry) was obtained from a large number of fish. WEIGHT TAKEN | AMOUNT OF IODINE FOUND ae baa IN DRY gram ee i gram F a 0.2015 0.000271 0.134 0.1999 0.000263 0.131 } | t Average, 0.133 Comparison of the iodine content of the parathyroids and thyroids of ™ the dog. ail a. ‘wenty-three external and ten internal parathyroids were obtained from twelve dogs. They yielded 0.077 gram dry mate- rial which contained 0.0000120 gram iodine corresponding to 0.015 per cent. The thyroids were analyzed with the following results: te WEIGHT OF THYROIDS AMouNE 1 Bere ows i orpos | ah Ss =e THYROID keel IN WHOLE GLAND kgm grams grams gram gram gram 23.0 4.916 1.276 0.505 | 0.0000534 00001349 13.0 4.793 1.379 0.502 | 0.0007797 00021418 16.2 2.119 0.672 0.672 | 0.0017798 _ 00017798 5.0 | 13.093 3.212 0.500 | 0.0004472 | 0 .0028728 3.0 3.977 1.119 0.500 0.0009311 ° 0 0020838 17.5°| 3.12 0.937 | 0.505 0.005003 0 0009450 2.0 | 6.438 1.460 0.503 — 0.0004009 0 .0011636 21.0 | 7.589 1.340 0.500 0.000576 0 0002120 16.5 | 1.178 0.372 0.372 0.0006614 0 0006614 16.0 | 2.684 0.792 0.792: 00005497 00005497 17.5 | 1-501 0.532 0.532 00001822 0 0001822 22.5 | 2.641 0.743 0.743 00009407 | 0 0009407 | Total, 14.334 | | | Total, 0.0136677 |Avers ze, The significance of this result has already been pointed out. SUMMARY OF RESULTS. Iodine is present in the thyroids of the pigeon, alligator, and — frog, in amounts corresponding with the diets of these animals, — It is also present in the thyroid of the dogfish (Acanthias), Pur-_ ' _ A..T. Cameron 473 ther support is therefore given to the theory that it is an invaria- ble constituent of thyroid tissue. All reliable data hitherto pub- lished point to this invariable concomitance. The negative figures obtained by some investigators have led to some doubt as to the bearing of the presence of this element on the function of the gland,” but it seems desirable to reject these negative figures entirely until more certain evidence is available. Iodine is absent from the ventral branchial body of the frog. The amount of iodine present in the parathyroids of the dog is of a less order of magnitude than that in the corresponding thyroids, if indeed the actual quantity observed be not wholly attributable to thyroid contamination. The results, so (C,7H33N He) -—> CysHgCN —> C,sH3,CONH. —_ C,sH;,COOH. Stil] another method has been published by Le Sueur* and Blaise* _ (the methods are the same, the interpretation of the course of the | reaction different). The C, acid is changed into the a-bromo- ! Krafft: Ber. d. deutsch. chem. Gesellsch., xii, p. 1664, 1879. * A. W. Hofmann: ibid., xvii, pp. 1406 and 1920, 1884; E. Lutz: ibid., xix, p. 1433, 1889. ’ Le Sueur: Journ. Chem. Soc., \xxxv, p. 827, 1904; Ixxxvii, p. 1888, 1905. * Blaise: Compt. rend. de lV Acad. des. Sei., exxxviii, p. 697, 1904; Bull. soc. chim. (3), xxxi, pp. 483-93, 1904. 475 476 Decomposition of Fatty Acids and e-hydroxy-derivative, from which the aldehyde of the next lower acid is prepared by heating to 275° for about one hour, car- bon monoxide being evolved during the heating. This aldehyde is then oxidized to the acid. Since the yield of the aldehyde is not over 50 per cent this method is not perfectly satisfactory, though Le Sueur claims it to be ‘‘a ready means of passing from an acid of the acetic acid series to the next lower homologue.”’ In certain cases, where the halide of the C,—: alcohol is easily accessible, as is cetyl iodide, the C,-; acid may be synthesized from the Grignard reagent and carbon dioxide.® _Edmed* has observed that dihydroxystearic acid may be oxidized ea. kaline permanganate at the place where the hydroxyls are attached. This method has been applied to the study of cere- bronie acid,’ where it has been shown that the principal if not the | only product of the reaction is.lignoceric acid. It was then con- cluded to apply this method of oxidation on a larger number of a-hydroxy-fatty acids. It is comparatively an easy task to trans- form a fatty acid into its a-hydroxy-acid. If the permanganate method of oxidation were successful we hoped to apply the process for the study of the structure of lignoceric acid and of other fatty acids, the structure of which is not yet definitely established. Considerable time after the publication of the work on cerebronie acid Lapworth’ made use of the same process for the preparation — of tridecylic from myristic acid. Since then we have had occasion to prepare a considerable quantity of margaric acid and have applied the reaction to a-hydroxystearic acid with equal success. It has further been tried out in the preparation of pentadecylic acid and we now believe that it isa general method for the decom- position of the carbon chain which may be easily carried out with fairly good yields. We obtained a yield of 80-85 per cent of mar- garic acid, calculated on the stearic acid used. It has the advan- tage over Le Sueur’s method in that the preparation of the alde- hyde with its consequent loss is avoided. | *Ruttan: Highth International Congress of Applied Chemistry, xxv, p. 431; Chem. Abstraets, vii, p. 2140, 1913, *Edmed: Journ. Chem. Soc., \xxiii, p. 627, 1898. ? Levene and Jacobs: this Journal, xii, p. 381, 1912; Levene and West: ibid., xiv, p. 257, 1913. * Lapworth: Journ. Chem. Soc., ciii, p. 1029, 1913. P. A. Levene and C. J. West 477 In agreement with Lapworth we find that the reaction is best carried out in acetone rather than in water. The potassium salt of the new fatty acid (especially of the higher acids) is insoluble in ‘ acetone and precipitates with the manganese dioxide, from which it is easily extracted with alcohol. | EXPERIMENTAL PART. _ Lignoceric acid. The preparation of lignoceric acid has been modified to the following: Fifty grams of cerebronic acid are dissolved in about... 1 liter of boiling acetone and this solution treated gradually with | @ warm acetone solution of potassium permanganate until the solution is slightly colored, indicating an e of permanganate. The mixture is then heated a short time on the water bath, cooled, the manganese dioxide and potassium lignocerate and cerobronate filtered off, and the potassium salt extracted with boiling absolute alcohol. Usually two or three extractions, using a liter of alcohol each time, is sufficient. The acid thus obtained is converted at once into the lithium salt and purified with methy! alcohol as be- fore. The acid from the insoluble lithium salt was recrystallized from acetone, when it gave the following numbers on analysis: 0.1232 gram of substance gave 0.3537 gram CO, and 0.1460 gram H,0. Calculated for CosHasO2: Found: ee ee, ates ie. oe aes 78 .20 78 .30 Pe en a ee 13 .20 13 .26 Margaric acid. Stearic acid was converted into a-bromostearic acid by Hell’s method and this into a-hydroxystearic acid according to Le Sueur. This was characterized by the preparation of a-Acetoxystearic acid. Fifteen grams of hydroxystearic acid were dissolved in 100 grams of acety] chloride and the solution heated an hour in a water bath. The excess of acetyl chloride was removed on a boiling water bath, the product treated with water to remove the last ™ 478 Decomposition of Fatty Acids traces of the chloride and hydrochloric acid, extracted with ether, the ethereal solution dried, the ether removed and the product recrystallized from a little absolute alcohol. It is a colorless crystalline body which melts at 70—70.5°. 0.1258 gram of the substance gave 0.3243 gram CO, and 0.1258 gram H,O. Calculated for ' CeoHssO4: Found: Cee eo eo a ES 70.12 70.31 Bho. ........ 2... ch ae ee bee 21/10 11.19 Oxidized with potassium permanganate in acetone as given - above, the hydroxystearic acid gave nearly pure margaric acid, which was purified by two recrystallizations out of acetone. It melted at 59-60°, as given by Ruttan.® Its purity was controlled by analysis: 0.1228 gram of the substance gave 0.3386 gram CO, and 0.1360 gram H,0. 1.0000 gram of the acid, dissolved in absolute alcohol and benzene ° required 37.0 cc. 7g NaOH for neutralization, using phenolphthalein as an indicator. Calculated for CyHaOe: =-Found: Bs ok Glare aac Uc Ss PS eee 75.60 75,.20 |: ee ieee ee eee ee a 12.60 12.40 DA Wc ie aic « s kcsethcnies + css SME Ss SMS ook 270 270 Pentadecylic acid. Pentadecylic acid was prepared in the same manner, starting with palmitic acid. The acid thus prepared melted at 53° and gave the following analytical figures. I. 0.1342 gram of the substance gave 0.3674 gram CO, (H,0 lost). Il. 0.1324 gram of the substance gave 0.3620 gram CO, and 0.1472 gram H,O. 1.0000 gram of the acid, as above, required 41.4 cc, y NaOH for neutral- ization. Calculated for Found: CysHyoO2: I i GO , .. 6 eR > 5 n pM sis onde. cal 74.40 74.67 74.57 DEES, ,'. . 0 ostMMMMNG > ov 0M A space 3 GC 12.40 12,44 SR SRR 242 241 * Ruttan: loc. eit. AUTOLYSIS OF MOLD CULTURES II. INFLUENCE OF EXHAUSTION OF THE MEDIUM UPON THE RATE OF AUTOLYSIS OF ASPERGILLUS NIGER. By ARTHUR W. DOX. (From the Chemical Section of the Iowa Agricultural Experiment Station.) (Received for publication, November 17, 1913.) In a previous Reiser it was shown that when molds are grown upon a fluid synthetic medium, the nitrogen is almost completely taken up by the mycelium during the vegetative period of the fungus and then gradually returned to the medium after the growth has come to a standstill. In the particular medium stud- ied where nitrogen and sucrose were present in the proportion of 1 to 50, most of the nitrogen had disappeared at the end of the first week, and during the subsequent seven or eight weeks a large part of it reappeared, principally in the form of ammonium salts. During this time the mycelium lost its turgidity and the medium became dark in color although it retained its original clarity. This phenomenon was ascribed to autolysis of the fungus. The changes observed both in the mycelium and in the medium were so striking as to be deemed worthy of further study. Subse- quent observations of numerous cultures showed that the rate of autolysis, as indicated by the visual appearance of the mycelium and of the medium, was influenced by a number of factors, such as the volume of the medium as compared with the surface area, the ratio of carbon to nitrogen, the temperature to which the cul- tures were exposed, etc. In other words, the supply of nutriment and the rate of growth seemed to be of primary importance in determining the point at which autolysis set in. As long as the presence of sugar could be demonstrated by Fehling’s test, autoly- sis was scarcely noticeable. 1 Dox and Maynard: this Journal, xii, pp. 227-31, 1912. 479 480 Autolysis of Mold Cultures The experiments herein described were made for the purpose of studying the effect of exhaustion of the carbohydrate in se me- dium upon the autolysis of the fungus. In the first series of cultures the following medium was used: distilled water 1000 cc., sucrose 30 grams, dibasic potassium phos- phate 1.0 gram, magnesium sulphate 0.3 gram, ammonium acid tartrate 4.0 grams, trace of ferrous sulphate. Two hundred cc. of this medium were placed in each of a number of liter Erlenmeyer flasks, sterilized in an autoclave, and inoculated with spores of Aspergillus niger. At the end of a week vigorous cultures with an abundance of black spores were obtained. Two of the cultures eated as follows: The cotton plug was removed, a sterile on inserted and the plug replaced. The medium was by suction, replaced by 200 ce. of sterile distilled water, and the latter removed in the same way after a few moments’ con- tact with the mycelium. This was in turn replaced by 200 ce. of sterile water in one flask and by 200 ce.of a sterile 2 per cent su- crose solution in the other flask, and the cultures allowed to stand another week. This operation was repeated at the end of each week for six successive weeks. All this was done with as little injury as possible to the mycelium, care being taken not to wet the — surface of the culture. Five other flasks were treated in the same — manner after the cultures had grown two, three, four, five and six — weeks respectively, except that the cultures were discarded after the medium and wash water had been obtained. The combined ‘TABLE I, a. b. c Q | MEDIUM REPLACED soe ov comune | WSEKLY BY Waren | epecnons 2% =| “TORDRD CULTURE. mgm. N in mediym mgm, N. in medium mgm, N weeks , a Or | giro | aaa 0 70.0 | | 70.0 — 70.0 1 3.7 me 2.7 2.5 | 2.5 2.4 2 12.4 | 15.1 .iey§ 686 9.4 3 12.8 | 27.9 2.3 7.9 16.9 4 5.9 33.8 1.4 9.3 23.5 5 3.6 37.4 1.0 10.3 26.1 6 3.0 40.4 | 123.6 27.3 Arthur W. Dox 481 6 ¢ A 9 oe 7) 4 a > 2 w 3 « > ur 5 at v w ° ws 2 © -free air may be driven several successive times into nitrometer C by simply opening pinchcock 9, In order to test whether or not the air thus treated is now free from carbon dioxide, the following experiment will be necessary. Remove the glass stopper S, and introduce into the chamber a known amount of mercury by means of the mercury burette H, so that the remaining volume of the chamber A is exactly known.’ * Instead of fusing the platinum electrodes into the wall of the chamber, they were fused into the glass stopper S. It will be clear by inspection of the figure that the wires are brought up high enough so that when the cham- ber is sealed with mercury there will be no short circuit established through mereury and electrodes, in case any electric current is used for the stimula-. tion of the tissue. * The exact capacity of chamber A should be calibrated once for all. [In this way, one can always work with a constant volume in the chamber by Shiro Tashiro 489 Turn stopcock 4 now about 45°, so that the connection between A and H is severed. Replace the stopper S, and seal the chamber with mercury. Keep stopcock 2 turned so that the connection is made only between nitrometer C and chamber A. The alkali in the nitrometer C is displaced by CO--free air by opening pinchcock 9. Collect about 300 cc. of the CO>-free air in the nitrometer C. While the stopcock 1 is closed, the chamber A is evacuated by means of suction, having the stopcocks 5, 6 and 7 opened (the three-way stopcock 6 should be opened in such a way that Ba (OH)> is completely shut off from the connection). When the evacuation is complete, CO.-free air is introduced into the chamber by opening stopcock 1. After the evacuation and washing out with pure air, which is repeated four or five times, the chamber now being filled with CO.-free air, the stopcock 7 is closed, and the pressure inside chamber A is made equal to the at- mospheric pressure by adjusting it at the nitrometer C by means of the alkali bottle D. Stopcock 5 is then closed, and the space between 5 and 7 is again evacuated so that the barium hydroxide can rush in, a process which is very advantageous in obtaining a clear barium hydroxide solution. In filling the tube with the barium hydroxide, it is advisable to open stopcock 6 so that the solution will first fill up the space between 6 and 7, then turn it in such a way that now the connection is made between the barium hydroxide tube and the space between stopcocks 5 and 6. By opening 5 very slowly and carefully, the barium hydroxide is now introduced into the chamber just so far that a small hemispherical drop stands upon the upturned end of the tube atd. After quickly readjusting the pressure by means of the nitrometer and the bottle D, the stopcock 1 is closed. If the air is completely free from car- bon dioxide, the drop should be clear not only at the start, but also, after several hours’ standing, free from any granules of the carbonate, when inspected with the lens. introducing the necessary odd cubic centimeters of mercury, thus making the remaining volume a convenient round number of cubic centimeters. For instance, the apparatus I am using has a capacity of 31.4 cc. I introduced 6.4 cc. of mercury for each experiment, so that the respiratory chamber then contains 25 cc. It will be needless to say that for an experiment to test the air for its purity, the knowledge of the exact capacity of the chamber is not at all necessary. : 490 ~+Estimation of Minute Quantities of CO, 3. For the qualitative detection of carbon dioxide. For the detection of carbon dioxide production from a tissue, this apparatus can be used in exactly the same manner as the pre- vious apparatus. After insuring that the air is free from the gas a given tissue® is placed inside of the chamber and the process is repeated as before.. If any COs is given off by the tissue not only will a deposit of carbonate appear, but it will also grow in size. 4. For quantitative estimation of the gas. ' The detailed method is as follows. Ypen stopcocks 3 and 2 so that they will connect the chamber A and the tube B only. Fill the mercury burette G and raise it till the mercury will completely fill the tube B and a little excess of it will stay in the capillary tube between the chamber A and the stop- cock 2. This stopcock 2 is now elosed so that it will connect the nitrometer C and the chamber A only. Increase the pressure inside the nitrometer C by raising the alkali bottle D much higher than the meniscus of the alkali in the nitrometer C and then open stopcock 1. In this way, the excess of mercury left in the capillary tube will be pushed over into the chamber and will flow through — the stopcock 4 into a receiving vessel. If the stopcock 2 is absolutely air-tight, having no air bubble in tube 3, then a known amount of mercury is introduced into the chamber by means of the mercury burette H, thus giving the respiratory chamber the desired volume. The tissue is intro- duced into the chamber, the glass stopper is replaced, the chamber is sealed with mercury, and the nitrometer C is filled with the pure air. After evacuation of the chamber and washing it with CO.- — free air several times, the stopcock 5 is closed and the time is re- corded; the pressure is adjusted, and stopcock 2 is turned 45°. At the end of the desired respiration period, any portion of the air in the chamber can be driven into the tube B. This is aecom- plished by raising the right hand mercury burette 7, and simul- taneously opening the stopecocks 2 and 4. Stopeock 2 is now The tissue can be placed on a cover slide and allowed to float on top of the mercury or it can be placed on the glass plate and hung on the elec- trodes as described in figure 1, p. 120, loe. cit. Shiro Tashiro 491 closed: the pressure of the air in B is kept equal to the atmospheric pressure by adjusting the mercury burette G. After removing the mercury seal and glass stopper S, the tissue is withdrawn and mercury burette H is lowered so that most of the mercury in the ‘chamber will now flow back into the burette H. The excess of mercury in the chamber A is withdrawn through stopcock 4 into a vessel. In order to analyze the air in the tube B, it is advisable to clean the whole chamber A once more with water,’ and then to perform the experiment in exactly the same manner as we described in con- ‘nection with the test for purity of air. Two things are imperatives namely; the capacity of the respiratory chamber A must be exactly after the known amount of mercury is introduced into it; and the bubble of barium hydroxide solution at d must be per- fectly clear at the start. If no deposit of barium carbonate forms on the surface of the drop within ten or fifteen minutes, it is a sure control that the apparatus is free from carbon dioxide. This point established, a portion of the sample of gas in the tube B is intro- duced into the chamber. This is very easily accomplished by withdrawing the mercury from the chamber A into a small gradu- ated cylinder® and adjusting the pressure by raising the left-hand mercury burette G; then close the stopcock 2 by turning it 45°. One now watches the surface of the drop at d with a lens to see whether any formation of barium carbonate occurs within ten minutes. If it does not, we should introduce more air from B until we get the first visible precipitate."° I have previously de- termined," by introducing accurately known quantities of carbon dioxide of very high dilution into the chamber in a similar manner, and have found with remarkable regularity that 1 * 10-7 gram of carbon dioxide is necessary as the minimum amount to give a 7 For the method of cleaning this apparatus and drying it in ten minutes without taking it apart, see a footnote on page 138, Amer. Journ. of Physiol., xxxii, 1913. 8 See pages 488-89. ®If more accurate measurement is necessary, the mercury withdrawn should be weighed. The detection ofthis precipitate is not a question of degree, but is a question of the appearance of some precipitate or none at all; therefore the end point is very sharp. " See p. 144, loc. cit. Be. 492 Estimation of Minute Quantities of CO, precipitate within ten minutes." Smaller amounts of the gas give no visible results, while larger amounts give a deposit more rapidly and in larger quantities. This minimum detectable amount of 1 X 10-7 gram is about the amount which is contained in 0.17 ce. of natural air in which we assume 3 parts of carbon dioxide in 10,000 by volume. The following example will illustrate the cal- culation of the exact amount of the gas a tissue gives off. The original volume of the respiratory chamber is 31.4 cc., to which 6.4 ec. of mereury are introduced, making the remaining volume exactly 25 cc. 10 mgms. of tissue are used and are allowed to respire in the chamber for ten minutes. Then about 10-15 cc. of the gas are withdrawn into the tube B. 0.5 ce. of this gas gave no precipitate during the first ten minutes; 0.5 ee. more of the sample gave no deposit in another interval of ten minutes; 0.5 ce. more, a total of 1.5 ec., was run into the chamber. A marked evi- dence of a precipitate appeared in five minutes. 1.5 cc. of this gas must therefore contain 1 X 10-7 gram of carbon dioxide. The apparatus is then cleaned and dried and a clear drop of barium hydroxide is again introduced upon the top of the tube d; and after again insuring the fact that the air is free from any COs, by waiting, 1 ec. of the sample gas which has been left undisturbed in tube B is introduced into the chamber; no precipitate was formed within ten minutes; 0.25 cc. more of the sample did not produce any precipitate; but when 0.25 cc. more is taken, crystals of barium carbonate now appeared after a few minutes. 1.5 ec. of the gas must contain, there- fore, 1 X 10-7? gram of carbon dioxide. From these duplicates, it becomes certain that 1.5 ec. out of 25 ce. of the chamber now contain 1 X 10-7 gram of carbon dioxide. Therefore the total amount of carbon dioxide produced by 10 mgms. of the tissue during ten minutes will be » 1 X 10-7 gram x? = 16.6 X 10-7 gram of carbon dioxide!’ 5. For a rapid collection of air for a later analysis. With this new apparatus one can also collect the air very rapidly and analyze it at leisure, thus enabling him to collect the air at successive short intervals of respiration by the same tissue or similar tissues at different stages of activity. For this purpose, a very simple special form of the gas pipette was devised. Figure “ This is the case when the analytic chamber has about 15-20 ce. It may take a longer time to produce a precipitate, when the chamber is much larger than this. " For correction for temperature and pressure, see footnote on p, 494. Shiro Tashiro 493 2 will illustrate the exact shape of the tube. Instead of tube B, this tube is connected to the arm of the stopcock 2 at a by means of rubber tubing,"* and the tube is connected to the mercury bur- ette G at b. With this arrangement, we should repeat the experiment exactly the same way as above, except that when we drive the air from the chamber to this tube, we should drive it so far as to push a few ‘cubic centimeters of mercury also from the chamber, so that the mercury will remain in the U-tube U, thus automatically sealing a if Fig. 2. A special gas pipette } actual size. the tube. By clamping the rubber connection at b, this tube is removed, and another pipette is connected, and the experiment is repeated with the same tissue or another tissue as the case may be. By this method, one can collect twenty or thirty samples of the gas a day with a single apparatus. 14 Use of rubber tubing is harmless, provided the mercury burette G is kept above the level of stopcock 2. . 16 Since this pipette has a capacity of 10-20 cc., and only about 10-15 cc. of air are introduced, there will usually be 4-5 cc. of mercury left at the lower end of the tube, thus sealing it automatically at both ends. It is obvious that one should keep the tube vertical, in order to keep it air-tight. 494 Estimation of Minute Quantities of CO, When we are ready to analyze the gas from these tubes, all we have to do is to connect the tube to the usual place (at 2) and raise the mercury burette G in such a way that the mercury in the U-tube is now driven up to fill all the capillary tubes between the chamber and the pipette, thus forcing all the atmospheric air out of the tube, and then stopcock 2 is turned so as to cut all connections. The air in the tube is then examined according to the method described before." , 16 One disadvantage of this new apparatus is that we must take into con- sideration temperature and pressure variation, which was entirely unneces- ary for the previous apparatus. If the respiration and analysis were done aie fferent temperature and pressure, the ratio between the minimum volume which gives the first precipitate and original volume of the chamber will not be rigid. In that case, the minimum volume should be translated to the volume at the temperature and pressure at the time of respiration. Such correction, however, will not be necessary if the analysis is done imme- diately after the respiration, during which the variation in temperature and pressure will not affect the result beyond the experimental error, as is shown in the following calculation: : Let us suppose that 10 mgms. of tissue respires for ten minutes at 18° at 760 mm. of pressure in 25 cc. of the chamber, and suppose 1.5 cc. of the same air at 22°, at 730 mm. of pressure (making a liberal estimate of the change in temperature and pressure) gave the first precipitate; then we will obtain the following results: 5 , a. Without any correction, we get 1 X 10-7 gram x = 16.6 < 10-7 gram. * P 25 b. With the correction, 110-7 gm. X 1 poe (270-+18) x730 ~ 17° X< 10-7 gm. (270+-22) X760 This is a little over 5 per cent error, which will be the maximum, and almost an impossible variation for ordinary weather in the laboratory for a short interval of time. Besides, we are dealing with a very small sample of moist tissue, the weight of which may easily vary within 5 per cent. ON THE RATE OF ABSORPTION OF CHOLESTEROL FROM THE DIGESTIVE TRACT OF RABBITS.! By EDWIN P. LEHMAN.? (From the Laboratory of the Medizinische Poliklinik, Freiburg. in Breisgau.) (Received for publication, November 24, 1913.) Our knowledge of the physiology of cholesterol, particularly of its absorption from food, has only in recent years made notable progress. The fact that it is, in truth, absorbed from the food, ac- cording to the uniform evidence of several investigators, may now be considered established. 4 The following studies were undertaken in Forder to determine, as accurately as possible, the rate of absorption in normal rabbits, as a basis for further investigations of the alterations of this rate in those pathological conditions, notably lipemia, in which the cholesterol metabolism is known to be greatly disturbed. As early as 1867; Tolmatschek* sought to prove the absorption of choles- terol from the food by estimating the intake and output of cholesterol in the breast-fed child. In 1890, Thomas,‘ working on dogs with experimental biliary fistula, found no increase of cholesterol in the bile with a diet rich in cholesterol. Jankau,® two years later, found no increase in the feces of rabbits and dogs after a single feeding of cholesterol; but, six hours after feeding, he found also no increase in cholesterol in blood and bile and liver- substance. With this contradictory evidence he was forced to leave the question still undecided. In 1906, the first work of real significance appeared. Pribram® fed rab- bits through a stomach-tube on three successive days with the palmitic ! These studies were made under the direction of Prof. P. Morawitz, to whom I wish here to express my gratitude, not only for valuable advice and assistance, but also for the opportunity to carry on the investigation. 2 John Harvard Fellow, Harvard Medical School. ’Tolmatschek: Hoppe-Sey’er’s Med.-chem. Untersuchungen, 1867, p. 272; cited from Oppenheimer’s Handbuch der Biochemie, 1908, Vol. IV, No. 1, p. 485. ‘Thomas: Inaug. Diss., Strassburg, 1890. 5 Jankau: Arch. f. exp. Path. u. Pharm., xxix, p. 237, 1892. 6 Pribram: Biochem. Zeitschr., i, p. 413, 1906. 495 496 Rate of Absorption of Cholesterol and oleic acid esters of cholesterol and with cholesterol itself, in amounts of 1 gram on each day. One or two days later, the animal was killed and the blood and organs were analyzed for cholesterol by saponification, ether extraction and weighing. He found with feeding of esters and pure sub- stance alike, an increase in the blood and no definite reliable results from the tissue estimations. He also showed that the serum of an animal fed with cholesterol would prevent or delay in small doses the haemolytic action of saponin on normal red blood corpuscles: Morgenroth and Reicher’ reported in the next year, the fotlowiele experi- ments with a series of rabbits. Rabbit A was fed 4 grams of cholesterol in 15 ce. of olive oil daily, rabbit B, the same amount of pure olive oil, rabbit C, nothing beyond the ordinary diet; all three were then injected alike with lecithide. After five days, the haemoglobin estimations were respectively _ 58, 20, and 30 per cent; and the cholesterol percentage in the blood, respect- ively 0.48, 0.03, and 0.026. The three sera showed corresponding influences on saponin haemolysis in vitro. These results confirm Pribram. Goodman,* in the same year, fed two sets of dogs respectively, white of egg and calves’ brains, and found no increase in the bile with the diet richer in cholesterol. In 1908, Kusumoto® fed dogs with ordinary diet with and without the addition of cholesterol, and estimated the cholesterol of the feces, finding that an average of 30 per cent of the amount ingested failed of excretion through the intestines. In the same year Dorée and Gardner,'® in a convincing series of experi- ments, supported the above cited evidence favoring absorption of choles- terol from the food. In the feces of rabbits fed in the course of several days with 2 grams of cholesterol, preceded and followed by three days of feeding with a cholesterol-free diet, they found that at least 25 per cent, and often more, of the ingested cholesterol failed to appear,in the feces. A rabbit re- ceiving only the cholesterol-free diet excreted no cholesterol. In blood-es- timations they employed the gravimetric method. Rabbits fed twenty days with cholesterol-free diet showed only a trace of the substance in the blood, whereas 0.0415 per cent cholesterol appeared in the blood of rabbits fed for ten days on the same diet with the addition of a total of 2.25 grams of choles- terol. With dogs they found some increase in the blood with foods richin cholesterol. Fraser and Gardner" reported inhibition of saponin haemo- lysis by the serum of rabbits fed (1) with cholesterol, (2) with the cholesterol esters, (3) with mixed diet, as compared in each case with the serum of rabbits fed on cholesterol-free bran. The same authors,! a year later, employing a modification of the new Windaus digitonin method - best scahan catinaae 7 Saiienroth and Reicha Berl. klin. oaliiteche,, XxXviii, p. 1200, 1907. *Goodman: Hofmeister’s Beitrdge, ix, p. 91, 1907; cited from Dorée and Gardner: loc. cit. *Kusumoto: Biochem, Zeitschr., xiv, p. 411, 1908. '° Dorée and Gardner: Proc. of the Royal Society, \xxxi, p. 109, 1908, ' Fraser and Gardner: Proc, of the Royal Society, \xxxi, p. 230, 1909. 18 Thid., \xxxii, p. 559, 1910, Edwin P. Lehman 497 tion, repeated the work of Dorée and Gardner and reported similar results. Studies carried out in 1912 by Ellis and Gardner," in the same laboratory, led them to conclude from evidence of the same nature, that the cholesterol content of the blood is dependent on the “‘sterol-content’’—i. e., phytos- terol and cholesterol—of the diet. An increase in the blood during starva- tion they attributed to the freeing of cholesterol by the destruction of the cholesterol-rich tissues of the animal. Klein, two years earlier, as reported by Magnus-Levy, found increase of absorption with increase of dose of cholesterol, as measured by the output in the feces of dogs. He found no difference i in absorption between pure cholesterol and its esters. Grigaut and L’Huillier," in 1912, studied, with the former’s colorimetric method, the curve of cholesterol content in the blood of dogs fed daily with 1 or 2 grams of cholesterol; and compared this curve with the curve of choles- terol present in the feces during the period of experimentation. His curves show a marked rise with the first feeding of cholesterol and a maintenance of the ‘“‘hypercholestérinémie’’ throughout the feeding period, with a prompt fall coincident with the return to normal diet. sad Anitschkow"' found, after prolonged feeding of cholesterol, pathological changes in various tissues of rabbits, particularly the walls of the aorta, representing an increased body-content of cholesterol. Others report similar observations. Rouzaud and Cabanis,'? using the Grigaut method, found an increase in the blood of only one out of eleven healthy young people, four to five hours after the ingestion of a meal consisting of thin soup, bread, meat, green peas, two eggs and wine. The other ten subjects showed nochange. Their work seems not to have been satisfactorily controlled. For the purpose of this research, the sole conclusion to be drawn from these investigations is that, after the feeding of cholesterol and its esters in relatively large doses, there is an undoubted in- crease in the blood as compared to the average figures with any one method of estimation. They give no hint as to the rapidity of absorption into the blood or disappearance from it. . METHOD. For the present investigation, rabbits were fed alike on the ordinary mixed diet of oats, hay or grass, and bread, and received daily as much as they would eat. They were thus under approx- imately normal conditions of metabolism. 13 Ellis and Gardner: Proc. of the Royal Society, |\xxxv, p. 385, 1912. ™ Klein (Magnus-Levy): Biochem. Zeiischr., xxix, p. 465, 1910. Grigaut and L’Huillier: Compt. rend. soc. biol., lxxiii, p. 304, 1912. 6 Anitschkow: Deutsch. med. Wochenschr., 1913, p. 741. 7 Rouzaud and Cabanis: Compt. rend. soc. biol., Ixxiv, p. 469, 1913. 498 Rate of Absorption of Cholesterol About 6 ce. of blood were withdrawn for estimation and imme- diately thereafter 10 cc. of a 3 per cent solution of cholesterol in olive oil (Merck), representing a dose of 0.3 gram of cholesterol, was introduced into the stomach by a tube. The control animals, with the exception of one (XVID), received the same amount of a solution of pure olive oil. The blood for the further estimations was withdrawn at inter- vals indicated in the tables to follow, in each instance about 6 ce. being taken. The degree of the consequent anemia was observed by haemoglobin determinations after Haldane. All bleeding was done by the Zahn method!’ with a suction-glass from the veins of Pett p., the ear; and the blood was treated with sodium oxalate to prevent coagulation. The Autenrieth-Funk’® colorimetric method of cholesterol esti- mation, with chloroform extraction, was employed, and, owing to the small amount of blood required by this method, it was possible, in most instances, to make two independent extractions and deter- minations with each sample of blood. The results of the two deter- minations, in most cases, agreed closely, as the tables indicate. On account of the small percentage of cholesterol that the bloods yielded, it was found necessary to modify the technique, as the authors suggest in the original description, to the extent of ex- tracting with 55 cc. instead of 100 cc., and, after the ordinary incidental evaporation, making up the test solution to 50 ce. Re- peated further extractions with no addition to the yield of choles- terol showed that, with small percentages at least, the less ex- tended extraction is as effective as a more thorough procedure. RESULTS OF EXPERIMENTS. The figures for the cholestero! content from the blood of twenty norma’ rabbits, as given in the seventeen tables below, together with those from three rabbits not there listed, range from a maximum of 0.1230 to a minimum of 0.0795 per cent with an average of 0.1020 per cent. This average is distinctly higher than that which Abderhalden*® reported in 1898, determined by the gravimetric method from the mixed blood of twelve hea thy rabbits. His figure for the whole blood is 0 0611 per cent. '*Zahn: Minch, med, Wochenschr., 1912, No. 16, p. 861. ' Autenrieth and Funk: Minch. med. Wochenschr., 1918,No. 238, p, 1243. * Abderhalden: Zeitschr. f. physiol. Chem., xxv, p. 65, 1898. er: Edwin P. Lehman 499 The records of the seventeen experiments and controls follow: Experimental series. Fed with cholesterol. Rabbit I. 1550 grams. 10 cc. 3 per cent cholesterol oil by mouth. | CHOLESTEROL IN BLOOD INTERVAL AFTER | Be eh ADMINISTERING Hgb. | Absolute Percentage CHOLESTEROL PARENT | Increase | Decrease a b Average a per cent a No p cent “per cent Normal............. 74 | 0.1200) 0.1195 | 0.1198 | EM ee ep pea 0.1185 | 0.1195 | 0.1190 | 0.67 OIE. ioe sees 60 | 0.1515 | 0.1490} 0.1503 25.46 _— OE eae ae 73 | 0.1090| 0.1050) 0.1070 ' 10. Rabbit IT, 1650 grams. 10 cc. 3 per cent cholesterol oil lh mouth. meee >... iS Rey 60 | 0.1230 0.123 WORDUIA... 6. scun gh 0.1550 | 0.1580 | 0.1565 27.23 | iins.........0h0 0.1515 | 0.1535 | 0.1525 | 23.99 | iPGays..:..... >a 60 | 0.1400 0.1400 | 13.82 | Rabbit III. 1650 grams. 10 cc. 3 per cent cholesterol oil by mouth. Normal... 75 | 0.1010} 0.1020| 0.1015 a 0.1050 | 0.1040} 0.1045 2.95 Va 68 | 0.1074/ 0.1050} 0.1062} 4.63 Des ....... 60 0.0950 | 0.0945 | 0.0948 6.60 Rabbit IV. 1520 grams. 10 ce. 3 per cent cholesterol oil by mouth. SVOPTiGl.... wavici cs. 80 0.0920 | 0.0920 | 0.0920 Se OUTS... ee eiieess 0.1190 | 0.1175 | 0.1183 | 28.58 BOIAVS... 5. <4 deteos 76 0.1050 0.1050 | 14.13 ON ae ee 70 0.1065 | 0.1050) 0.1058 | 15.00 | ' MPVS. . ose ee sak 78 0.1015 | 0.1040 | 0.1028) 11.74 Rabbit V. 2340 grams. 10 cc. 3 per cent cholesterol oil by mouth. Normal. 75 | 0.0900! 0.0860! 0.0880 | oS a Sa 0.0960 0.0965 | 0.0963 9.43 | os ne 0.0965 | 0.0940 | 0.0953 8.29 Rabbit VI, 2570 grams. 10 cc. 3 per cent cholesterol oil by mouth. Nonna k....... 70 | 0.1120 0.1120 | AZUOURGe ss... 0.1190 | 0.1165 | 0.1178 5.18 SORVR Tea... .... 0.1170 | 0.1050) 0.1110 0.89 500 Rate of Absorption of Cholesterol Experimental series. Fed with cholesterol—Continued. Rabbit VII. 1930 grams. 10 ce. 3 per cent cholesterol oil by mouth. — 64 | CHOLESTEROL IN BLOOD Absolute Percentage b 0.1105 0.1220 0.1715 0.1780 0.1700 | per cent 9.74 Rabbit TX. 1750 grams. i ee ee 8.92 3.47 per cent cholesterol oil by mouth. 76 0.1070) 0.1060} 0.1065 / 0.1150 | 0.1170} 0.1160 68 | 0.1055 | 0.1000} 0.1028 10 cc. 3 | 92 | 0.0900 | 0.0910 | 0.0905 0.0795 | 0.0820) 0.0808 | 75 | 0.1130 24.86 10.72 . 10 ce. 3 per cent cholesterol oil by mouth, Normal..........:.. 96 | 0.0810} 0.0820 | 0.0815 Cheers... «64a «3% 0.0730 | 0.0825 | 0.0778 4.54 Ri. mies see 82. 0.0955 | 0.0950 | 0.0953 | 16.93 Rabbit XT, 1610 grams. 10 ee. 3 per cent cholesterol oil by mouth. Normal. issn. ......abeae 0.0910 | 0.0910 | 0.0910 BOUTS a aeds >.> An 0.0790 | 0.0900 | 0.0845 7.14 Rabbit XIT. 1570 grams. 10 ce. 3 per cent cholesterol oil by mouth. Normigisa....<. steer 70 0.0855 | 0.0880 | 0.0868 Shouting... ..:.osane 0.0680 0.0680 20.65 Edwin P. Lehman 501 Control series. Rabbit XIIT. 1450 grams. Control. 10 ce. olive oil by mouth. | CHOLESTEROL IN BLOOD INTERVAL AFTER YO Ie j ADMINISTERING Hgb. | Absolute Percentage ] OLIVE: Om — ——_—— , Increase | Decrease | a b Average a sO cent ee Gab cont | per cent Mertial.::..cysse. 78 | 0.0865} 0.0870 | 0.0868 | WOUNE). Gis oes een 0.1100 | 0.1095 | 0.1098 | 26.49 | ar mee 60 | 0.0972 0.0972 |-11.98 | Rabbit XIV. 1480 grams. Control. 10 ce. olive oil by mouth. — OSS ei 82 0.1100 | 0.1095 | 0.1098 NOLS 2. «coe — 0.1185 | 0.1170 | 0.4178 7.29 PNG, 3. 0h Se 68 .| 0.1110! 0.1140) OFF 2.46 SO. x... vo, caheeeeiae 62 | 0.1150 0.1150 4.73 Rabbit XV. 1700 grams. Control. 10 ce. olive oil by mouth Normal....... «2m 72 0.0770 | 0.0820 | 0.0795 12houmeme.... 0.0760 0.0780 | 0.0770 3.14 GME ays oes vee 0.0815 | 0.0790 | 0.0803 1.00 Rabbit XVI. 1900 grams. Control. 10 ce. olive oil by mouth. 70 | 0.0990 Normnee..:,...... | 0.1020 0.1005 PAT OUTH A os cs... 0.1015 | 0.1045 0.1030 2.48 sdave ok. i:.. | 70 0.1000) 0.0955 | 0.0978 2.68 Rabbit XVIT. 1920 grams. Control. Received nothing. oe MUOTIVNAl. «5 cis sioebieka | 76 0.1130 | 0.1145 | 0.11388 | MTG. os cs ae pone | 0.1070 | 0.1095 | 0.1083 4.83 From the above figures the appended curve was constructed, showing the average blood-content at the intervals indicated in the twelve animals that received the usual dose of cholesterol. It must be noted, that, although the points form a not irregular line, the average figure for any one interval is; in most cases, drawn from widely differing individual figures. Thus at six hours, the average indicated on the curve, —3.24 per cent, is a product of results varying from — 10.72 per cent tu. +2.95 per cent. In reckoning the 502 Rate of Absorption of Cholesterol average figure for three days, it was necessary to omit the record of Rabbit VII, which shows at that interval an increase twice as great as the next greatest in the series and six times as great as the average; this discrepancy is not to be accounted for. The point X shows the deflection of the curve if the cholesterol content of the blood from Rabbit VII is allowed to enter as an element in the average. 24hbrs on 8hre sof Composite curve from seventeen rabbits, showing the average variation in pheneers content of the blood after the feeding of 0.3 gram of choles- terol. i The great differences in the individual reactions both in the experimental animals and in the controls, do not appear to be de- pendent on the body weight. Although Rabbits V and VI, the heaviest in the series, are among those showing the least effect from the procedure, yet Rabbit III, about 1000 grams lighter, shows even less effect, and Rabbit VII, 300 grams heavier than Rabbit III, shows the greatest effect of all. That the repeated . withdrawal of blood is alone sufficient to cause great variations from the normal content is suggested by Mauriac." This is quite ** Mauriac: Compt. rend. soc. biol., xxiii, p. 675, 1912. Edwin P. Lehman 503 possible, but an analysis of the figures shows that the control ani- mals show less average variation from the normal than the ex- perimental animals. There is one notable exception; a single control, Rabbit XIII, fed with pure olive oil, at the end of six hours showed an increase of 26.49 per cent whereas among the ’ animals fed with cholesterol it will be noted that in only one in- stance (IIT) was there an increase so soon after feeding, and in that case an increase of only. 2.95 per cent. The reason for this exceptional occurrence in Rabbit XIII is unexplained. This, however, can hardly vitiate the conclusion that, in general, the curve established from the estimations on the blood of the choles- - terol-fed animals is based on actual absorption of the cholesterol placed in the digestive fract and not on properties of the technique employed. , To turn to the figures from the cholesterol-fed’ animals from which the curve was constructed, we find that several of the ani- mals, as mentioned above (III, v, VI), show less, or little more, increase of cholesterol in the blood after the administration of cholesterol than do the animals that received no cholesterol. The majority however, tend to correspond with the curve—to show an initial fall (due, perhaps, to the recent hemorrhage with dilution of the diminished blood-vo ume from tissue-fluids) lasting for six to eight hours, followed by a rise reaching its maximum at the end of about twenty-four hours, and a more gradual fall through the _ next two or more days. The small number of experiments performed does not permit more extended generalization; one must expect that any individual rabbit may depart widely from the tendency that the curve ex- | Wiresses,” fea CONCLUSIONS. It is possible by giving rabbits small doses of cholesterol by mouth to demonstrate in the majority of instances, an increase of this substance in the blood in the course of a few hours. ® Three experiments, with subcutaneous injection of cholesterol oil in 10 cc. amounts suggest that the absorption from the subcutaneous tissues is much slower. The maximum amount in the blood appeared to be reached between the third and sixth days or even later. These animals were not controlled and the othér conditions of the experiments make the results worth recording only as being suggestive. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 GLYOXALASE. PART IV. . By H. D. DAKIN anp H. W. DUDLEY. (From the Herter Laboratory, New York.) (Received for publication, November 27, 1913.) The object of the following paper is to record some new exper- iments dealing with enzymes of the glyoxalase type and with the inhibitory action of the pancreas upon these enzymes. It may be recalled that we have recorded the presence of glyoxalases in a variety of tissues from various animal species.'’ The tissues examined with positive results included liver, kidney, thyroid, spleen, heart muscle, skeletal musele, tongue, lung, brain, blood cells and gastric mucosa. Negative results were obtained with saliva, urine, bile and blood serum. On the other hand, pancreatic — tissue and juice were found not only to be free from glyoxalase but to contain a thermolabile substance, not improbably an en- zyme, which exerts an intense inhibitory action upon glyoxalase derived from other sources. The inhibitory substance, named for convenience antiglyoxalase, is not identical with trypsin, lipase or diastase. The inhibitory action of the pancreas upon glyoxalase appeared so definite a phenomenon and had so suggestive a relation to the function of the pancreas in carbohydrate and lactic acid me- -tabolism, that we considered it desirable to examine other glands of the body, particularly those which in recent years have been brought into relation with sugar metabolism, in order to learn if antiglyoxalase was peculiar to the pancreas. We have been unable to obtain any evidence of the presence of antiglyoxalase in the thymus, thyroid, suprarenal, pituitary, or salivary glands, or testicle, but on the contrary we have deter- mined the presence of glyoxalasein allof theseorgans. The amount of glyoxalase in the salivary glands is very small. In the case of the abdominal lymph glands of the dog, we have observed a con- 1 This Journal, xv, p. 463, 1913. 506 Glyoxalase stant absence of glyoxalase, but efforts to obtain evidence of the presence of antiglyoxalase have given doubtful or negative re- sults. It is certain, however, that if antiglyoxalase be present in the lymph glands, its amount is utterly insignificant when com- pared with the pancreas. It is perhaps conceivable that the anti- glyoxalase of the pancreas reaches distant parts of the body by way of the lymphatic system, but of this we have no precise evi- dence. Moreover, we are inclined to believe that, notwithstand- ing the fact that antiglyoxalase is present in the external pan- creatic secretion, it is likely that antiglyoxalase is largely contained in an internal secretion. The reasons for this belief are as fol- _ lows: In the first place we are under the impression that the con- centration of antiglyoxalase in pancreatic juice is relatively small compared with that in the tissues. Secondly, antiglyoxalase ap- pears to be non-dialyzable and does not pass through animal mem- branes and hence would probably not undergo ready absorption from the intestine. In the light of our present results, it appears _ that the production of antiglyoxalase is a specific function of the pancreas. | We wish to record at this point that in a private communication from Prof. F. G. Hopkins we learn that in some experiments made several years ago, he found that lactic acid production in muscle was markedly inhibited by the action of pancreas extract. We be published shortly. In a recent paper Neuberg has taken exception to the name glyoxalase. He writes? Das Enzym, das die Umwandlung von Methylglyoxal in Milchsiiure bewirkt, ihnelt nach meinen Ausfihrungen in seinen Eigenschaften und seiner Wirkungsweise der bekannten Aldehydmutase. Durchgreifende Un- terschiede von dieser sind nicht offenbar geworden. Js liegt daher bislang keine Veranlassung vor, das Ferment als Vertreter einer neuen Gruppe zu % understand that these experiments are being amplified and will — betrachten, wie es Dakin tut. Besonders ist aber der von Dakin gewiihlte . . Name “Glyoxalase’’ héchst ungliicklich, da gerade das Glyoxal bisher nicht nachweisbar beeinflusst wird. Zweckmiissiger erscheint daher mein Vorschlag (1. ¢.), das Enzym den Aldehydasen anzureihen und es vorliufig Ketonaldehydmutase zu benennen, da diese Bezeichnung nichts priiju- diziert. * Biochem. Zeitachr., \v, p. 502, 1913, H. D. Dakin and H. W. Dudley 507 We believe that these objections are without weight for the fol- lowing reasons. In the first place, contrary to Neuberg, we find that glyoxal is converted into glycollic acid by enzyme action. On perfusing a dog’s liver with blood to which glyoxal had been added, we recovered almost 2 grams of pure calcium glycollate. Since ordinary commercial glyoxal is a mixture of highly poly- merized substances, it is not surprising that it should be acted on less readily than some of the other glyoxals. Secondly, we have found that the inhibitory action of pancreas extract upon glyoxalase furnishes us with an excellent method for the differentiation of glyoxalase from other enzymes. Parnas’ aldehydemutase is scarcely affected by pancreas extract under con- ditions which completely inhibit glyoxalase, so that we believe that the two enzymes can have nothing in common. | Incidentally it may be noted that the distribution of haar enzymes in the body is quite different and the reactions with which they sever- ally are concerned have only a superficial resemblance. We there- _ fore propose to retain the name “glyoxalase” for the enzymes which we have shown to effect the conversion of various glyoxals into the corresponding hydroxy-acids: R.CO.CHO + H.0 = R.CHOH.COOH fi _ Thus far we have made use of glyoxal itself, methyl glyoxal, _ isobutyl glyoxal, phenyl glyoxal and benzyl glyoxal, and in every case we have obtained the corresponding hydroxy-acid by the action of glyoxalase. Finally we wish to mention that we have _ been able to demonstrate the formation of amino- as well as hy- _ droxy-acids from corresponding glyoxals when perfused through | the liver. We are also making experiments in which the forma- tion of hydroxy-acid is suppressed by addition of pancreas extract to the blood used for perfusion. A description of the synthesis of the hitherto unknown isobutyl and benzyl glyoxals, correspond- ing to leucine and phenylalanine, together with a study of the _ formation of amino- and hydroxy-acids from them, will be pub-— lished shortly. 508 Glyoxalase EXPERIMENTAL. I. Glyoxalase in certain glands. The method employed was: identical with that already de- scribed in a previous paper,® except that in these experiments in- stead of using extracts the tissues themselves after washing free from blood were added to the digestion mixtures. Phenyl glyoxal was used as substrate and in all positive experiments the mandelie acid produced was isolated in crystalline form. From blank experiments in which the enzyme was destroyed by preliminary heating, no mandelie acid could be obtained. The following results are typical. ROTATION | PRESENCE ANIMAL TISSUE perprry | ern ee ACID ALASE ce. deg. Calf:4..........| Thymus (10 gms.) 2.2 | —0.53 + Hosa { Suprarenal (20 gms.) 10.0 —2.56 + re Pituitary ( 3 gms.) 4 BY —0.27 + OMe. . 00. an Pituitary (10 gms.) 5.4 —1.25 Testicle (10 gms.) 5.8 —1.75 + Abdom. lymph ’ glands ( 5 gms.) 0.8 0 a See kc Sane Abdou lymph glands ( 6 gms.) 12 +0.05 - Salivary glands (10 gms.) 2.2 —0.22 We owe the experiments on the suprarenal and pituitary glands of the horse to the kindness of Dr. C. Ten Broeck of Harvard, and we wish to express our thanks also to Dr. F. Fenger of the research laboratories of Armour and Company, for making the experiments with ox pituitary. II, Examination of abdominal lymph glands of the dog for antiglyoxalase. Many experiments were made in which finely chopped lymph ; glands‘ were added to 20 per cent skeletal muscle extract. ti * This Journal, xv, p. 466, 1913. ‘ Glands from near the pancreas were purposely rejected. H. D. Dakin and H. W. Dudley 509 incubating this mixture for several hours in the presence of chalk, phenyl glyoxal was added and incubation continued for about eighteen hours. As control an equal amount of 20 per cent ex- tract was incubated in the presence of chalk, phenyl! glyoxal being added at the same time as to the lymph gland experiment. Only a few typical experiments are here reported. a. 50 ce. 20 per cent skeletal muscle extract were incubated for three hours with 9 gms. minced lymph glands and 5 ce. of a chalk suspension. Then 0.2 gm. phenyl glyoxal was added and incubation continued for twenty hours. In the control experiment 50 cc. of the same muscle extract were ineu- bated in the presence of chalk for three hours, after which 0.2 gm. phenyl glyoxal was added and incubation continued for twenty hours. Lymph gland experiment. Rotation of mandelic acid: —0.82°; Acidity, 3.8 cc. Control. Rotation of mandelic acid: —1.35°; Acidity, 3.8 ce. b. As in (a), 14 gms. lymph glands being used. Lymph gland experiment. Rotation of mandelie acid: —0.37°; Acidity, 3.2 ce. Control. Rotation of mandelic acid: —0.22°; Acidity, 2.0 ce. c. As in (a) 5 gms. lymph glands taken. Lymph gland experiment. Rotation of mandelic acid: 0°; Acidity, 1.8 ce. Control. Rotation of mandelic acid: —0.32°; Acidity, 1.8 ce. d. As in (a) 3.5 gms. lymph glands added. Lymph gland experiment. Rotation of mandelic acid: —0.18°; Acidity, 2.2 ce. Control. Rotation of mandelic acid: —0.17°; Acidity, 2.2 ce. The results of these experiments show that under the observed conditions the presence of antiglyoxalase in lymph glands cannot be definitely asserted. III. Dialysis experiments with antiglyoxalase. Ten grams of fresh dog’s pancreas were ground up with sand mixed with 50 ec. of water and dialyzed for twenty hours in a -condome made from the caecum of the sheep. This was chosen as an appropriate membrane to employ in this experiment. The contents of the condome and the dialysate were then tested for ,antiglyoxalase by adding them each to 50 ce. of 20 per cent dog’s ‘skeletal muscle extract and incubating for three hours before adding phenyl glyoxal, at the same time carrying out a control ean 510 Glyoxalase: ~ experiment with the muscle extract. The following results were obtained: Muscle extract. Rotation of mandelic acid: —0.9°; Acidity, 4.4. ce. Muscle extract + contents of dialyzer. Rotation of mandelic acid: +0.1°; Acidity, 0.5 ec. Muscle extract + dialysate. Rotation of mandelic acid: —0.85°; Acidity, 4.6 ce. It will be noted that there is no inhibition in the third experi- ment, whereas in the second the enzyme has been completely paralyzed showing that no antiglyoxalase has passed through the membrane. ‘on * IV. Formation of glycollic acid from glyoxal in the liver. The technique of perfusion was similar to that already reported ® in connection with other experiments made in this laboratory. The dog (4.8 kgm.) was starved for twenty-four hours previous to the operation. The liver (170 grams) was perfused with a mix- ture containing 170 ec. of the animal’s own blood, 500 cc. of fresh blood from another dog and 200 ce. of saline. During the first ten minutes of the perfusion solutions of about 5 grams glyoxal in 100 ec. water and of 5 grams sodium bicarbonate in 200 ce. water were added in small portions to the perfusion fluid. At the end — of the perfusion, which lasted an hour, the liver was washed out — with 200 ce. of saline. The perfusion fluid, after removal of pro- teins according to Schenck’s method, when tested with p-nitro-_ phenylhydrazine, gave a precipitate of the characteristic dinitro- phenylhydrazone of glyoxal, showing that there was still unchanged glyoxal present. An acetone determination of an aliquot part of the filtrate showed that only 29 mgm. of acetoacetic acid had been formed. The mercury was removed from the clear filtrate of the perfusion fluid by means of hydrogen sulphide and the still acid liquid was then evaporated to dryness in vacuo. The residue was washed with alcohol, the alcohol extract evaporated, taken up in water and after the addition of ammonium sulphate and phosphoric acid, extracted with ether in a continuous extractor. The ether * This Journal, ix, p. 146, 1911, Hin he ae oe H. D. Dakin and H.W. Dudley 511 extract, after treatment with calcium carbonate gave a salt crys- tallizing in the characteristic form of calcium glycollate. Almost 2 grams of crystallized calcium glycollate were obtained. Analysis of air-dried salt: 0.1176 gm. lost 0.0260 gm. H;0 at 140° and gave 0.0269 gm. CaO. Calculated for Found: CaCsHs0s.3H20: ers... . 5 «4 cael eee 22.1 22.1 per cent. eee... 5s 5 16.4 16.4 per cent. The remainder of the calcium salt was decomposed with oxalic acid, the filtrate evaporated and taken up in ether... The residue from the clear ether solution crystallized at once on seeding a crystal of glycollic acid, and the hah so obtained, dried on on porous plate, melted at 76—-78°. ‘ see if V. The differentiation of glyoxalase from aldehydemutase by means of the action of pancreas extract. As is known, an extract of pancreas inhibits the action of glyox- alase. The following experiment was made to determine whether pancreatic extract exerted a similar effect on aldehydemutase, an enzyme described by Battelli and Stern’ and also by Parnas,’ whose _method of investigation was substantially followed. The enzyme extract was made by stirring water with an equal weight of minced ox liver and straining through muslin. 250 ce. of this extract were then placed in each of four flasks. To the contents of flasks (1) and (2) were added 2.1 grams of sodium bicarbonate and 2 cc. of isovaleric aldehyde. The extract in flask _ (3) was digested forty minutes with 1 gram pancreatin before add- ing 2.1 grams sodium bicarbonate and 2 ec. isovaleric aldehyde. Flask (4) was heated on the water bath for fifteen minutes and then similar amounts of sodium bicarbonate and isovaleric alde- hyde were added. All the mixtures were incubated at 37° for five and one-half hours after addition of the aldehyde. The reaction was then checked by adding 10 cc. of phosphoric acid, and the mixtures were immediately distilled in steam, 500 ec. of distillate being collected * Biochem. Zeitschr., xxviii, p. 147, 1910; xxix, p. 130, 1910. 7 Ibid., xxviii, p. 274, 1910. , é 512 Glyoxalase in each case. The distillates were then titrated, the results being as follows: (1) Acidity = 6.7 ce. (2) Acidity = 6.4 cc. (3) Acidity = 5.7 cc. Pancreas added. (4) Acidity = 2.7 cc. Blank experiment. This experiment shows clearly that the action of the enzyme aldehydemutase is not abolished by pancreas extract, as is the case with glyoxalase. A second experiment was made i in order to compare directly PPR agp the aldehydemutase and glyoxalase contained in the same extract both before and after treatment with pancreas extract. To this end an emulsion of dog’s liver in five times its weight of water was taken. In two flasks (1) and (2) were placed 400 cc. of the emul- sion and to flask (2) was added the dog’s pancreas (16 grams) finely minced. In flask (8) 350 ec. of the liver emulsion were heated up on the water bath for fifteen minutes to act.as a control in the aldehydemutase determinations. After flasks (1) and (2) had been incubated at 37° for two hours, 50 ce. from each were measured off into flasks (4) and (5) respec- tively; to these were added 0.2 gram phenyl! glyoxal and 5 ce. of a chalk suspension and a typical glyoxalase determination was car- ried out. To each of flasks (1), (2) and (8) were added 3 grams of sodium bicarbonate and 3 ce. of isovaleric aldehyde. Incubation was continued for fifteen hours, when the reaction was checked by the addition of 15 ee. of phosphoric acid and the mixtures were immediately distilled in steam, 400 cc. being collected in each dis- tillation. The distillates were titrated, giving figures indicative of the activity of aldehydemutase, while the determination of the mandelie acid produced in flasks (4) and (5) is a measure of the givoxalase contents of the same mixtures. Aldehydemutase; Flask 1. Pancreas absent. Acidity = 3.2 ce. Flask 2. Pancreas present. Acidity = 3.0 ce. Flask 3. Blank. Acidity = 1.6 ce. Glyoralase: Flask 4. Pancreas absent. Rotation of mandelic acid, —1.13°; Acidity, 2.0 ec. Flask 5. Pancreas present. Rotation of mandelie acid, ~—0,12°; Acidity, 0.25 ce. H. D. Dakin and H. W. Dudley 513 The glyoxalase determinations (4) and (5) were actually made on samples taken from flasks (1) and (2). It is seen that the activity of aldehydemutase is not appreciably diminished by the action of pancreas extract, while the glyoxalase under the same treatment is practically completely inhibited. The two enzymes are undoubtedly distinct. SUMMARY. The presence of glyoxalase and the absence of antiglyoxalase has been determined in all the glands of the body we have examined, with the exception of the pancreas and abdominal lymphatic glands. aed The lymphatic glands contain no glyoxalase and compared with the pancreas, the inhibitory action of extracts of lymph glands upon glyoxalase is trifling or non-existent. The formation of antiglyoxalase, so far as can be at present de- termined, appears to be a specific function of the pancreas, and some reasons are adduced for suspecting that it acts mainly by way of an internal secretion. Contrary.to Neuberg’s statement, we find that glyoxal may be converted into glycollic acid by enzyme action. Furthermore we find that Neuberg’s suggested relation between glyoxalase and -aldehydemutase is incorrect. We have shown that the enzymes are entirely distinct since, unlike glyoxalase, aldehydemutase is substantially unaffected by pancreas extract. SOME NEGATIVE EXPERIMENTS ON THE INFLUENCE OF THE PANCREAS UPON ACETOACETIC ACID FORMATION IN THE LIVER. By H. D. DAKIN anp H. W. DUDLEY. ° (From the Herter Laboratory, New York.) (Received for publication, November 27, 1913.) The incentive to make the following experiments to determine a possible influence of the pancreas upon acetoacetic acid forma- tion in the liver arose from the following facts. First, the known acidosis with excretion of acetoacetic acid observed to follow extirpation of the pancreas. Second, the fact that we have found in the pancreas a mechanism for the regulation of the action of the enzyme glyoxalase, whose function, at least in part, is con- cerned with the formation of another acid produced in interme- diary metabolism, namely, lactic acid. Third, the fact that it is much more difficult to evoke a considerable excretion of aceto- acetic acid in the non-diabetic intact animal under normal condi- tions than it is to demonstrate its production in the excised liver - on perfusion. It seemed possible that the pancreas might furnish some en- _ zyme or hormone the absence of which leads to acidosis in the _ diabetic animal. So far as we are aware, no experiments have hitherto been made to determine this point, although a number of workers have investigated the influence of the pancreas upon the capacity of various tissues to effect the oxidation of glucose. More recently Paderi! has found that the addition of pancreas extract to the fluid used for perfusing a glycogen-containing liver, was not followed by a diminished glucose production. In our attempt to detect any influence that the pancreas may have on acetoacetic acid production we have perfused dogs’ livers with blood containing added substances, known from Embden’s 1 Arch. d. farmacol. sperim., xvi, p. 54, 1913. 515 516 The Pancreas and Acetoacetic Acid Formation experiments to yield acetoacetic acid freely. Tyrosine and the sodium salts of butyric and homogentisic acids were the substances chosen. In some of the experiments we added fresh extract of pancreas to the blood used for perfusion, while in others we added skeletal muscle extract or heated pancreas extract to serve as a control. The tissue extracts were prepared by grinding up per- fectly fresh tissue with sand and ten parts of distilled water. The suspension was then strained, a suitable quantity of salt added, then whipped with clotting blood and again strained before add- ing it to the blood used for perfusion. The methods of perfusion and analysis were those previously used by us. _ Although the results of our experiments, taken literally, might be considered to show a slightly lessened acetoacetic acid produc- tion in those experiments in which pancreas extract was added to the blood, we believe that the results do not warrant any such interpretation, owing to wide individual variations in similar experiments. We conclude that under the conditions of our experiments ad- dition of pancreas extract to the blood has no marked effect upon acetoacetic acid formation in the liver from butyric acid, homo- gentisic acid or tyrosine. MILLIGRAMS svseramon apna | MMUMLABRRRACT apou> vo. | yassoe || Ar FORMED | | | grams 1. Butyric acid 2 gms.......| 100 ce. pancreas....... | 267 7 2. Butyric acid 2 gms.......| 250 ec. panereas........ | 318 lll 3. Butyrie acid 2 gms...... | 100 ce. heated pancreas; 339 106 4. Butyric acid 2 gms....... | 100 ec. muscle ......... | 261, |" 352 5. Tyrosine 1 gm............| 100 ec. pancreas ....... | 293 | 89 6. Tyrosine 2 gms...........| 200 cc. pancreas........| 267 ed 7. Tyrosine 2 gms.. ...++} 200 cc. pancreas........) 169 114 8. Tyrosine 2 gms.. .....| 200 ce. heated pancreas) 281 126 9. Homogentisic ac sid 2 2 gms.| 100 cc. pancreas,.......| 289 | 178 10. Homogentisic acid 2 gms. 200 ce. pancreas........, 151 | 174 Il. Homogentisic acid 2 gms. | 100 ec, muscle ......... | 159 | 325 —_ — ———e ——— ee oe ee re ee ON FAT ABSORPTION. III. CHANGES IN FAT DURING ABSORPTION. By W. R. BLOOR. (From the Laboratories of Biological Chemistry of Washington University, St. Louis, Mo.) (Received for publication, November 28, 1913.) The belief regarding fat absorption which is now almost uni- versally accepted is that the fats are saponified in the intestine, absorbed in water-soluble form as soaps, and resynthesized into neutral fats during the passage through the absorbing cells. The more important evidence! in support of this belief is as follows: 1. Fatty substances such as glycerides, other readily saponifiable esters, fatty acids and soaps, lecithin, etc., which are soluble in water or can be changed by digestion into compounds soluble in water (or bile) at body temperature are readily absorbed. They appear in the chyle as triglycerides. 2. Fatty substances which cannot be changed into water-soluble form in the intestine under these conditions are not absorbed, no matter in what form they are presented. In this class fall difficultly ; -saponifiable esters such as those of wool fat, also the petroleum hydrocarbons, etc.,—substances which are soluble in the ordinary E fats and fat solvents, and in most cases form very good emulsions - with water. 3. The abundant provision made for saponification and for the. absorption of soaps in the intestine by the supply of large amounts of lipase and bile—whose chief function is now believed to be to aid in fat absorption. 4. The presence of soaps in the intestine in relatively large _ proportion, while only small amounts are present in the chyle. 5. While emulsified neutral fats are found on both sides of the absorbing cells, the particles of emulsified fat in the lacteals are ‘The evidence is discussed in greater detail in the preceding paper of this series: this Journal, xv, p. 105, 1913. 517 518 Changes in Fat during Absorption very much finer than those in the intestine. They are of “dust- like fineness,’ apparently of the same order of magnitude as the “‘haemakonien” of Neisser,’ and are probably formed by the par- tial flocking of the newly resynthesized fat molecules under the influence of the electrolytes of the lymph stream. The purpose of the complete breaking down of the fats in the intestine and their immediate resynthesis in the passage from the intestine is not clear in the light of our present knowledge. One reason which has been suggested‘ is that the process is a protec- tive one for the purpose of excluding undesirable fatty substances, such as wool-fat and the petroleum hydrocarbons, which differ Ps: gertrom ordinary food fats mainly in that they cannot be changed into water-soluble substances in the intestine, and which but for -this mechanism would be absorbed with the fats. But this reason is obviously not sufficient to explain the changes, since these sub- stances rarely occur in the food. - A comparison of the absorption of fats with that of proteins and carbohydrates suggests another reason. It is now believed that during digestion all (organic) food- stuffs alike are broken down into their component “Bausteine”’ in the intestine, the object being to provide material in a sufficiently elementary form for use in building up the characteristic body tissue. Proteins are broken down to amino-acids, carbohydrates to monosaccharides and fats to fatty acids and glycerins. Pro-_ tein and carbohydrate “Bausteine” pass directly into the blood — stream and are not rebuilt into body protein and carbohydrate complexes (accepting the findings of Folin and Denis) until they reach the tissues and organs. In their passage from the intestine to the system they pass through theliver. Fats are unique in that they are rebuilt before they leave the intestinal wall, and enter- ing the blood stream by way of the thoracic duct, avoid the liver, which appears to take part in fat metabolism only after the fats have passed to the tissues. The protein complexes rebuilt by the tissues from the protein building stones are different from the proteins ingested and are characteristic of the species, and of the tissue. Since the fats are rebuilt during their passage through the intestinal wall, it is logical to expect a change in their chemi- * Munk: Virchow’s Archiv, exxiii, p..239, 1891. * Neisser and Bratining: Zeitschr. f. exp. Path. u. Ther., iv, p. 747, 1907. * Bloor: loc. cit. W. R. Bloor 519 cal structure during resynthesis unalogous to that undergone by the proteins in the tissues. Another, and perhaps the main reason, then, for the phenomena of fat absorption may be looked for in changes which the fats undergo during absorption. What changes may be expected in the fats during absorption? The chemical structure of the fats is not definitely known, but it is believed that many fats contain in addition to simple triglycer- ides a considerable proportion of mixed triglycerides. The pres- ence of mixed triglycerides is significant because many lecithins and similar substances contain a mixed glyceride residue; and since these lipoids are more closely identified with tissue structure, than the fats, it may be assumed that the mixed glycerides are — - of special importance in those phases of fat metabolism which have to do with tissue repair. That mixed glycerides are also potentially optically active may also be considered significant in view of the possible importance of molecular structure in the utili- zation of other foodstuffs.® We may look for two kinds of change in the fats during their passage through the intestinal wall: (a) a physical change con- sisting in a rearrangement of the quantities of the different gly- cerides—more or less of the liquid fats or of the solid fats—result- ing in a mixture of different physical properties which may be more suitable for transport or storage than the fat fed, or (b) a chemical change consisting in a rearrangement of the fatty acids _ in the molecules of some or all of the glycerides, which in addition _to changing their physical properties, would make them presum- ably more adaptable for use as tissue fats—for “endogenous” ~metabolism. Observed changes in fats during absorption. That fats do not pass into the chyle in exactly the form in which they occur in the food has been noted by several observers. _Arnschink® on feeding mutton fat to dogs found that the feces fat had a higher melting point than the food fat and drew the 5 An excellent example of the influence of molecular arrangement on pro- tein utilization is reported in the recent work of Dakin and Dudley (this Journal, xv, p. 271, 1913) on racemized casein. ® Arnschink: Zeitschr. f. Biol., xxvi, p. 434, 1890. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 520 Changes in Fat during Absorption conclusion that there was a discrimination in favor of the softer fats during absorption. Munk’ experimenting with a case of human chyle fistula after feeding mutton tallow found that the chyle fat had a lower melting point than the food fat, thus sup- plementing the findings of Arnschink. He explained the change also as being due to a discrimination in favor of the softer fats. Some results which he obtained later with the same patient indi- cate, however, that there is another factor. After excluding other fat from the diet, and feeding his patient cetyl palmitate, m.p. 55°, he found that the chyle contained in addition to palmitic acid (as triglyceride) about 14 per cent of combined oleic acid, also that the chyle fat had a melting point of 36°C. while tripal- _mitin Its at 65°C. He proved that the oleic acid could not have come from the food and must therefore have been supplied by the intestine. — Frank® in feeding experiments with dogs using fats of different melting points found that the melting point of the chyle fat was generally closer to that of the body temperature of the animal than that of the fat fed. In the same paper (Experiment 11) after feeding ethyl palmitate, he found that the melting point of the chyle fatty acids was 50.5°C. (palmitic acid 63°), while their iodine number was 32.6—corresponding to an oleic acid content of 36 per cent. In Experiment 12, after feeding ethyl] palmitate (other fats were excluded from the diet in both experiments) and fractionating the chyle fat, one fraction was obtained which melted at 39° and had an oleic acid content of 25 per cent’ (calculated from the iodine number). As in Munk’s experiment the oleic acid was demonstrated not to have come from the food and was believed to have been supplied by the intestine or liver (via the bile). From calculations based on figures for the chyle fat (per hour) of dogs given by v. Walther'® and by Munk," Frank concluded 7 Munk: loc. cit. * Frank: Zeitschr. f. Biol., xxxvi, p. 568, 1898. *It is worthy of note that an oleo dipalmatin (oleic acid content = 33.8 per cent) prepared by Kreis and Hafner (Zeitschr. f. Untersuch. d. Nahr. u. Genussm., 1904, p. 665) had a melting point of 37-39°. Its caleu- lated iodine number would be 30.4. vy, Walther: DuBois Raymond's Archiv, 1890, p. 329. Munk: loc, cit. W. R. Bloor 521 that the extra fat found mixed with the food fat in the chyle, was only that normally present in the chyle and was not a pur- posive addition. A fact which he overlooked, however, was that the dogs used by Munk and v. Walther were as a rule very much larger than his dogs and would therefore yield more chyle per hour. Recalculating his results on the more logical percentage basis, it will be seen that even if all the fat of the fasting chyle were con- sidered as oleic acid, the fat of fasting dog chyle does not account for the amounts of oleic acid found in his experiments. _ For ex- ample in Experiment 11, p. 576, 27 ec. of chyle yielded 0.372 gram of fat containing 40 per cent oleic acid = 0.14 gram of oleic ac ae 7 27 ce. of fasting dog chyle would have yielded 0.07 gram of (0.25 per cent of the chyle—average of Munk’s and V. Walther’s figures). In Experiment 12, p. 576, 40 ce. of chyle cotullined 1.43 grams fat with 13 per cent oleic acid = 0.186 gram; 40 ce. of fasting dog _ chyle contains 0.1 gram of fat. egy OEE ae Since, as will be shown later, the fat of fasting dog chyle does not consist entirely of oleic acid, the fact of these additions to the chyle fat becomes the more remarkable. Bloor,” after feeding pure isomannid dilaurate, found that the purified chyle fat had a melting point of 32° (pure trilaurin melts at 45-46°). The fatty acids prepared from the chyle fat had a melting point of 30°, a mean molecular weight of 211, and an iodine number of 16.5. The corresponding figures for lauric acid are: m.p., 43.6°; mean molecular weight, 200; iodine number, 0. Calculating the unsaturated fatty acids as oleic acid from the iodine number, 18 per cent of the chyle fatty acids was oleic acid. The mean molecular weight of the chyle fatty acids, as- suming them to consist only of lauric and oleic acids would then have been 215.3. Since the actual mean molecular weight was 211, oleic acid was probably not the only acid added, although it was the main one. Taking into account in this experiment the jatar amount of chyle collected and accepting the figures of Munk and vy. Walther for the fat content of fasting dog chyle it cannot be denied that the added oleic acid may have had its origin in this experiment in fat normally present in the fasting chyle. 2 This Journal, xi, p. 429, 1912. 522 Changes in Fat during Absorption Raper," after feeding cocoanut oil, which contains a consider- able amount of lower fatty acids, observed the following differ- ences between the food fatty acids and those of the chyle fat. Cocoanut oil fatty acids: mean molecular weight, 212; iodine num- ber, 7.7. Chyle fatty acids:.mean molecular weight, 236; iodine number, 19.1. The change in molecular weight was consider- ably greater than could be accounted for by the oleic acid added, from which (and other evidence) Raper believes that the lower fatty acids were absorbed through some other channel than the chyle. Bloor,“ in feeding experiments with cocoanut oil, found some- a a: similar differences between the cocoanut oil fed and the fat of the chyle. The cocoanut oil fed had an iodine number of 7.3 and a melting point of 26°. The chyle fat had the same melt- ing point, but the iodine number had increased to 24°. The few results available on this point then indicate (1) that changes in the fats may be produced during absorption; (2) that changes are probably greater than could be produced by the fat present in the normal fasting chyle. — In the work recorded below evidence is submitted of further various changes in fats during absorption which make it seem probable that the intestine is able to radically modify the com- position of fats during absorption. The tendency of the changes observed is toward the production of a fat more nearly like the body fat of the animal than the fat fed. The changes appear to be purposive, since not only are they as a rule much greater than 1e fats present, in fasting chyle, but they. vary ‘in ‘Kind and degree with the nature of the fat. fed. EXPERIMENTAL, The changes above recorded consist essentially in lowering of the melting point and raising of the iodine number by the addi- tion of oleic acid. With the possible exception of the earlier work of Frank, no results are available which indicate a change in the chyle fat in the reverse direction, ¢.e., an elevation of the melting point and lowering of the iodine number—the production of a chyle fat less liquid than the food fat. Raper: this Journal, xiv, p. 117, 19138. “ Bloor: loc. cit, Pe a Lael W. R. Bloor 523 In the course of the study of another phase of fat absorption” it was observed that when olive oil (mixed with various hydro- carbons) was fed to dogs, the melting point of the chyle fatty acids was generally considerably higher than that of the olive oil fed, while the iodine number was lower. A further examination of the fat of these samples of chyle was made and the data which pertain to the present discussion are given below (Table I). TABLE I. FATTY ACIDS FROM CHYLE FAT VOLUME | WEIGHT OF EXP, NO. MATERIAL FED OF CHYLE | CHYLE wie sh a Three ec | “joe? | Nimbegill cc. | gma | re | Olive oil and hydro- | | carbon oil........ 16545 1.2 | 30-32 + 868 II | Olive oil and hydro- / | Zi | carbon oil........ 65 | 0.8 30 79.4 Ill | Olive oil and hydro- ) carbon jelly... 80 146 « 2 Shoe IV | Olive oil and — ) | carbon jelly...... P75 eS below 20°C. | 80.9 V_ | Olive oil and hydro- carbon oil emulsi- | | Ds a h179) ae | 29.5 i= 80.3 VI | Olive oil and hydro- | carbon oil emulsi- Hey Rs. oe a eee 100 |: "eee 30 | 4 0 The fatty acids of the olive oil fed had a m.p. of about 16°, and an iodine number of 86.1°. The chyle fatty acids obtained above had with one exception in each case a much higher melting point and a lower iodine num- ber than the fatty acids of the olive oil fed. The unused fatty acids from these experiments were united, dissolved in ether, treated with bone black until the solution was nearly colorless, then the ether evaporated and the residue dried. The melting point of the combined fatty acids was 29.5°. The acids, on frac- -tionation from chilled petroleum ether, yielded two fractions, the first of which was crystalline and amounted to 22.5 per cent of the whole. After repeated recrystallization it gave a melting point of about 56° (a fraction of constant m.p. could not be ob- 18 Bloor: loc. cit. 524 Changes in Fat during Absorption tained), and an iodine number of 4.1. This fraction was probably a mixture of the higher saturated fatty acids. The remaining portion, soluble in cold petroleum ether, athe repeated chilling in small amounts of petroleum ether, until no more would separate, had an iodine value of 85.9 and a melting point of about 14°C. It was therefore nearly pure oleic acid. There was obtained then after feeding olive oil whose fatty acids consisted of 96 per cent oleic acid, a chyle fat containing approximately 22.5 per cent of solid fatty acids and 77.5 per cent oleic acid and with an average melting point of 29.5°C. ,. me, As already noted, these results were obtained after feeding the olive oil in admixture with hydrocarbons and although none of the hy: bons were absorbed, they may have had an influence on the composition of the chyle fat. Further experiments on this point and to determine the nature of the glycerides obtained, are in progress. In order to obtain more data regarding the changes in the fats during absorption and to determine whether there was any rela- tionship between the amount of change and the nature (especially m.p.) of the fat fed, as well as to settle definitely whether the for- eign fat found in the chyle was normally present in fasting chyle or was purposely added, experiments were conducted in which esters of pure fatty acids of various melting points were fed and the chyle fat examined. Feeding of pure fatty acid esters and collection of the chyle. Ethyl esters of stearic, palmitic and laurie acids were prepared by the action of their chlorides upon ethyl alcohol. The chlo- rides were prepared from pure fatty acids (mainly Kahlbaum’s K preparations) by the method of Krafft and Biirger," and in the preparation of the esters were added slowly with stirring to ex- cess of absolute aleohol. The solution was allowed to stand over night, then poured into excess of water. After thorough washing the material was ready for use. The animals (dogs) were starved for forty-eight hours before the feeding. The esters, which were all liquid at body temperature, were given by stomach tube and the feeding was followed by about 59 grams of bread which had been rendered fat-free by boiling with alcohol. Krafft and Birger: Ber. d. deutsch. chem. Gesellsch., xvii, p. 13878, 1884. W. R. Bloor 525 The operations for the insertion of the cannula into the thoracic duct (for the earlier of which I am indebted to Doctor W. M. Marriott of this laboratory), were performed under ether anaes- thesia. Special efforts were made to make the shock of operation as light as possible and to bring the animal after the operation into a condition as nearly normal as possible for the experiment. — The operations were done with aseptic precautions making a small wound; and by the use of a paraffined cannula of narrow lumen (2 mm.), clotting was prevented without the use of special de- vices. As soon as the cannula was safely in the duct and the wound closed, the animal was removed to a padded table, coveredy 4 warmly and allowed to recover from the anaesthetic. The re- turn to consciousness was always followed by an improvement in the fat content of the chyle. Collection of chyle was continued as long as convenient. The animals rested quietly most of the time. If they became restless they were removed from the table and allowed to walk around for a while, after which they were returned to the table and generally went to sleep. Water was given as often as desired. As previously mentioned, the chyle was occasionally found to be flowing on the second day and another experiment could be made with the same animal. The cannula generally dropped out on the third day and the animal itt almost all cases made a good recovery. The chyle was collected in a vessel containing a little dry mag- nesium sulphate to prevent clotting and when collection was com- plete it was transferred to a separatory funnel, shaken well with ether and the mixture allowed to stand over night. The extracted chyle, now clear, was run off, evaporated to dryness on a water bath, powdered and again extracted by boiling out two or three times with ether. These ether extracts were added to the first and the whole evaporated to dryness. The essential points of the experiments are as follows: piv ry as a= 5 Eruyt Stearate. Dog, weight 10 kgm., a fat female, had been used in a _ similar experiment the day before. The chyle was still flowing and was clear. At 10 a.m. she was fed 7.3 grams of the pure ester of which a con- siderable portion was vomited shortly afterwards. The volume of chyle was not noted. Total chyle fat collected, 0.3 gram. The fat was saponi- fied and the fatty acids separated from unsaponifiable matter and purified with bone black. Melting point of the fatty acids, 45°C. 526 Changes in Fat during Absorption Iodine number, 56.21—containing therefore 62.5 per cent oleic acid. Mean molecular weight, 285. Ernyt Patmitate. ExperimentI. The dog, a female, weight 25 pounds, thin and active, was fed 20 grams of ethyl palmitate at 8.30 a.m. The operation was complete and collection was begun at 11.45. Collection was continued for ten hours with a total yield of chyle of 147 cc., one 1.3 grams of fat—0.9 per cent. Melting point of the fat, 55°. Iodine number, 66.9. Experiment Il. Next morning the chyle was still flowing and clear, so another feeding of 20-25 grams of ester was given together with 50 grams of coagulated egg white, at 10.30 a.m. By 12 0’clock the chyle had begun to a wammepres: milky. Collection was then begun and continued until 7.45 p.m. Total chyle collected, 79 cc., containing 1.6 grams of fat—2 per cent. Melting point of the fat, 57°C. Iodine number, 52.3. The chyle fat from the two palmitate experiments was united. Average melting point, 56°. Average iodine number, 59.6. It was fractionated from ether in the cold, yielding two main fractions of which the first and largest—1 gram—after several recrystallizations, yielded well formed crystals with a melting point of 61°C. and an iodine number of 7.6. It was probably nearly pure tripalmitin. The second fraction, from which no other substance could be separated was liquid at room temperature. Its iodine number was 62.5, corresponding to an oleic acid content of 69.4 per cent. (This value is suggestively close to that of a dioleopalmitin oleic acid, 65.7 per cent.) Erayt Lauvrate: A female dog weighing 20 pounds, in fair condition, was fed 18 grams of ethyl laurate, together with 50 grams of fat-free bread and 10 grams of glycerin, at 8.30 a.m. Collection of chyle was begun at 1.15. From 1.15-3.15, 58 ec. containing but little fat were collected and treated separately (see below). At 3.15 the chyle was becoming richer in fat and continued of good color until 9.45 p.m., when collection was stopped. This portion (II), total 106 cc., was extracted separately (see below). Next morning the chyle was still flowing and of a good white color. Collection was begun again at 8.30. At about 11.00 a.m. it began to lose its white color and 9 grams more of ethyl laurate were fed through a stomach tube. Collection was continued until 9.45 p.m. The chyle from this period, total 131 ce., was treated separately (portion III). The extractions of the sep- arate portions were made as in the other experiments. Portion I. 58 ce.; total fat, 0.21 gram, 0.36 per cent; m.p. 22-24°C. Portion LI. 106 ce.; total fat, 1.82 grams, 1.71 per cent; m.p. 22-24°C, Portion III. 181 ec.; total fat, 1.15 grams, 0.88 per cent; m.p. 30°C, lodine number, 56.35. Portion IV. The chyle from the above three portions was evaporated to dryness, powdered and extracted with hot ether. Weight of extract 0.57 gram. Total laurate chyle collected, ce., containing 3.75 grams of fat—1.27 | W.R. Bloor 527 per cent. Iodine number of the whole chyle fat, 44. Attempts were made to fractionate this fat but without success. The essential facts of the ester experiments are collected in table II. The results of Munk and Frank already cited (pp. 520- 521) are added for comparison. From the results given it may be seen: 1. That the amount of ‘‘oleic acid’’ in the chyle fat is generally much greater than could be accounted for by the fat of fasting chyle, accepting the average value 0.25 per cent already quoted (p. 521) and supposing it to be entirely oleic acid. 2. That there is a parallelism between the melting point of they _ fatty acids fed and the amount of “oleic acid” added—the higher the m.p. of the fatty acid, the more oleic acid. 3. That the fat of the fasting chyle is not always entirely, or even mainly, oleic acid, as may be seen from the laurate experi- ments. Since the chyle in portion I has the lowest fat content it should contain most of the fasting chyle fat, and if this fat were oleic acid, the iodine number should be highest of the three por- tions, while it is actually lowest. 4. That there is a marked similarity in the results of Frank’s experiments with ethyl palmitate and those reported in this paper. The chyle fat of both yielded two main fractions, one of which was undoubtedly tripalmitin and the other with an iodine number and melting point close to those of mixed glycerides of palmitin and olein. It seemed of interest to know also whether any change would be produced in a fat with an already high iodine number and low melting point. For this purpose cod liver oil, iodine number 148, was fed to a dog, the chyle collected and the chyle fat examined as in the earlier experiments. Total chyle, 76 ce. ' Fat of chyle, 1.05 gram—1.4 per cent. Iodine number, 118. The iodine number was reduced from 148 to 118 during the absorption. Here for completeness may be mentioned again the results of experiments with cocoanut oil which have already been men- tioned above (p. 522), as examples of a change of a somewhat different nature. Changes in Fat during Absorption 528. uso sty wiouy poyeIMoqe | ee aililiion oa wenveansie ir pocinemaiis doa: Pwe s/o ey,» ‘urgraypediay Ayqeqoig © *(‘9]89) FTE “OU oUIpOT ‘gg-1e “dul O18 UTVIUITedIp-o9]O UF IOJ SON[eA | *(‘o]89) Z'6¢ ‘OU OUTpoT ‘uTy1UI[edo9]0-1p A] qISsog ‘e-eg “deur ‘urqrunyped11y Ajquqoig ‘pajou you ofkyo Jo oumnjo, | SyuVNaU Ss oo = ARS cocoon _ CAR) AT pur TJ suonovsy "RAT Sa (es skies Mea ecard 8 ] wornsvsy i UOTPBUOTIIEIT UG 61 0 Ps ee 1 Cp ee ZI dxq tl Qwest dey eens It “dxq ) (juvsq) oyeqrayed [Aq 96 {46€ | OOZt |" (qunyY) eyezrarped [Aya er reesiert!) tet fs eee: III wors0g 0 lor r-2e [ee fer eee I] wors0g NS a ak eae he I wort0g | aqemmel Agi SEO |Oe MOO PS eam I] worRsey o2 | 1 | |. patteeeereccene [ wonoeay UOTIVUOTPIBIT UO ¢ ze 1g 9° I 62 ee Il “dx¥ 6 99 ce e I L¥I Se ee ee ee I “dxq 1° oveqruped [qa OUP log | | |ees-teceee-orees ayexeays [Aun "oD “bap *supsb “20 ioq re: ae ays ae ouiper ‘dW 3PM Ray _— — g@waT0A Gia “IVIESIVER iva @IAHO ‘TL WTAaVL W. KK. Boer: 529 EXPERIMENT I (Raper): The cocoanut oil fed had an iodine number of 7.7, and a mean molecular weight of 212. The chyle fat collected had an iodine value of 19.1, and a mean molecular weight of 236. Experiment II (Bloor): Cocoanut oil (fed together with hydrocarbon oil). Iodine number, 7.3; m.p., 26°; chyle fat iodine number, 24; m.p., ee The changes in cocoanut oil consist in the addition of “oleic acid” without a change in the melting point. SUMMARY AND CONCLUSIONS. Evidence is presented of changes in fat during absorption as follows: 1. A lowering of the melting point of high melting point fats _ by the addition of an unsaturated fatty acid, probably oleic acid. _ The addition is proportional to the melting pointiof the fatty acid fed. 2. An elevation of the melting point and latting of iodine ~ number of a low melting point fat (olive oil) by the addition of saturated fatty acids. 3. Addition of “oleic acid” together with a change in the mean molecular weight of the fatty acids, without change of melting point in a fat which consists mainly of glycerides of saturated fatty acids (cocoanut oil). 4, Lowering of the iodine number of a fat (cod liver oil) which i contains a large percentage of glycerides of highly unsaturated fatty acids. The intestine appears to have the power to modify radically the composition of the fats during absorption. The changes are - apparently purposive in that they vary in kind and degree with the nature of the fat fed and also show in general a tendency toward the production of a uniform chyle fat, presumably the charac- teristic body fat of the animal. In the preceding paper of this series one reason was suggested _ for the peculiar mechanism of fat absorption—that it serves to _ exclude undesirable fat-like substances such as the petroleum hydrocarbons, etc. The observations presented above, suggest a second—that the mechanism serves to permit adaptive changes in the fats during absorption. Work is being continued along similar lines. ea 4 iA | Ps _ essarily the minimum amount present. For arginine the results are practically the same, 7.4 to 7.8 per cent of the total nitrogen, THE HEXONE BASES OF CASEIN. By DONALD D. VAN SLYKE. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, November 28, 1913.) In our preliminary description of the method for analysis of pro= — ; teins by determination of the chemical groups characteristic of the _ different amino-acids, we published an analysis of casein.! The results agreed quite well with those previously obtained by other 4 authors with the Kossel method for determining the bases of pro- teins. Consequently, although our method was improved before its final publication, we did not repeat the casein analysis. The if discrepancy, noted in the preceding article, between the free amino nitrogen of casein, and the lysine content previously determined, rendered a repetition of the nitrogen distribution in this protein _ desirable. We have, therefore, determined the bases by the method of Kossel and Patton, as modified by Osborne, Leavenworth, and _ Brautlecht,? and have also redetermined the bases and nitrogen dis- Hi _ tribution by our previously published method of group analysis. The most significant differences between our present results and previous ones occur in the lysine. The percentages of the casein nitrogen previously found in the lysine were 6.66 to 7.24% by Kos- _ sel’s method and 7.86 by our own. By exercising particular care in the Kossel method we have now obtained 9.36 per cent of the casein nitrogen in the form of lysine weighed as the analytically pure picrate. Our group determination method gave 10.3 per cent, and we believe that this figure is even more nearly correct, as the amount of lysine picrate which one can crystallize represents nec- 1 Ber. d. deutsch. chem. Gesellsch., xliii, p. 3179, 1910. * Amer. Journ. of Physiol., xxiii, p. 183, 1908. 5 Ibid. 531 532 Hexone Bases of Casein as those previously obtained by both methods. The histidine results are a little higher than previously, but not to a marked extent. The source of error in our own former results for lysine lay in the cystine determination. The lysine is estimated from the total amino nitrogen of the bases precipitated by phosphotungstic acid, after the cystine nitrogen has been subtracted. The cystine was estimated from the amount of organic sulphur precipitated with the bases. The original form of the method, however, made the cystine figures liable to error from the fact that sulphates could be dissolved from a glass flask used in one stage of the opera- tion. Although this source was recognized‘ and a correction, de- aay ee controls, attempted for it, the results for cystine were ertheless much too high, those for lysine being conse- quently low. In the form to which the method was modified be- fore being published in detail in this Journal,® the above source of error in the cystine and lysine determinations was eliminated. In the determination by the pierate method, as usually per- formed, it appears that the most probable source of loss lies in the decomposition of lysine phosphotungstate with barium hy- drate. In this operation one insoluble precipitate (lysine phos- photungstate) is transformed into another (barium phosphotung- state), a process the completeness of which is necessarily difficult. to judge. Moreover, the bulky barium phosphotungstate has — marked adsorptive properties, so that even skill and experience — might not insure against loss from this source. In working out the details of the group determination method, we noticed that several percent of the total nitrogen of the protein could be lost from the base fraction through adsorption or occlusion by the © barium precipitate. We therefore made a practice of reducing this loss to a minimum by completely dissolving the phosphotung- states of the bases with alkali, and precipitating the barium phos- photungstate in a dilute solution.* In the more successful of our present determinations by Kossel’s method we have dissolved the lysine phosphotungstate in ammonia and diluted the solution to a large volume before treating with barium hydrate. * Ber, d. deutach. chem, Gesellsch., xliii, p, 3177, 1910, footnote. * This Journal, x, p. 16, 1911. ¢ Ibid, x, p. 25, 1911. Donald D. Van Slyke 533 First analysis by Kossel’s method. In this determination the casein was completely hydrolyzed with hydrochloric acid and all the bases were precipitated with phos- photungstic acid. The precipitate was decomposed with barium hydrate, using a large volume of solution and a mechanical stirrer to make the decomposition quantitative and keep loss by adsorp- tion as low as possible. The bases were then separated by Osborne, Leavenworth, and Brautlecht’s modification of Kossel and Pat- ton’s method. - The details follow. Forty grams of Merck’s ‘‘Casein nach Hammarsten”’ were boiled thirty hours with 400 ec. of 20 per cent hydrochloric acid. The solution was, diluted to 1 liter, and three samples of 5 cc. each removed for Kjeldahl de- terminations. The amounts of ;5 acid neutralized were 18.70, 18.67, and 18.65 ec., the average indicating 5.15 grams of nitrogen in remaining 985 cc. of solution. ¥ The latter was diluted with water to 2 liters, and the bases were precipi- tated with 150 grams of purified phosphotungstie acid. After two days the precipitate was filtered with suction and washed with a solution of 2.5 per cent phosphotungstic acid in 5 per ‘cent sulphuric acid until the chloride reaction disappear rom the filtrate. The precipitate was then suspended in 5 liters of water and thoroughly stirred with a machine while an excess of barium I drate solution was added. The stirring was then continued for abou wo | ours. The filtrate from the barium phosphotungstate was con- ‘trated in a vacuum, the ammonia being driven off in the process. The ss barium was then removed with carbon dioxide, and the solution con- centrated to 1000 cc. Twenty-five cc. were removed for analyses, which gave a total nitrogen of 1.176 grams, or 22.8 per cent of the entire casein nitrogen, and an amino nitrogen of 0.754 gram, or 14.63 per cent. __ The remaining 975 cc. of the solution, containing the basic portion of §.02 grams casein nitrogen, were concentrated in a vacuum, and the his- _ tidine precipitated as described by Osborne, Leavenworth, and Brautlecht. _ The histidine solution was brought to 100 cc. volume. 2.000 ec. for NH determination gave 2.51 ce. N gas at 21°, 774 mm. 10.00 cc. for Kjeldah! determination neutralized 14.92 ce. of 7h acid. Amino nitrogen in histidine solution, 0.0723 gram. Total nitrogen in histidine solution, 0.2090 gram = 4.16 per cent of total casein nitrogen. Ratio (total nitrogen) : (amino nitrogen) = 2.89 Ratio calculated for histidine . = 3.00 The arginine solution was also brought to 100 ce. - 2.000 ec. for NH: determination gave 3.41 cc. N gas at 23°, 764 mm. 5.00 ce. for Kjeldahl determination neutralized 13.44 and 13.50 ee. of to acid. 534 Hexone Bases of Casein Amino nitrogen in arginine solution, 0.0962 gram. Total nitrogen in arginine solution, 0.3770 gram = 7.51 per cent of total casein nitrogen. Ratio (total nitrogen) : (amino nitrogen) = 3.92 Ratio caleulated for arginine = 4.00 The filtrate from the arginine was freed from barium and silver, and the lysine reprecipitated with phosphotungstic acid at 1 liter volume. The lysine phosphotungstate was dissolved in 2 liters of dilute ammonia and freed from phosphotungstic acid with barium hydrate. The ammonia was boiled off in a vacuum, the excess barium removed with carbon dioxide, and the lysine solution brought to 100 ee. ‘ 1.000 ec. for NH» determination gave in 0.5 hour 8.28 cc. N gas at 20°, 746 mm. Ps. TM, 5.00 ce, for Kjeldahl determination required 16.48 cc. of to acid. ail 3.00 ce. for Kjeldahl determination required 9.71 cc. of 74 acid. Amino 1 nitrogen in lysine solution, 0.462 gram = 9.21 per cent of total casein nitrogen. Total nitrogen in lysine solution, 0.461 to 0.460 gram. Ninety cc. of the solution were concentrated to 40 cc., heated to boiling and 1 equivalent (3.39 grams) of picric acid was dissolved in the hot solu- tion. The lysine picrate obtained weighed, when dried to constant weight, 5.063 grams, equivalent to 0.378 gram of lysine nitrogen in the 90 cc. of the original solution, or 0.422 gram in the total solution, the latter amount being 8.39 per cent of the nitrogen of the casein. Allowing for the solubility of lysine picrate in water at 20° (0.5 gram per 100 cc.) increases the lysine picrate to 5.26 grams, and the percentage of lysine nitrogen in casein to 8.70. The lysine which crystallized was analytically pure. ANALysiIs: 0.1532 gram substance; 20.6 cc. N at 22°, 748 mm. (nitrous acid method). alculated for CeHpOst NH) CeHsNsOr: = Found: Amino-nitrogen.. .. 4... «i sesiabe? s 7.47 7.45» Of the total amount of nitrogen in the lysine fraction, 92 per cent was recovered as pure lysine picrate, or 95 per cent if the solubility of the pierate is taken into account. This indicates strongly that in casein which has been completely hydrolyzed and freed from ammonia there are, aside from arginine, histidine, ly- sine, and the cystine not destroyed by the hydrolysis, no amino- acids precipitated by phosphotungstic acid under the usual con- ditions, 7.e., at room temperature, in the presence of mineral acid of ¥ concentration and a moderate excess of phosphotungstic, the concentration of the proteolytic products being about 2 per cent. It will be noted that all the non-amino nitrogen of the total Donald D. Van Slyke 535 phosphotungstate precipitate is accounted for by two-thirds the . histidine nitrogen + three-fourths the arginine. Of the 14.6 per cent amino nitrogen in the first precipitate, however, the amount _recovered in the histidine, arginine, and lysine fractions respectively was 1.19 + 1.88 + 9.21 = 12.28 per cent. The 2.3 per cent loss is doubtless in part due to adsorbed lysine, but a portion is probably due to mono-amino-acids which were not completely removed from the first phosphotungstate precipitate. The latter was washed on a suction funnel in the ordinary manner, and not triturated in the way which we have found necessary for a quan- titative washing of these precipitates. As the bases were repre- cipitated separately before they were determined, however, the final results are not affected. Second analysis by Kossel’s method. In this analysis we attempted to keep possible losses from ad- _ sorption by barium phosphotungstate at a minimum. The pre- liminary precipitation of all the bases with phosphotungstic acid was left out, and the regular technique of Osborne followed for the precipitation of arginine and histidine directly from the solution of all the amino-acids by means of silver nitrate and barium hy- drate. The lysine was then precipitated as phosphotungstate. In “the decomposition of the latter, it was completely dissolved with ammonia and diluted before treatment with barium hydrate. in The casein was hydrolyzed as in the first analysis and the solution diluted ¢ to 1 liter. Three samples, of 5 ce. each, taken for Kjeldahl determinations required 19.55, 19.65, and 19.75 cc. of 7 acid, the average indicating 5.42 grams of nitrogen in the remaining 985 cc. of solution. The latter was freed as completely as possible from hydrochloric acid by concentration under diminished pressure. The residue was taken up in water, made alkaline with barium hydrate, and the ammonia was driven off by concentrating again under diminished pressure. The solution was then acidified with ni- trie acid, and the remaining hydrochloric acid removed with silver nitrate. __ The filtrate from the silver chloride was brought to a volume of 1 liter, and __ the histidine and arginine precipitated with excess silver nitrate and barium hydrate. The solution of the two bases was brought to 250 ec., from which 12 ce. were taken for amino and Kjeldahl determinations. These showed 0.707 gram nitrogen and 0.211 gram amino nitrogen present. The remaining 238 ec. of the solution, containing the arginine-histidine fraction of 5.16 THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 i 536 Hexone Bases of Casein grams of casein nitrogen, were submitted to the Osborne modification of Kossel and Patton’s technique for separation of the two bases by precipi- tation of the histidine with mercuric sulphate. The histidine solution was brought to 100 ce. 2.000 ce. for NH: determination gave (1) 2.59 cc. N gas at 17°, 779 mm. ; (2) 2.69 ce. at 25°, 765 mm. 3.00-ce. portions for Kjeldahl determination required 5.00 cc. each of — acid. Amino nitrogen in histidine solution, 0.0758 to 0.0752 gram; average, 0.0755 gram. Total nitrogen in histidine solution, 0.2330 gram = 4.51 per cent of " the casein nitrogen. Ratio (total nitrogen) : (amino nitrogen) = 3.09 ia Y__ Ratio calculated for histidine _ = 3.00 The arginine solution was also brought to 100 cc. 2.00 for NH: determination gave 3.60 cc. N gas at 17°, 770 mm. 3.00 ee. for Kjeldahl required 8.60 to 8.65 ec., average 8.63 cc., 7 acid. Amino nitrogen in arginine solution, 0.1052 gram. Total nitrogen in arginine solution, 0.4040 gram = 7.83 per cent of the casein nitrogen. Ratio (total nitrogen) : (amino nitrogen) = 3.84 Ratio calculated for arginine = 4.00 From the filtrate of the first arginine-histidine precipitate the lysine was, precipitated as phosphotungstate in the usual manner. The precipi- tate was redissolved in 2 liters of dilute ammonia and the phosphotungstic acid removed by addition of barium hydrate. The filtrate from the barium phosphotungstate was freed from ammonia by concentration under dimin- ished pressure, from barium by means of carbon dioxide, and was then brought to 100 ce. Analysis showed that 0.605 gram of amino nitrogen was present. To 75 cc. of the solution one equivalent (3.75 grams) of picric acid was added, and the mixture was heated until solution was complete. After two days’ standing 4.69 grams of lysine picrate, equivalent to 6.26 grams for the entire 100 cc. of solution, crystallized analytically pure. The amount of lysine nitrogen calculated from the picrate is 0.467 gram, or 8.62 per cent of the total nitrogen of the casein. ANALysIs: 0.1433 gram substance; 18.85 cc. N gas at 19.5°, 760 mm. freed by nitrous acid method. Calculated for CoH002(N He)2.CeHsNsO7; = Found: Bitte Dittomeiiss. .... iokeds oso one 7.47 7.50 The filtrate from the above main crop of picrate was acidified with sul- phuric acid, freed from picrie acid with ether, and treated again with phos- photungstic acid in a volume of 100 ce. A second crop of lysine phospho- tungstate was obtained, which eventually yielded 0.40 gram of pure lysine picrate. ANALYsis: 0.0327 gram substance; 4.35 cc. N gas at 24°, 760 mm., by the nitrous acid method with micro-apparatus. Donald D. Van Slyke 537 Calculated for CsHiw02(NH2)2.CeH3sNsO7: Found: Amino nitrogen..................... 7.47 7.42 This second crop of lysine brings the total lysine nitrogen up to 0.507 gram, equal to 9.36 per cent of the casein nitrogen. Analysis by the nitrogen distribution method. Ten grams of casein were boiled twenty-four hours with 200 cc. _ of 20 per cent hydrochloric acid.? The acid was driven off as com- _ pletely as possible by concentrating under diminished, pressure, and the solution was brought to 150 ec. Three samples of 5 cc. each required 35.05, 35.20, and 35.05 cc., an average of 35.10 ce., ge. A of > acid in Kjeldahl determinattinail Tot the nitrogen - _ bution’ 50 cc. of the solution, containing 0.491 gram of nitrogen, were taken. The analysis was performed sggpeeetibed in the _ paper on the method.‘ Fr Ammonia. The amount of 3) acid neutralized was 36.00 ec., equivalent to 0.0504 gram of ammonia nitrogen. Melanine. The amount of 35 acid neutralized was 4.50 cc., equivalent to 0.0063 gram of melanine nitrogen. Cystine. The weight of barium sulphate was 0.0035 gram, equivalent to 0.0010 gram of cystine nitrogen i in all. Ar, nine. The volume of jy acid neutralized was 6.50 cc., equivalent to 0. gram of arginine nitrogen. Total nitrogen of the bases. The amount of 74 acid neutralized in the - Kjeldahl determination was 35.65 cc. Added to the amount neutralized in the arginine determination, this gives 42.15 cc., equivalent to 0.1181 gram of nitrogen, or 24.27 per cent of the total nitaaeen of the casein. _ Amino nitrogen of the bases. Two-cc. portions of the solution of the bases gave, in the micro-apparatus, 5.02 cc. of nitrogen gas at 23°, 778 mm., _ and 4.94 cc. at 20°, 778 mm., indicating respectively 0.0722 and 0.0719 gram of amino nitrogen in the bases, equivalent to 14.6 per cent of the total nitrogen of the casein. Amino nitrogen of the filtrate. Duplicates gave each 31.40 cc. nitrogen - gas at 20°, 781 mm., equivalent to 0.274 gram of amino nitrogen. D: _ Total nitrogen of the filtrate. The amounts of 7% acid neutralized were 36.65 and 36.90 ce., the average, 36.78 cc., indicating 0.309 gram of nitrogen in ‘the filtrate. eho results are tabulated on the ‘gerne page. FE 1 Van Slyke: Conditions for Complete Hydrolysis of Proteins, this Journal, xii, p. 295, 1912. * This Journal, x, p. 15, 1911. 58 °c Hexone Bases of Casein NITROGEN | TOTAL NITROGEN grams per cent A WATAOMA wae ee 8. 6s eee ret ee 0.0504 10.27 Moelanimeinimeete... os <2 ----- 3 hee 0.0063 1.28 Cystint:Gabeeee nee 0.0010 0.20 Arginin@ste geese + ee 0.0364 7 Al Tistiditiveteeaas.s-- s--- seer | 0.0305 6.21 Viyeilie-vaete = tee | 0.0506 10.30 Amino nitrogen of filtrate........--- 0.2740 55.81 Non-amino nitrogen of filtrate....... 0.0350 7.13 Re. ; ee Total nitrogen recovered.....-.--- 0.4842 98 .61 co SUMMARY. . The following figures for the bases of casein were obtained by the method of Kossel, and by the author’s nitrogen distribution — method. The figures represent percentages of the total nitrogen — of the casein. + ke ie } " at Se ae — ae KOSSEL’S METHOD | NITROGEN DISTRIBU- ieee oe TION METHOD (Uncorrected for solu- First Analysis Second Analysis bility of bases) Histidine........--.---> 4.16 4.51 6.21 Arginine........-.++++: 7.51 7.83 7A Lysine......6.+--++++4 8.70 9 .36 10.30 | It appears probable that low results for lysine (7 per cent) ob- tained previously by the Kossel method were due to adsorption or occlusion of lysine by barium phosphotungstate, a source of error which we attempted to avoid, especially in the second Kossel analysis. That the lysine crystallized as picrate represents the entire amount present is improbable, and the lysine content obtained by the nitrogen distribution method is doubtless more nearly correct. From the data in this and previous papers,’ which permit a comparison of results by the two methods, it appears that the nitrogen distribution method if somewhat more reliable than the Kossel method for lysine determination in proteins, that both methods are quite accurate for arginine, and that the Kossel-Pat- ton method, as modified by Osborne, Leavenworth, and Brautlecht, gives more consistent results for histidine. * This Journal, x, p. 16, 1911. THE NATURE OF THE FREE AMINO GROUPS IN PROTEINS. By DONALD D. VAN SLYKE anp FREDERICK J. BIRCHARD. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) P (Received for publication, November 28, 1913.) Z — The presence of basic groups in the proteins has long been as- _ sumed because of the ability of the proteins to neutralize acids. _ The specific nature of these basic groups appears to have been _ first indicated by work on the protamines in Kossel’s laboratory. _ These simplest proteins are all unusually rich in one or more of _ the hexone bases, arginine, histidine, and lysine, and are also markedly basic, forming salts of constant composition with sul- phurie acid. Goto! found that clupeine, which contains a large amount of arginine, binds approximately 1 equivalent of sulphuric acid for each molecule of arginine. Arginine contains two amino groups, one (a) in the a-position to the carboxyl group, the £g a NH.—CH(NH)—NH-—(CH:);—CH(NH:) -COOH 7 FANT a ~ other (g) in the guanidine nucleus, and it was uncertain which of _ these was free in the clupeine molecule. The point was settled by - Kosseland Cameron?in favor of the guanidine group. They nitrated the free amino groups of clupeine under conditions which avoided _ hydrolysis of the protein. The nitroclupeine was then hydrolyzed, and nitroarginine obtained from it. The amino group in this Be rorernine could be determined by the nitrous acid method. As» he guanidine group does not react with nitrous acid, the amino group freed by hydrolysis was evidently the a-group. The other ' Zeitschr. f. physiol. Chem., xxxvii, p. 114, 1902. * Thid., \xxvi, p. 457, 1912. * 339 540 Nature of Free Amino Groups in Proteins amino group, free in the intact protamine, is therefore in the guan- idine nucleus. Other evidence indicated that, in some proteins at least, one of the two amino groups of lysine, NH>.(CHe)4-CH(NHe).COOH, is also free. Skraup* found that after casein, gelatin, and serum globulin had been treated with nitrous acid, no lysine could be obtained from them. Similar results were obtained by Levites.* Van Slyke demonstrated in edestin and egg albumin the presence of definite amounts of free amino nitrogen, which, though small, could be determined by the nitrous acid method, and pointed out Sine fact that, considering the results of Levites and Skraup, it was probable that one of the amino groups of lysine furnished a large part of the free amino nitrogen of the proteins.’ Since then, having determined the lysine contents of a number of protein preparations in this laboratory, we have also determined their free amino nitrogen. The results, reported in detail in the present paper, were published in a preliminary abstract in May, 1912.° ~ They showed that the free amino nitrogen of the native proteins is approximately one-half of the lysine nitrogen, indicating that one of the two amino groups of lysine is free in the protein mole- cule. At the time of our preliminary report the paper by Kossel and Cameron also appeared.’ They settled the location of the free amino group of arginine in clupeine, as mentioned above, and also showed that clupeine, which contains no lysine, gave off no | nitrogen when treated with nitrous acid to determine the free amino — groups. Cyprinine and sturine, which contain lysine, showed con- — siderable free amino nitrogen. Later Kossel and Gawrilow® per- formed the formol titration of Sérensen on hordein, zein, and sev- — eral protamines, and found that the protamines containing lysine revealed amino groups by the formol method, while those which contained none, as well as zein, Which also contains none, re- vealed no amino groups. No quantitative relations were ascer- . Ann. d. Chem., cccli, p. 379, 1906. ‘ Biochem. Zeitschr., xx, p. 224, 1909. ‘ This Journal, ix, p. 196, 1911. ¢ Proc. Soc. Exp. Biol. and Med., May 15, 1912. ? Loe, cit. * Zeitachr f. physiol, Chem., Ixxxi, p. 274, 1912. D. D. Van Slyke and F. J, Birchard 541 tained, however, between the lysine and free amino nitrogen con- tents of the proteins.°® In our preliminary report we gave determinations of free amino * nitrogen on several native proteins, and quoted those previously published from this laboratory on proto- and heteroalbumose. The native proteins showed, as stated above, free amino nitrogen equal to approximately half their lysine nitrogen. In the hetero- and protoalbumose we found more of the total nitrogen in the form of free amino nitrogen than is calculated by halving the lysine nitrogen. This result was to be expected, as the hydrolytic cleay- _ age from which the albumoses result sets free a-amino groups from) _ the peptide linkings into which they are condensed in the native proteins. _ We have in the meantime analyzed, in addition, gliadin, re- peated with present technique the amino determimation performed _ several years ago on the albumoses, and confirmed the results re- _ ported for the native proteins, except casein. The amount of _ free amino nitrogen in casein (3.4 per cent of the total nitrogen) _ was, through an error, calculated too low. The correct figure is 5.5 per cent. The former, incorrect figure was almost exactly one- half the lysine nitrogen previously determined in casein. The lysine content of casein had been determined some years before by the picrate method, but in the case of casein, unlike most of the other proteins, the estimation had not been checked by our _ group determination method in its present form. The hexone base content of casein was, therefore, carefully redetermined by ie both the Kossel and the group determination methods. The results are recorded in the preceding paper. The correct lysine nitrogen was found to be, as in the other proteins, nearly twice the free amino nitrogen. For the hemocyanin preparation we thank Dr. C. L. Alsberg - of the Bureau of Chemistry; for the hemoglobin, Dr. Butterfield of this Institute, and for the zein, Dr. Thomas B. Osborne of New Haven. ®“Tm allgemeinen scheinen die lysinreicheren Prptamine auch reicher an formoltitrierbarem Stickstoff zu sein, doch sind die bisher vorliegenden Analysen noch nicht zahlreich genug, um hieriiber zu entscheiden.”’ a 542 Nature of Free Amino Groups in Proteins EXPERIMENTAL, Methods. The proteins were brought into solution in 2 to 4 per cent concentration, using when necessary acetic acid or sodium carbonate in the cold to assist the process. There was no evi- dence of the occurrence of any hydrolysis during the preparation of the solutions. These were analyzed immediately after they were made up. When they were allowed to stand several hours, only slight increases in the amino nitrogen were noted. All de- terminations were made in the standard size amino apparatus As em@pdescribed in this Journal, xii, p. 275, or in the micro-apparatus deseribed in xvi, p. 121. The mixtures were shaken constantly witha oo, each determination, octyl alcohol being used to prevent foaming. The proteins, or deaminized proteins, are pre- cipitated as soon as they are mixed with the nitrous acid solution. When the mixture is kept well stirred by shaking, however, the precipitation does not appear to influence the results, which were uniformly definite and constant. That the amino nitrogen thus determined represents amino groups free in the protein, none of the latter being hydrolyzed by the nitrous acid, is indicated by two facts: 1. Peptides of varied composition and containing up to four- teen amino-acids in the molecule have been analyzed by our method and found to give theoretical results.'° . 2. The evolution of nitrogen is complete inside of twenty or thirty minutes, following practically the course found in analysis of lysine," of which the w-NH, group reacts somewhat more slowly than the a-groups of the amino-acids in general. We do not believe that any part of the amino nitrogen determined comes from acid amide groups in the protein molecule. As determined by analysis of asparagine and acetamide, acid amide groups give off no nitrogen at all when treated with nitrous acid under the con- ditions of the determination. Casein, Three grams of Hammarsten casein (air-dried) were dissolved — in 100 ec, of water with 0.375 gram of sodium carbonate. This amount of carbonate is sufficient to dissolve the casein without rendering the solu- tion alkaline, and autohydrolysis occurs only at a very slow rate. '© Abderhalden and Van Slyke: Zeitschr. f. physiol. Chem., xxiv, p. 505, 1911, Van Slyke: this Journal, xii, p. 275, 1912. D. D. Van Slyke and F. J. Birchard 543 Kjeldahl nitrogen: 5-cc. portions; 14.80 ec. of 7 HCl (average of 3 de- terminations), indicating 8.29 mgm. of nitrogen in the 2 cc. of solution used for amino determination in the micro-apparatus. In this table the uncor- rected as well as corrected results are given to show the magnitude of the correction for reagents, and its variation with the reaction time. In the subsequent tables only corrected results are given. Amino nitrogen: 2-ce. portions. | Att / } DURATION OF |NITROGEN GAS EVOLVED PER CENT OF wiTH "NITROUS Volume riooreaail “TURE i. en yuan Nu ACID for reagents min. oe | ce. | deg.C. | mm. mgm. | 10 0.86 | 0.76 | ° 22 756 | 21130 | 5.14 al 15 092 | os0 | 23 | 756 | 2.230 | 5am ° 20 0.96 | 0.82 | 23-| 756 | 2.285 | .«iiimE 30 098 0.82 | 23-56 | 2.285 | 5.1 After the solution had stood forty-sighiil hours at_ temperature the proportion of nitrogen as free NH, had increased to 5.84 per cent (thirty- minute reaction), indicating an appreciable, but very slow, autohydrolysis. Gelatin. Kjeldahl nitrogen: 10-ce. portions; 24 cc. 7) HCl. Total ni- trogen in 10 ce., » Shana egy a nitrogen: 10-ce. portions. DURATION : | | PER CENT OF _ OF N Gas TEMPERATURE | PRESSURE amino N TOTAL N as REACTION Be i mt. Ue FREE NH: min. ce. f deg. C. mm, | mgm. 10 1.80 23 ek + 026) | 9.05 30 1.90 28 762 1.048 | 3.12 30 1.90 28 762 1.048 | 3.12 Ox hemoglobin. Solution I. Kjeldahl nitrogen: 10-cce. portions; 12.93 cc. 7y HCl;18.10mgm.N. SolutionTI. Kjeldahl nitrogen: 10-cc. portions; ‘ 10 cc. 4 HCl; 14.01 mgm. N. Amino nitrogen: 10-cc. portions. ] NO. CG eas | Teen | PRESSURE amino N Pan OE eas REACTION iB FREE NH: min. ce. deg. C. mm, mgm. : Ta 10 neo. | 1898 760 1.033 | 5.70 Ib 10 1.80 18 4 760 1.033 5.70 Ila 30 1.50 26: ea 75 0.825 | 5.89 IIb 30 1.60 26 | 7 0.881 6.29 } | 544 Nature of Free Amino Groups in Proteins Edestin. Solution I. Kjeldahl nitrogen: 10-cc. portions; 24 ec. 7s HCl; 37.60 mgm. N. Solution II. Kjeldahl nitrogen: 10-cc. portions; 19.33 ce. +5 HCl; 27.06 mgm. N. Amino nitrogen: 10-cc. portions. | 10.) | Mos | ee PRESS URE aurvo N- | Toran Nas: | REACTION | | FREE NH ' min, | cc. =| «deg. C. | mm. mgm. Loapet 1.00. |, 20°) ae 0.570 | 1.69 | tk® 1.00 20 | 762 0.570 | 1.69 | 30 0.80 25 | 766 0.448 | 1.65 | 30 0.90 26 | 758 0.495 1.83 Hemocyanin. The substance was oropai for Pacatysis “ pulse to a fine powe en rubbing it up with 5 per cent sodium carbonate until a colloidal ion w: btained. This was poured into an excess of glacial acetic acid, and the ure formed a clear solution, which was diluted with water. i Kjeldahl nitrogen: 5-<0. pomions; 20.4 ce. gy HCl. Total nitrogen im 10 ce., 57.12 mgm. . i ‘Amin’ nitrogen: 10 ec. of 20°, 765 mm.; amino N, 2.41 mgm.; per ¢¢ gen, 4.28. oy Zein. The substance was dissolved in ell acetic acid. Kjeldahl nitrogen: 10-cc. portions; 24.36 ce. 7 HCl; total N, 34.00 mgm. Amino nitrogen: 10 cc. of solution gave in thirty minutia the same » ume of gas as 10 ce. of glacial acetic acid alone in the control determina- _ tion. Amino nitrogen not present. i: Gliadin. Two grams of gliadin were rubbed up with 5 ce. of glacial acetic | acid, the turbid solution diluted to 40 ce. with water, and cleared by centri- fugalizing. Kjeldahl nitrogen: 5-cc. portions; 22.65 cc. ~y HCl, indicating 12.70 mgm. of nitrogen in 2 cc. in thirty minutes, 4.20 cc. at pritrogen as amino nitro- Amino nitrogen: 2-cc. portions. aa | eee Bor N as | TEMPERATURE | PRESSURE amino N | ® gro a *min, ce. | deg. C. mm, mgm. G4 oP) | (OO. Sie 21 764 0.12 0.94 30 0.25 21 764 0.14 1.10 Heteroalbumose (from Witte peptone), dissolved i in 0.5 per cent N x00, solution. Kjeldahl determination: 5-ce. portions; 11.65 cc. 4) HCl, indicating 6.52 mem. N in 2 ce. solution. D. D. Van Slyke and F. J. Birchard 545 Amino nitrogen: 2-ce. portions. % DURATION ye | | PER CENT OF ie nok N Gas | TEMPERATURE PRESSURE | AMINO N xOeas an min. ce. ie 9 mm. mgm. 15 08 | 2 762 | 0.488 7.48 30 OO | A he SR, 0.636 8.06 Protoalbumose (from Witte peptone). Solution I. 0.750 gram albu- mose dissolved in 25 cc. water. Kjeldahl nitrogen: 5 cc. portions; 15.44 ce. ty HCl, indicating 8.65 mgm. of nitrogen in 2 ee. Solution II. 0.750 gram albumose dissolved in 25 ce. of 0.5 per cent NazCO; solution. Kjeldahl nitrogen: 5-ce. portions; 15.49 cc. 4 HCl, i 4 dicating 8.67 mgm. of nitrogen in 2 ee. Pa —— Amino determinations: 2-cc. portions. DURATION ee? CENT jon NHs IN sitehaa mmaction i | oh es: | scomaieg fe aT ora N nig min. ce. deg. C. mm. mgm. I 5 | 1.394 | 19 °1%Nmes | 0.773 | 8.04 II 5 1,32 18 | 764 | 0.760 | 8.77 I 15 ano 24 762 | 0.953 | 11.03 II 15 eeees 18 764 | 0.881 | 10.16 I. \aapeso 1.74 24 762 | 0.975 | 11.28 | 9.93 I | 30 1.67 18 764 | 0.965 | 11.13 | 9.78 The final, thirty-minute results, are no higher when sodium carbonate is used to assist solution (solution II) than when pure water is used (solution I). This fact shows that the dilute sodium carbonate solution had no immediate hydrolyzing effect on the protein. Unlike the heteroalbumose, the preparation of protoalbumose™ contained an appreciable amount of free ammonia. Portions of 0.500 gram, used for determination of free ammonia by vacuum distillation from solution made alkaline with calcium hydrate, gave 0.75 and 0.80 cc. of 7) ammonia, equal to 1.46-1.55 per cent of the total nitrogen. In thirty minutes approxi- mately 90 per cent of the nitrogen of ammonia is given off in the amino de- termination. The free amino nitrogen determined should therefore be re- duced by 1.35 per cent of the total to correct for the ammonia. The figures for free amino nitrogen in hetero- and protoalbumose are higher than those given in the papers from this laboratory on the composi- 12 Levene, Van Slyke, and Birchard: this Journal, viii, p. 269, 1910. 546 Nature of Free Amino Groups in Proteins tion of the albumoses.'* The earlier determinations were run for intervals of only five minutes. These suffice for all the other amino-acids, but are not long enough for complete decomposition of the w-NH: group of lysine. Summary of results. In the following table the average results of the complete (thirty- minute) determinations are collected. The figures obtained by halving the lysine nitrogen are given for comparison. PER CENT OF TOTAL NITROGEN AS PROTEIN ANALYZED = <> an = | Free — One-half - lysine A. Native Sal 4 Hemogl oe 6.00 5 .80* Casein. . a TES ae . Wael 5 .15t Hemoeyanin ee MM. siete ss 4.30 | 4.25t Gelatini. |... (ARE. . 5... | 3.12 | 3.15t Edestin.............ntaee. 1.80 1.90t Gliadin....... ... jee: oye oo 1.10 0 .38t Zein ...5 «+. «ov eh hes 0.00 0.00§ B. Albumoses from fibrin: Ps Heteroalbumose................... 8.06 5.15** | * 9.86 | 4.80** Protoalbumose: 2... 3.:..cs seb encwe * Unpublished result. t Van Slyke: Preceding pa t Van Slyke: This packets x, p. 16, 1911. § Osborne and Jones: LErgeb. d. Physiol., x, p. 99, 1910. ** Levene, Van Slyke and Birchard: This Journal, x, p. 57, 1911. CONCLUSIONS, In all the native proteins investigated the amount of free amino nitrogen is equal to one-half the lysine nitrogen, no deviation exceeding the limit of experimental error of the amino and lysine determinations being found in any case with the possible exception of gliadin, in which the difference is 0.7 per cent. The period required for complete reaction of the proteins with nitrous acid (thirty minutes) is longer than that required by the a-amino groups (three to four minutes), but corresponds to that found for Levene, Van Slyke, and Birchard: this Journal, viii, p. 272, 1910; x, p. 50, 1911, D. D. Van Slyke and F. J. Birchard 547 lysine, with an w-amino group free. The facts support the following conclusions. 1. One of the two amino groups of lysine, the w-group, exists free in the protein molecule. 2. This group represents, within at most a fraction of a per cent of the protein nitrogen, the entire amount of free NH, determin- able in the native proteins by the nitrous acid method. The a-amino groups, which constitute the remaining and greater part of the free amino nitrogen found after complete hydrolysis, are, in the intact protein molecule, practically all condensed into peptide linkings. J 3. With the primary albumoses the relations are different. _ The free NH, in hetero- and protoalbumose exceeds half the ly- sine nitrogen by 3 and 4.8 per cent, respectively, of the total pro- tein nitrogen, indicating that an appreciable on of the a-amino groups is uncovered in even the primary digestion products. 4 Osborne, Leavenworth and Brautlecht have demonstrated the prob- able presence of the acid amide groups of glutamine and asparagine in the protein molecule (Amer. Journ. of Physiol., xxiii, p. 180). Acid amide _ groups, however, like the guanidine nucleus of arginine, give off none of their nitrogen when treated with nitrous acid, and consequently are not determined by our method. > ON SPHINGOSINE. SECOND PAPER. THE OXIDATION OF SPHINGOSINE AND DIHYDROSPHINGOSINE. By P. A. LEVENE anp C. J. WEST. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) In 1911 Levene and Jacobs! announced the first information regarding the chemical structure of sphingosi They regarded the substance as a dihydroxy derivative of an unsaturated pri- mary amine. Later Thierfelder and Thomas* corroborated some of these conclusions. Levene and Jacobs* in their full publication on sphingosine stated that further work on the structure of the base was in progress in this laboratory. It was made clear that the respective positions of the hydroxy groups, of the amino group and of the double bond were under iavestigation. The progress of the work was not quite so rapid as expected i. for the reason that the reduction of dihydrosphingosine into the _ primary amine offered unexpected difficulties. In the course of _ these reduction experiments unexpected substances were obtained _ which may prove of considerable interest and the study of which somewhat delayed the completion of the work. While this work was in progress there appeared during the course of the present summer a publication by Lapworth‘ on the structure of sphingosine. The investigations of Lapworth were begun in 1910, but the experiments reported in his publication are apparently based on the knowledge of the structure of the sub- stance furnished by the work of Levene and Jacobs. Lapworth demonstrated that on oxidation of the base with chromium tri- (Received for publication, December 1, 1913.) . ! Levene and Jacobs: this Journal, xi, p. xxix, 1912. 2 Thierfelder and Thomas: Zeitschr. f. physiol. Chem., xxvii, p. 511, 1912. ’ Levene and Jacobs: this Journal, xi, p. 547, 1912. ‘ Lapworth: Journ. Chem. Soc., ciii, p. 1029, 1913. 549 550 Sphingosine oxide a tridecylic acid was obtained, which he regarded as the normal acid and for this reason argued that in sphingosine the carbon atoms are linked in a straight chain. Our work on the structure of sphingosine is as yet not completed but we wish to present some of the results of our experiments, particularly in view of the publication of Lapworth. Only those experiments will be discussed in the present communication which deal with the oxidation of sphingosine and dihydrosphingosine. On oxidation of the unsaturated base a tridecylic acid was ob- tained, while the reduced base under the same conditions of ex- a erimentation gave rise to a pentadecylic acid. Regarding the silliai ie. ucture of the carbon chains of the two acids, we as yet have no definite information. On the basis of the melting points of the acids one seems justified in concluding that the carbon atoms are not linked in a straight chain, since the normal tridecylic acid has a melting point of 43°,5 the normal pentadecylic acid melts at 53°, while the melting points of our two acids were 47—48° and 60-61° respectively. On this point our results do not agree with those of Lapworth. It will be the aim of the future work to es- tablish the exact structure of the carbon chains of these two acids. Our results, however, are important principally for the reason that through them the position of the double bond is made clear, namely, between the fourth and fifth carbon atoms (from the right end); and further, that through them the possibilities of the posi- tion of the two hydroxy and of the amino groups were limited to _ the carbon atoms 1, 2 and 3. On the ground of this one may express the structure of sphingosine approximately as follows: Cy,H,»CH = CH.CHOH.CHOH.CH.N He. . 5 4 3 2 1 The nature of the carbon chain of the part from 5 to 17 is not yet clear and the distribution of the hydroxy and amino groups on the carbon atoms 1-3 is not yet determined. However if the position of a hydroxy! or of an amino group had been removed further than the third earbon atom. then on oxidation of the dihydrobase, instead of the pentadeeylic acid, a hydroxy- or an amino-acid should have been formed. * Le Sueur: Journ, Chem. Soc., \xxxvii, p. 1905, 1905, P. A. Levene and C. J. West 551 EXPERIMENTAL PART. Our first experiments in the study of the oxidation products of sphingosine were attempts to repeat the work of Lapworth. We ' added an excess of chromic acid to a glacial acetic acid solution of _ sphingosine sulphate, keeping the reaction at about 70° on the water bath, the lowest temperature which caused effervescence. The excess of chromic acid was then reduced with sulphur dioxide and the reaction product distilled with steam. Only a very small amount of solid distillate passed over with the steam, 5 grams of sphingosine sulphate giving not more than 0,5 gram of crude acid. There remained in the flask an oily, green residue, soluble. | of in ether, which we attempted to reoxidize, with little success, how- ever. This green product was then boiled with concentrated hy- drochloric acid, but it did not give a colorless,a¢id as stated by Lapworth. Since then we have found that this oil contains most of the reaction product which may be obtained by distillation in vacuum, but have not repeated the work under exactly these conditions. In view of this we do not wish to condemn the method until we have tried it again. After this we tried many conditions of oxidation and finally found the following to be the most suitable. Three grams of sphingosine sulphate were dissolved in about 50 cc. glacial acetic acid, warmed on a boiling water bath to about 85-90°, and a warm _ solution of 12 grams of chromic acid in 120 ce. glacial acetic acid ’ slowly dropped into it, the flask being shaken quite frequently. . After the addition of the acid, the reaction product was diluted _ with water and distilled with steam until nearly all of the acetic acid had been removed. A small amount of solid acid distilled over with the steam, which, after being dried in ethereal solution, _ weighed 0.350 gram. The main part of the acid was found in the _ oily, green residue, which floated on the surface of the dilute acetic _ acid in the flask. This solidified upon cooling in the ice box, and _ after drying the ethereal solution weighed 1.03 grams. It con- _ tained about 10 per cent ash. The theory from 3 grams sphingo- sine sulphate is about 1.9 grams. The green product, when dried in vacuum, gave the following _ numbers upon analysis, which, calculated on an ash-free sub- ' stance, indicated the presence of a Cys acid. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 552 Sphingosine 0.1170 gram of the substance gave 0.2790 gram CO, and 0.1088 gram H,0. The ash weighed 0.0125 gram. Calculated for ~ 13 H2asO2: Found: Coie. ee es cL ee 72.80 73 .1Z | ae io ere Sa 12.20 11.82 This dried product was then distilled in vacuum, when a color- less distillate was obtained, which, when distilled twice, was found to be a pure tridecylic acid. 0.1222 gram substance gave 0.3248 gram CO» and 0.1338 gram H,0. Calculated for Pm . CusHarOs:,. : Jouhals RE een ae 72.80 72 .58 | Sse traeer a 12.20 12.25 dar weight estimation. 0.5300 gram of the acid, dissolved in abso- lute methyl aleoholjand benzene, required 24.8 ec. { NaOH for neutraliza- tion, using phenolphithalein as an indicator. pe Calculated for 3" CisHasO2: Found: Molecular welgmmmramusie....... 2c taueeeee.... ae 214 The acid on the last distillation boiled between 190-200°, a glass water pump being used. The col product melted at 46—47°, when cooled and reheated at 46-47° and after standing over nigitt and carefully heated, 47-47.5°, the last to be considered as the correct melting point. When recrystallized from dilute the melting point was not changed. = Since the melting point of the acid was higher than that fo inc by Lapworth (39-40°) and differed by 4~5° from the melting poi of the normal tridecylic acid (43°), we repeated the oxidation on a larger quantity of sphingosine sulphate. The acid, after the sec- ond distillation, melted fairly sharply at 42-43°, after recrystal- lization from pure formic acid at 42—43°, but when recrystallized from about 90 per cent acetone, the melting point was raised to 46.5-47.5° as found above. Oxidation of dihydrosphingosine. A preliminary experiment was also carried out using dihydrosphin- gosine. Three grams of dihydrosphingosine sulphate, dissolved in about 125 ce. glacial acetic acid, were treated with 12 grams of chromic acid dissolved in 120 ce. glacial acetic acid and the reac- P. A. Levene and C. J. West 553 tion product worked up as given above. The green mixture was distilled in vacuum and the colorless acid twice redistilled. The product thus obtained melted at 48-49°. When recrystallized from a little dilute acetone it melted at 60-61°. The normal pen- tadecylic acid, which we prepared from a-hydroxypalmitic acid, melted at 53°.° 0.1288 gram substance gave 0.3506 gram CO, and 0.1410 gram H,0. 0.1159 gram substance gave 0.3154 gram CO, and 0.1306 gram H,0. 0.5282 gram substance required 21.72 ec. 7g NaOH for neutralization. : Calculated for CysHs02: ~~ ~—“ Found: Ds Ral Sotho = eyes Ee 2 74.40 74.24 74.28 Ts... . 5. sole a acaleanhin 12.40 12.25 It may be mentioned that Liebermann’ dese 7 ¢ itadecylic coccery! alcohol. Whether our acid i is ide tic not yet been determined. 1 ; ¢ Levene and West: this Journal, xvi, p. 475, 1914. _ 7 Liebermann and Bergami: Ber. d. deutsch. chem. Gesellsch., xx, p. 959, 1887. met \ a, “ a = Rak ON THE ACTION OF LEUCOCYTES AND OF KIDNEY TISSUE ON AMINO-ACIDS. By P. A. LEVENE anp G. M. MEYER. (From the Laboratories of the Rockefeller Institute for Medical Research, New York.) (Received for publication, December 1, 1913.) ge 2 Since the publication of Lang! in 1904 it was generally accepted that through the action of tissues amino-acids underwent a deam- inization which led to the formation of the corresponding hydroxy- acids. It was also considered probable, on the basis of the work of Neubauer,? that the primary product of the reaction is the a-ketonic acid, which is subsequently reduced to the hydroxy- acid. Very recently the work was repeated in Cathcart’s labora- tory by Gertrude D. Bostock,’ who in the main corroborated Lang’s conclusions. This writer, however, makes a passing remark that the liver and intestinal mucosa failed to act on alanine. Considerations which were discussed in a previous publication* led us to test the action of leucocytes and of various tissues on ala- nine under conditions in which bacterial growth was completely excluded. It was then definitely proven that under absolutely aseptic conditions, or under conditions of absolutely effective antisepsis, no deaminization of alanine took place. Our observa- tion was soon corroborated in Embden’s laboratory. This made it urgent to extend the experiments to a larger number of amino- acids. The experiments were carried out under aseptic condi- tions, and it was found that not one of the amino-acids tested suffered a deaminization through the action of the leucocytes or of the kidney tissue. 1 Lang: Hofmeister’s Beitrdge, v, p. 321, 1904. * Neubauer: Deutsch. Arch. f. klin. Med., xev, p. 211, 1909. * Gertrude D. Bostock: Biochem. Journ., vi, p. 48, 1912. * Levene and Meyer: this Journal, xv, p. 475, 1913. > Griesbach and Oppenheimer: Biochem. Zeitschr., |v. p. 329, 1913. 555 556 Action of Tissues on Amino-Acids These observations make it ‘necessary to repeat all the older work on deaminizing action of tissues. Experiments on purine bases are in progress. > EXPERIMENTAL. Tissues. The leucocytes were obtained from dogs by the in- jection of turpentine into the pleural cavity. Rabbit kidneys were removed aseptically from exsanguinated animals. The tis- sues were finely minced before being added to the solutions. Solutions. Glycocoll, aspartic acid, and leucine in approximately er cent, and asparagine in 1 per cent conceniration in 1 per ; enderson phosphate mixture were used. Two kidneys each experiment and control. M analysis. The amino-acid nitrogen was determined by the Van Slyke thod, total nitrogen by the Kjeldahl process and ammonia nitrogen by distillation in vacuo. The details of analysis have been outlined i in a previous communication. Bacteriological control. Aerobic and anaerobic cultures and smears were made of all solutions. _ We wish to thank Dr. H. L. Amoss for his courtesy. te Kidney. 13s | 8 Efa| Bo te | eae og SR WES) Ria | ce cc. |deg.C.| mm a" solution: ! Before................+.+--| 10.0] 2.40| 22 | 750| 1.95 | O.0n8 After one w weak Soy GRRL a 3 110.0} 2.60) 20 750 1.46 0.015 Glycocoll solution: Berore..iovas...- ss ackn...| eee ad. 20). 22 759 9 .64 0.385 After one week............ | 2.5) 17.20) 20 750 9.65 0.386 Aspartic acid solution: | i Before.............++:+--..{ 5.0 | 18.00] 22 | 759] 10.18 | 0.208 After one week. 5.0 | 18.00, 20 750 | 10.10 0.202 sat rinse solution: Before.. os eee itles ss ck RO | 16.2 758 9.30 / 0.184 After one weeks.) SRR ose 9:0 OE 5.0 | 16.20) 21 760 9.18 | 0.184 Leucine solution: Befloreias... cove cebees oeeat O.0') 18 CO 8 | 10,24 0.205 After one week............ 5.0| 18.30, 21 | 760| 10.85 | 0.207 P. A. Levene and G. M. Meyer 557, Leucocytes. - eee. won : 48 P _ z B i ie B s | : N és PER ; aS 2 86 oe 4 : 100 ce Leucocytes and phosphate so- | lution: MR ore... /....| 2.600 20 sae | 0.117 0.0468 After one week............ 2b 0 20). ke 752 |} 0.117 0.0472 Asparagine solution: hoe oe ag 50 After one week............ 40 Leucine solution: SO 40) After one week............ .60 AMMONIA N PER CENT SOLUTION Before; After | Diff. Phosphate... 0 | 0,005) 0.005 Asparagine | 0.006 Ae 8) 0.008 i 0 | 0.005) 0.005 0 | 0.005) 0.005 q Leucocytes. ~ Phosphate. . 0 | 0.008 0.008) 0.047 0.047; 0.000 | 0.040 0.070 0.030 ' Asparagine | 0 | 0.012) 0.012) 0.199) 0.196} —0.003 | 0.168) 0.201 0.033 ' Leucine..... 0 | 0.010, 0.010, 0.267, 0.271) 0.004 | 0.096 0.131 0.035 ON “SUCRE VIRTUEL” AND BLOOD GLYCOLYSIS. By Pror. R. LEPINE (Lyons), Correspondent de |’ Académie des Sciences de Paris. (Received for publication, December 5, 1913.) In an interesting communication which recently appeared in this Journal, Macleod' states that according to Lépine and Bar- ral “the concentration of actual sugar may be greater in bl that has stood for from fifteen minutes to an hour at body tem- perature, outside the body than in freshly drawn blood,” and he refers to page 64 of my book on diabetes (P. 1909). Neither on that page nor elsewhere in any of my publications can be found the words “at body temperature” but exactly in the middle of page 64 there appears the following title: AUGMENTATION DU GLycosE DANS LE Sana, in Vitro, A 58°. As a matter of fact, it is necessary to inhibit glycolysis and in order to obtain this result a temperature of 58° is essential; but even this does not always prevent a great loss of sugar in the blood of certain dogs.* But leaving aside for the moment exceptional cases, it may be said that if arterial blood be permitted to flow simultaneously by means of a bifurcated canula into two tared flasks, one of which, A, contains a known quantity of a solution of mercuric nitrate, while the other, B, immersed in water at 58-59°, contains a weighed quantity of water sufficient to prevent coagulation of the blood; and then if after a quarter, one-half, or even one hour the contents of flask B be poured into a solution of mercuric nitrate and the sugar determined in A and B, it is generally found that the amount of sugar per 1000 grams of blood is greater in B than in A. This 1J. J. R. Macleod: this Journal, xv, p. 497, 1913. 2 Italics are mine. 8It is not easy to explain this exception. It is certain, as Barral and the writer showed in 1891, that the glycolytic enzyme nearly always loses its activity at 58°. It may be supposed that sometimes the enzyme which decomposes the actual sugar (see below) while continuing its diastatic ac- tion decomposes the sugar which is present in a nascent state. I advance this hypothesis with great reserve. 559 560 ‘Sucre Virtuel’’ and Blood Glycolysis slight increase in sugar, commonly observed in healthy and nor- mal dogs (in some instances, as mentioned above, a loss is ob- served), is much greater in dogs which have previously undergone an operation, particularly in such as have been bled.‘ The increase in sugar is still greater if a few hours previous to the bleeding the animal is injected subcutaneously or intrave- nously with a small quantity of pancreatin, invertin, phlorhizin, adrenaline, morphine, or antipyrine—in short with any substance which brings about a rapid modification in the quantity of “sucre virtuel.” More interesting than the increase of sugar in vitro is that which occurs in the circulation. I discovered in 1903, in collaboration with ud, that in a dog that had been fasting about fifteen hours, the d of the carotid (7.e., of the left ventricle), con- trary to the op of Cl. Bernard, very often contains more sugar than the bl the right ventricle (obtained by means of a sound introduceli™ the right jugular vein). Many of our numerous experiments have made it possible to state that this increase of sugar in the blood of the carotid takes place at the expense of the combined sugar.’ For — Blood of the right ventricle........ ae 0.90 Blood of the carotid .: 2.72... cc0ee eis 1,10 a oy, 0.75 165° 0.50 1.60 The increase of sugar in the blood of the renal vein in 7a . hizinized dogs discovered by Levene in 1895 (erroneously denied by Zuntz) has the same origin.’ Recently we have demonstrated * Macleod, having kept the blood at body temperature, naturally noted a loss of sugar which was due to normal glycolysis, In any case, if he had greatly increased the number of his experiments he might have found in some exceptional cases a slight gain. I have observed this in two or three ~ instances out of more than one thousand experiments. This can be ex- plained by assuming a considerable decomposition of “‘sucre virtuel’’ in an animal whose glycolytic power is weak. * Regarding the difficulties in estimating the amount of the combined sugar, see Lépine and Boulud: Journ. d. physiol. et d. path. gén., pp. 183- 184, 1911. *Lépine: Revue d. méd., 1913, p. 614, et seq.; Semaine méd., Sept. 24; Lépine and Boulud: Compt. rend. de Acad, des Sci., October 6, 1913, R. Lépine 561 that the cleavage of the combined sugar is produced by an en- zyme which can be extracted from the vascular wall.’ This lib- erated sugar deserves the name of “sucre virtuel’’ inasmuch as it is ready to be utilized as soon as it is liberated (apparently spon- taneously but in reality under the influence of an enzyme) from the combination in which it was present and in which combina- tion it could not be detected by the ordinary reagents. The enzyme concerned hydrolyzes phlorhizin. From this fact one is justified in drawing the conclusion that the “sucre virtuel’’ is of a glucosidic nature. In reference to glycolysis, Macleod might well have adde to his bibliography the article on glycolysis in the Dictionnaire s @ physiologie by Richet, and the chapter on glycolysis in my book on diabetes (pp. 152-191). According to Macleod the absence of glycolytic power of the serum was discovered 1and Déblin. This assertion is inaccurate because I carefully recorded this fact - with Barral (in a note at the Académie es Sciences, 1890) and this is furthermore pointed out by Levene and Meyer (this Jour- nal, xi, p. 364). With reference to the glycolytic activity of the leucocytes, which I was the first to discover, Van de Put might also be mentioned (Arch. internat. d. physiol. ,ix, p. 292,1910). Macleod studied glycolysis in vitro after having kept the blood for variable intervals at 40°. In any case, he should have made the estima- tions after one hour in order to have comparable results. He seems _ to deny the diminution of glycolysis in the blood of depancre- ' atized dogs. I wish to refer him to page 357 of the Journal de _ physiologie et de pathologie générale, 1911. There he will find an - explanation, at least a partial one, of the errors which may be committed in that respect (see also my book, p. 342). I cannot finish without referring to the important fact discov- ered by Levene and Meyer that the sugar of the blood can be regained after glycolysis. I have observed an analogous fact (see Journ. d. physiol. et d. path. gén., p. 184, 1911, note by Hugouneng and Morel). But this disturbing factor does not take place in the course of the first hour of glycolysis. Thus, no importance need be attached thereto, if the blood be left for only an hour in the incubator, as has always been my practice. 7 Lépine and Boulud: Compt. rend de l’ Acad. des, Sci., October 20, 1913. THE CHEMISTRY OF GLUCONEOGENESIS. VI. THE EFFECTS OF ACETALDEHYDE AND PROPYLALDEHYDE ON THE SUGAR FORMATION AND ACIDOSIS IN THE DIABETIC ORGANISM.' By A. I. RINGER anp E. M. FRANKEL. (From the Department of Physiological Chemistry of the University of Penn- sylvania, Philadelphia, Pa.) (Received for publication, December 5, 1913.) G It has long been recognized that aldehydes are capable of effect- ing a great many syntheses in the animal plant kingdoms. Thus, formaldehyde is generally accepted now to be the building stone from which sugars are synthesized in the plant kingdom? and Grube* demonstrated the possibility of this synthesis taking place in the liver of the turtig perfused with a fluid containing formaldehyde. The first suggestion that acetaldehyde may play a réle in the synthetic processes of the animal body was made by Spiro. He suggested the possibility of 8-hydroxybutyric acid arising from a condensation of two molecules of acetaldehyde, going through an aldol stage. CH; CHs CHs CH; | 9 oe CHOH CHOH > co H +O | —_ : - ra CHs O O O ye 0 of 4 a H H H H Acetaldehyde Aldol B-Hydroxybutyric Acetoacetic acid acid . Aided ‘= a eee Bis the Rockefeller Institute iol 1 Medical Haskaiel. 2 v. Bayer: Ber. d. deutsch. chem. Gesellsch., iii, p. 63, 1870. 3 Grube: Pfliger’s Archiv, cxxi, p. 636, 1908; CXxxxix, p. 428, 1911. ‘Spiro, quoted by Magnus-Levy: Arch. f. exp. Path. u. Pharm., xlii, p 225, 1899. 563 564 The Chemistry of Gluconeogenesis Friedman’ subjected these views to the test of experimentation and found that the perfusion of a dog’s liver with blood to which acetaldehyde had been added, was actually followed by an in- crease in the acetoacetic acid in the perfusion mixture. On test- ing the effects of aldol he likewise oh a very marked increase in the acetoacetic acid. We were led to the study of the fate of acetaldehyde in the dia- betic dog while searching for possible intermediary compounds in the metabolism of pyruvic acid. As was shown in a previous communication® pyruvic acid in different experiments does not ield glucose to the same degree. Several paths of catabolism of ic acid suggested themselves, and we hoped to be able to a, "show: that pyruvic acid passes in part through acetaldehyde when it gives 1 o only small quantities of glucose. The experiments, however, yielde d entirely unexpected and contrary results. The methods 9 in these experiments were the same as de- scribed in the previous communications of this series. Female dogs were used which were phlorhizinized by daily injection of 1 gram of phlorhizin ground up in : oil. The animals were catheterized at the end of each period of twelve hours, after which the bladder was washed three or four times with warm distilled water. The acetaldehyde used was prepared by Merck (deriva- tive of absolute alcohol) and was given subcutaneously diluted with water. In several instances the acetaldehyde was redis- tilled immediately before injection.? The nitrogen was deter-_ mined by Kjeldahl, glucose by Alihn, ammonia by Folin, acetone, — acetoacetic acid and 6-hydroxybutyric acid by Shaffer’s methods. The glucose determination was controlled by the polariscopic meth- od, and after aldehyde feedings was also controlled for its fer-— mentability by yeast. Effect of acetaldehyde, CHs—CHO., In experiment X XVII, period XIV, 8.8 grams (-*) of acetalde- _ hyde dissolved in 35 ec. of water were given subcutaneously. From * Friedman: Hofmeister's Beitrdge, xi, p. 202, 1908. * Ringer: This Journal, xv, p. 145, 1913. ? Attention must be called to the fact that acetaldehyde undergoes considerable deterioration on standing, which lessens its effects very mate- rially. A. I. Ringer and E. M. Frankel 565 the results of this experiment we see that the acetaldehyde exerts a very profound influence on the nitrogen as well as the glucose elimination. The nitrogen excretion of this dog which stayed above 5 grams per period for thirteen periods was reduced to 3.02 and 3.52 grams in periods XIV and XV to rise again to 6.60 grams in period XVI. The glucose elimination which was 18.75 grams in the foreperiod rose in the experimental period to 21.07 grams in spite of the very marked drop in the nitrogen excretion. The D :N ratio rose from its level of 3.65 in the foreperiod to 7.19 and 4.62 to come down again to 3.03 in period XVI. The amount of extra glucose eliminated in periods XIV and XV was 16.10 grams. In experiment XXVIII, period IX, 8.8 grams (-*.) of acetalde-— hyde dissolved in 25 cc. of water were given subcutaneously, Here as in the preceding experiment we note a depression in the nitro- gen and a rise in the glucose elithination. itrogen elimina- tion which was 6.42 grams in the foreperi eriod VIIT) was reduced to 5.03 and 4.95 grams in period LX anc X respectively to come up again to 6.18 and 6.23 in period XIand XII. The glu- cose elimination which was 22.39 grams in the foreperiod, rose to 29.70 in spite of the reduction in the nitrogen elimination. The D:N ratio rose from 3.48 to 5.91 and 4.89 to come down again in periods XI and XII to 3.54 and 3.40. The amount of extra glucose eliminated in periods IX and X was 18.9 grams. In tk iis experiment the effect of acetaldehyde on the acidosis was also studied. As is seen from the table, page 576, acetaldehyde possesses a very marked antiketogenetic effect. The amount of 8-hydroxybutyric acid elimination in the foreperiod was 2.35 grams. After the administration of acetaldehyde it was reduced to 1.42 and 0.50 gram in periods [X and X respectively, to come up again to 1.50 and 1.72 in periods XI and XII. Similar was the effect on the acetone and acetoacetic acid elimination. In period VIII (foreperiod) 630 mgm. of the two ketones were eliminated. Af- ter the acetaldehyde administration in periods [IX and X it was reduced to 360 and 180 mgm. to rise again in periods XI and XII to 380 and 470 mgm. There was also a reduction in the ammonia elimination following the acetaldehyde administration. This was relative as well as absolute. In experiment XXIX, period III, 8.8 grams (-*) of acetalde- hyde dissolved in 28 cc. of water were given subcutaneously. & diet 566 The Chemistry of Gluconeogenesis The results of this experiment corroborate our findings in the pre- ceding two experiments. ‘The nitrogen elimination which was 5.15 grams in the foreperiod (period II) was reduced to 3.27 and 2.77 in periods III and IV and rose again to 6.16 and 6.53 grams in periods V and VI. The glucose elimination, in spite of a very marked drop in the nitrogen output, rose from 17.59 in the fore- period to 24.85 grams in period III resulting in a D : N ratio of 7.61. The amount of extra glucose eliminated in periods IIT and IV was 20.45 grams. The effect of the acetaldehyde on acidosis was also studied in this case. Although distinct, it was, however, not as marked as in experiment XXVIII. The reason for this probably lies in the fact that the elimination of acetone bodies in this dog was very low to start with. In experiment XXX, period XII, 5.0 grams of acetaldehyde dis- solved in 20 ec. of water were given subcutaneously. Here too the elimination of the acetone bodies and of nitrogen was depressed very considerably. The rise in the glucose output, however, was very slight, resulting in only 3.2 grams of extra glucose. It is noteworthy that this dog had a very low (for phlorhizin gluco- suria), almost abnormal D:N ratio to start with, and we are in- clined to believe that the failure to yield more extra glucose may have some relationship to it. , In experiment XXX], period IT, 8.8 grams (-*-) of acetaldehyde dissolved in 30 cc. of water were given subcutaneously. There followed only a slight diminution in the nitrogen elimination, but _ a considerable rise in the glucose output. The amount of extra glucose eliminated in periods II and IIT was 10.7 grams. On examining the tabulated results of these experiments we note a number of very striking effects brought about by the acet- aldehyde. I. A very marked depression of the nitrogen elimination which lasts for about two periods (twenty-four hours) after the acetalde- hyde administration. II. A rise in the absolute amount of glucose eliminated during the period of acetaldehyde administration, in spite of the drop in the nitrogen, accompanied by a very high rise in the D : N ratio. III. A very marked depression in the acetone, acetoacetic acid and $-hydroxybutyrie acid eliminations where acidosis is high. As will be seen from a subsequent communication the admin- A. I. Ringer and E. M. Frankel 567 istration to diabetic dogs of ethyl alcohol and acetic acid which stand in such very close chemical relationship to acetaldehyde are not followed by any of the effects enumerated above. This sug- O gested the possibility that the aldehyde radical ct may be responsible for the effects brought about by the acetaldehyde. We, therefore, decided to study the effects of its homologues. Effect of propylaldehyde, CHs—CH2— CHO. In experiment XX VII, period XVII, 11.6 grams (-¥-) of Kahl- baum’s propylaldehyde dissolved in 30 ec. of water to which 3 ec. of ethyl alcohol had been added, were given subcutaneously. The results were very similar to those obtained with adetaldehyde. There followed a very marked depression in - oe elimina- tion and a rise in the D:N ratio. The amount of extra glucose eliminated in periods XVII and XVIII was 11.65 grams. Much more convincing results were obtained in experiment XXVIII. In period XIII of this experiment 11.6 grams (-*-) of propylaldehyde were administered as above. The nitrogen elim- ination which stood at the level of 6.18 and 6.23 in periods XT and XII dropped down to 3.27 and 4.83, to rise again to 6.24 grams in period XV. ‘The glucose elimination in period XIII rose to 24.02 grams in spite of the very marked reduction in the nitrogen elim- imation, resulting in a rise in the D:N ratio to 7.35. The amount of extra glucose eliminated was 19.09 grams. The effect of prophylaldehyde on the ammonia, acetone, aceto- acetic acid and 6-hydroxybutyric acid eliminations was as marked as that of the acetaldehyde. All were depressed very consider- ably. The ammonia nitrogen elimination which was 0.65 and 0.70 gram in periods XI and XII was reduced to 0.25 gram in period XIII. The reduetion of the ammonia nitrogen from 10.5 and 11.2 per cent to 7.6 per cent of the total nitrogen is also note- worthy. The acetone and acetoacetic acid elimination was re- duced from 380 mgm. in period XI and 470 mgm. in period XII to 120 mgm. in periods XIII and XIV. The f-hydroxybutyric acid elimination which was 1.50 and 1.72 grams in the two fore- periods was reduced to 0.36 and 0.25 gram in periods XIIT and XIV. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 Tm os ~ We have attempted several experiments to study the effect of 568 The Chemistry of Gluconeogenesis In experiment XXXII, period VI, 11.6 grams of propylaldehyde were administered as above. The results are very similar to those obtained in the preceding experiment. The glucose elimination and the D:N ratio rose very considerably, yielding 19.75 grams of extra glucose. The reduction in the acetone bodies was very marked. The acetone and acetoacetic acid having come down from 640 mgm. to a little above 200 mgm. while the 8-hydroxybutyric acid was reduced from 1.87 grams in period V to 0.98 gram and 0.378 gram in periods VI and VII respectively. Effect of formaldehyde, H—CHO. formaldehy, e on the diabetic dog, but all have failed so far be- cause of the toxicity of the aldehyde. The animal usually dies within twelve hours. We hope, however, to come back to these. experiments in the near future by administering the aldehyde in very small quantities at a time. Perhaps this may yield satis- factory results. | Discussion of results. From a review of the preceding pages we note a very remark- able reaction which apparently seems to be characteristic of sub- stances possessing an aldehyde radical, for neither the alcohols, nor the acids corresponding to the aldehydes studied possess the power of effecting such deep seated changes in the metabolism of — the diabetic animal. It is true that propyl alcohol and propionic — acid possess the power of glucogenesis but they do not affect the nitrogen metabolism nor the acidosis, to the extent that propylalde- hyde does. The effect of the acetaldehyde is much more remark- able, because neither the alcohol nor the acid that corresponds to it has any appreciable influence on the metabolism of the diabetic dog. : The outcome of our experiments leads to conclusions diamet- rically opposed to those of Friedman.’ As was stated above he found that acetaldehyde on perfusion through the surviving liver gives rise to acetoacetic acid. On giving acetaldehyde stbeu- taneously, however, we found that it has just the opposite effect. How can we explain these differences? * Friedman, loc. cit. A. I. Ringer and E. M. Frankel 569 It. is possible that in perfusion of the surviving extirpated liver one may deal with a metabolism that is decidedly abnormal, and which does not correspond with the process in the same organ when in normal condition. Under such circumstances, wé can readily understand, why a substance should follow one path of metabolism in one case and an entirely different one in the other. It must also be borne in mind in this connection that the func- tions of the liver in the animal body may constantly be influenced by the other organs or by products of their metabolism. An illus- tration of such influence can be found in the works of Levene and Meyer® who showed how essential the codperation of the different organs is for carbohydrate metabolism, and also to what erro- neous and misleading conclusions one may be drawn by, studying the influences of individual organs on the processes of'metabolism, without taking into consideration the possil influence of the other organs upon the one studied. a But most of the results obtained by mbden Tad his collabora- tors on perfused surviving livers have been found in a general way to be identical with those obtained in feeding experiments. Is it not possible, therefore, that in Friedman’s experiments another factor played a réle which is of great moment in determining the path of the metabolism of acetaldehyde? Is it not possible, for example, that the presence or absence of glycogen in the liver may have influenced his results? Surely, this is a matter which re- quires further study, especially since it was shown by Friedman” _ that the perfusion of sodium acetate through a liver poor in glyco- gen will cause an increase in the acetoacetic acid formation whereas the perfusion through a liver rich in glycogen will be followed by negative results. The glucogenetic effect of acetaldehyde and propylaldehyde. One of the most remarkable phenomena in our experiments is the very large amount of “extra” glucose that was eliminated after the administration of the acetaldehyde. If all the carbon of the acetaldehyde molecule were converted into glucose the administered 8.8 grams would give rise to 12 grams of glucose. ® Levene and Meyer: This Journal, ix, p. 97}; xi, pp. 347, 353, 361;, xii, p. 265. 10 Friedman: Biochem. Zeiischr., lv, p. 436, 1913. vi? lil ictal 570 The Chemistry of Gluconeogenesis CH; | Be + 80 — CHO; COH 132 -+ 48 = 180 The amount of “‘extra’’ glucose found in experiment X X VII was 16.10 grams, in experiment XXVIII, 18.9 grams, in experiment XXIX, 20.45 grams and in experiment XXXIJ, 10.7 grams. From this we see very clearly that considerably more glucose was eliminated in the urine after acetaldehyde administration, than can be accounted for by a complete conversion of the acet- ehyde into glucose. In other words, some substance or sub- ces that are ordinarily non-glucogenetic have contributed to the formation of glucose. The conclusion, therefore, seems justi- fied that. dehyde possesses the power of converting some sub- stance in the animal metabolism that is non-glucogenetic to one that is glucogenetic and ‘that the substance so formed possesses a greater number of carbon atoms than does acetaldehyde. Whether acetalde- hyde itself takes part in the yglucogenetic process will be discussed later. What is true for acetaldehyde i is true for propylaldehyde, but to a lesser extent. If all the arbon of the propylaldehyde were converted into glucose 11.6 grams of propylaldehyde could yield 18 grams of glucose. CHs = * t ue 2 CHe +, 40 - CgH205 COH 116 + 64 = 180 The amount of “extra” glucose obtained in experiment XXVII was 11.65 grams, in experiment XXVIII, 19.10 grams and in ex- periment XXXII, 19.75 grams. The ‘‘extra’’ glucose in experi- ment XXVII is lower than in any of the others and may be ac- counted for by the fact that it was performed on a dog that had had glucosuria for seventeen experimental periods, outside of the preparatory periods. At this time the animals find themselves in a very low state of vitality and this may account for the dif- ference. The results obtained at this stage are usually taken for corroborative purposes only, eo A. I. Ringer and E. M. Frankel 571 In experiments XXVIII and XXXII there is clearly a greater amount of glucose eliminated than can be accounted for by a com- plete conversion of the propylaldehyde into glucose, and similar to the acetaldehyde it seems to possess the power of converting non-glucogenetic substances into glucogenetic ones. The question arises now, what is the nature of this change? In what way may the acetaldehyde or propylaldehyde exert its influence upon non-glucogenetic substances? Does the acetal- dehyde or propylaldehyde bring about its effects by modifying the normal path of metabolism of those substances, or does it enter with them into a chemical union, thus changing their struc- tural configuration and thereby modifying their path of cola olism? When we come to examine the effects of the aldehydes on the diabetic organism as a whole and associate ee phenom- ena, a theory suggests itself which seems to harmo all the facts. Our experiments bring to light three important facts: I. That the administration of aldehydes in diabetic animals is followed by a very marked rise in the glucose elimination. II. That concomitant with this phenomenon there is a consid- erable drop in the elimination of acetone bodies. III. That aleohols and acids related to the aldehydes do not possess these effects. From the above it becomes evident that the aldehyde radical , O - c is the determining factor in bringing about the effects described. In a subsequent communication additional evidence will be presented in support of this view. The aldehyde radicals possess great combining powers and it is generally recognized now, in what a complexity and multiplicity of unions the alde- hydes are capable of entering. In the disaccharides, glucosides, and glucoronates and a great many other compounds this fact is evident. When we bear this fact in mind and realize that the acetalde- hyde and propylaldehyde in our diabetic animals have brought about a diminution in the acetone bodies on the one hand and an increase of the glucose elimination on the other, it seems reason- able to assume that the two phenomena may be causally related, cad 572 The Chemistry of Gluconeogenesis i.e., the aldehydes, because of their great combining power, may have the property of combining with the secondary alcohol radi- cal of 6-hydroxybutyric acid, and by changing its structural con- figuration convert it into a substance that is glucogenetic. An — illustration of the possibility is given in the following reaction, CHs = CH, | a H COH co CH; — CHOH | + OH 2 — CH;— - ; CHe 8. a CH. COOH B Hy droxytiityrc Acetaldehyde 8-Methyllevulinic acid | " acid which results in the formation of 6-methyllevulinic acid, 7.e., the conversion of a normal fatty acid into an iso compound. It has been shown by Baer and Blum, Embden and his collaborators, and by Ringer, Frankel and Jonas the iso compounds in the animal body undergo demethylation. $-Methyllevulinic acid would therefore be converted into levulinic acid, which, as will be shown in a subsequent communication, does poste glucogenetic properties. It must be realized that between the 6-hydroxybutyric acid and acetaldehyde combination and levulinic acid there may be a num-— ber of possible intermediary compounds of the keto and enol forms — which will be discussed elsewhere. At present we wish only to sketch our conception of the possible reaction and indicate the possibility of the conversion of a compound with an even number of carbon atoms as 8-hydroxybutyric acid to one with an uneven number of carbon atoms, the conversion of an acetone-genetic compound to one that is glucogenetic. Objection may be raised to this theory because the increase of glucose elimination in our experiment is much greater than is the " Baer and Blum: Arch. f. ewp, Path. u. Pharm., lv, p. 89, 1906; Embden, Salomon and Schmidt: Hofmeister’s Beitrdge, viii, p. 129; Ringer, Frankel and Jonas: This Journal, xiv, p. 525, 1913. A. I. Ringer and E. M. Frankel 573 drop in the acetone elimination. But as will be shown in a sub- sequent communication, this objection is not valid. For, the acetone bodies that are eliminated in the urine are only a small fraction of that which actually plays a part in the intermediary metabolism, and any change that is evident in the urine may be greatly magnified in the intermediary metabolism. The path of levulinic acid in metabolism is the subject of our present inquiry. It may be one of the following; it may undergo B-oxidation giving rise to pyruvic acid, or it may undergo further CO CO CO. — GLUCOSE CHe — CHOH — COOH COOH COOH COOH Levulinie $-Hydroxy levu- i acid linie acid NY CH; CH: * — GLUCOSE oxidation in the y-carbon, breaking up into acetic acid and pro- pionic acid. In either case we get a three-carbon compound which is glucogenetic. Judging from the amount of glucose we obtain after acetaldehyde administration we are inclined to believe that propionic acid is the final product. In the case of the propylaldehyde we may conceive of the following reactions: Sd —_ chibi (A ey, pal ga 574 The Chemistry of Gluconeogenesis CH, oe, CH, br, bu, but, Ban | a to CH,;-CHOH + _ ca > he CH, | CH, tia ee | COOH ners B-Hydroxybutyric Propylal- B-Methyl propion- 8-Propionyl- acid dehyde yl-propionie acid propionic acid These experiments have Spal a ‘a new field fo tation which, in our estimation, throws a great deal ¢ mechanism of antiketogenesis. They also bring up the que sti of the influence of the higher aldehydes on metabolism. Most of these experiments are completed and will be published in the near future. The effect of the aldehydes on the protein metabolism ’_ will be discussed in a separate communication. : 575 BY = Serpe xix | & S —_uaai3 ‘poppe wooq puy joyooye [Aye Jo °99 ¢ YT Scag | © TAX | &% thie sStreik J0 +00 gg ts paaqouerp enkigpreédoid ured y | wot | WE AX | 2% s ‘¢ | w'6r | 099 IAX ra : “A[snoouvjnoqns uaAtz ‘y | OT z's | AX IZ {x} toJWM JO “90 OE UT padjosstp opAYap]ezeo" surIBIZ g*g ‘. | ore | we | AIX IZ a . ‘¢ | osst | ws | OL | Ix | 02 I ‘¢ | SG ZT | 909 IIx | of : zs | O8OL | IX | OI © 4) oe | sro eo | 6 q cose | wo | Sel | xr | 8st a | STO | THA | St * whevnies ; a aia —— | LHOIGA | aoraa tet XY af ft ‘sporsad snoy-aapny, “JIAXX LNANTIGaXyY 2 pea 5 a. it ae The Chemistry of Gluconeogenesis 576 *A[Ssnoouey -noqns udAts ‘sAoqe se joy -O9]8 pUw 107BAL Ul PpeA]|OSSIp | epkyopyetséd | -o1d Jo sures 9° [1 | | “AT -snoouryns qns UOAIS I04BM JO "09 GZ UL PaAlos -Stp oepAyopye -jo0v SUIRIZ Q'S | SwUVNaY aiov OMMALOR IOV OLLGOV |-OLS0V NV 9L Lo N 1¥a04 AXOUGAH-g) ANOLMOV | :N-HN & “sporsad “ gnoy-aayam J, iti AXX. LNANIUAANG <4 . aso0o0n 1D VULXa SRRSS oD 0D OD OD OD N'd 0S' ST 6E° 2 OE ZZ SE FG OS € 96° 92 asoon'19 ‘1VLOL i N@OOULIX | IVLOL cr Or eo 01 LHOISA aomad xe Si6t “AVR aava A. I. Ringer and E. M. Frankel 577 T I 6 LH K al LE [EX ThA Z j |X | i A 4 je i aie! ct le = Ay F 3 LA s s 29 ; \ i “ \, * Be. ? eT ai \ ra ac NG “N / N A 5 4 oo PR ae as 5 : “oa \ \ E ~ - _4 ‘ aa > ‘ 1. J Mi ysis ‘ i. : NS AN [ad os ; og wage | \ — NITROGEN \ / 5* ---- GLUCOSE —— D:N Ratio y 3 eae * | 5 ie — S$ £ on 8 8, 8) 1t “4 4 lo =I PrriopDs. ExPERIMENT XXVIII. The Chemistry of Gluconeogenesis 578 “A[snoouey -Noqns UdAIS 19}VM Jo ‘00 QZ UI ee 0 a peAajossip opAyopyeyooe surviZ Q'¢ re'8 Ix II % It 9¢°8 XI or IIIA or ss°8 IIA 6 SXaVAEE aHom | aonaa = “9 89° OT IA ¢ 9 A ¢ ‘A[snoou ; : ; -B{NOqNS UOAIT 10jVM Jo “00 gz 209 06 &T es AI : UI poAlossip opAYopyezoow survid g*g TOL 9 8s Ill o's 6F2T STs 00°11 II £ I £ sxUVWaY N:a | STeton | aeaon | smOtmm | comma | StGl Sune 3 eed noy-ayan yt, “XIXX LNAWINSAXG 579 cos ° Z0Z 0 rE € 61 6z cL'8 XI 61 ( 91z2°0 06°€ 96° €E os cI IIIA 61 “Ajsnooueynoqns i — udAld oAoge sev joyoore | .e¢t+ | seo : ; ar9 onze z2E9 IIA 8ST v Lt per, , QL 6T +4 pu JojeM Ul poAjossIp | .€9T+ | 860 | 88B =we 97s WL MFI IA SI a epAyopyeyAdoid suread 9° [] \ 5 ny o6t' I+ | Z8°T whe | 62 | 082 | A LI Be o90' T+ | e242 96° go'se | StL | #'ST | AI vA . | op! asoon1e | NESOULIN e16r tx} SUUVINGAY esti Banter . N:a avaaia aves | LHDIG9M | aoludd iors = “sporsad unoy-aapon, “TIXXX INGNIYTAX tas) : tee 3) on 1 & ce 0 €1 0 08" 2 AI 91 Ns A “AJsnoouv} 3 ‘ ; : i + -hoqns Waals J0zWM JO ‘99 OG UI pide rth LOT pep 92 aI 5 Ff bog ; , Padjosstp opAyopyezooe sured gg < sro | 90°0 16°¢ I tI amy jay o1aov| __ : ‘ ; SxUVNGH Laoag ouiov axv eee ae ‘a = mepossiN | anommm | coraza | ®6l NM “*sporiad unoy-2aan,, "TXXX INDEX TO VOLUME XVI. Absorption, changes in fat during, 517; of amino-acids from the blood by the tissues, 197; of cholesterol from digestive tract of rabbits, 495. Acetaldehyde, effects of on sugar formation and acidosis in the diabetic organism, 563. Acetoacetie acid formation in liver, influence of pancreas on, 515. Acetole, effect of on animal organ- ism, 455. Acetone, determination of, 281, 289. Acidosis in the diabetic organism, effects of acetaldehyde and pro- pylaldehyde on, 563. Acids, free mineral, effect of on en- dogenous nitrogen metabolism, 299. : d-Alanine, separation of from d- valine, 103. ; Alcohols, polyatomic, as sources of carbon for lower fungi, ¥. Aliphatic amino nitrogen, determi- nation of, 121. Alkali soils, influence of salts of upon growth of rice plant, 235. Amino-acid nitrogen, determination of in urine, 125, 385. Amino-acids, absorbed, locus of chemical transformation of, 213; absorption of from blood by tis- sues, 197; action of leucocytes and of kidney tissue on, 555; in tissues, effect on of feeding and fasting, 231. Amino groups, free, nature of in proteins, 539; —— nitrogen, ali- phatic, determination of, 121; —— nitrogen in tissues, deter- mination of, 187. Ammonium carbonate, formation of urea from in the liver, 399. Aspergillis niger, influence of exhaus- tion of the medium upon rate of autolysis of, 479. Autolysis of mold cultures, 479. Bacterial changes in milk and cream at 0°C., 331. BauMANN, Emi J. Baumann, 135. Benepict, 8. R. and J. R. Murwin: Note o cai determination of ein nitrogen in urine, penaale acid, liience of on endog- enous nitrogen metabolism, 321. BircHarD, Freperick J.: see Van Slyke and Birchard, 539. Blood, absorption of amino-acids from by tissues, 197; determina- tion of 8-oxybutyric acid in, 293; gaseous content of after clamp- ing abdominal vessels, 79; gly- colysis in, 559; of fish, 389. Bioor, W. R.: On fat absorption. III. Changes in fat during ab- sorption, 517. Branchial cleft organs, iodine con- tent of, 465. Burrewi, J. I.: see Pennington, Hepburn, St. John, Witmer, Stafford and Burrell, 331. Butter-fat, influence of on growth, 423. : see Ji ohns and Cameron, A. T.: The iodine con- tent of the thyroid and of some branchial cleft organs, 465. Carbohydrates in diet, influence of on rate of nitrogen elimination, , 37. 581 ' 582 rae * Carbon dioxide, and oxygen con- tent of blood after clamping ab- dominal vessels, 79; apparatus for estimation of minute quan- tities of, 485. Casein, hexone bases of, 531. Cholesterol, rate of absorption of, 495. Cream, bacterial and enzymic changes in at 0°C., 331. Creatine and creatinine, administra- tion of, influence of on creatine content of muscle, 169. aun, H. D. and H. W. Duptey: Glyoxalase. Part IV, 505; Some negative experiments on the in- we pancreas upon acetoacet acid formation in the liver, 515. ~ Denis, W.: Metabolism studies on cold-blooded animals. II. The blood and urine of fish, 389; Note on the tolerance shown by elasmobranch fish towards cer- tain nephrotoxic agents, 395. Determination, of acetone, 281, 289; of aliphatic amino nitrogen in minute quantities, 121; of ami- no-acid nitrogen in urine, 125, 385; of amino nitrogen in tis- sues, 187; of B-oxybutyrie acid, 265, 293. Diabetes, effects of acetaldehyde and propylaldehyde on sugar forma- tion and acidosis in, 563; forma- tion of glucose from propionic acid in, 375; theory of, 455. Diet, influence of carbohydrates and fats of on rate of nitrogen elimination, 37; influence of character of protein of on rate of nitrogen elimination, 55; in- fluence of texture of on rate of nitrogen elimination, 19. Dihydrosphingosine, oxidation of, 549, Index 2,6-Dioxy-3,4-dimethyl -5-nitropyri- midine (@-dimethylnitrouracil), 135. 2,8-Dioxy-1,6-dimethylpurine, 135. Dox, ArTHUR W.: Autolysis of mold cultures. II. Influence of ex- haustion of the medium upon the rate of autolysis of Asperg- illis niger, 479. _ Duptey, H. W.: see Dakin and Dud- ley, 505, 515. EDELMANN, Lzo: see Murlin, Edel- mann and Kramer, 79. Elimination of nitrogen, influence of carbohydrates and fats on rate of, 37; influence of character of ingested proteins on rate of, 55; influence of texture of diet on rate of, 19. Enzymiec changes in milk and cream at 0°C., 331. Esterase, purification of, 1; and sodi- um fluoride, compound between, mation of minute quantities of carbon dioxide, % Fast ect of on amino-acid, con- tent of tissues, 231. ie 517; feeding, influence of on en- Fat, changes in during absorption, 4 dogenolll nitrogen metabolism, — ve 317. Fats in diet, influence of on rate of nitrogen elimination, 37. Fatty acid, saturated, of kephalin, 419. Fatty acids, method for converting into their lower homologues, 475. Feeding, effect of on amino-acid con- tent of tissues, 231. Ferments, effect of on growth of to- bacco, 439. Ferry, Epna L.: Mendel, 423 see Osborne and 7 Fine, Morris 8.: see Myers and Fine, 169. Fish, blood and urine of, 389; elas- mobranch, tolerance towards ne- phrotoxic agents, 395. Fisxp, Cyrus H. and Howarp T. Karsner: Urea formation in the liver. Astudy of the urea-form- ing function by perfusion with fluids containing (a2) ammonium carbonate and (b) glycocoll, 399. Folin and Denis, uric acid and phe- nol reagents of, 369. FranKet, E. M.: see Ringer and Frankel, 563. Fungi, polyatomic alcohols as sources of carbon for, 143. onie acid in diabetes mellitus, 375. Glycid, effect of on animal organ- ism, 455. Glycocoll, non-formation of urea from in liver perfusion, 399. Glycolysis, blood, 559. -Glyoxalase, 505. GREENWALD, Istpor: The forma- tion of glucose from propionic acid in diabetes mellitus, 375. Greer, J. R., E. J. Wirzemann and a R. T. Woopyarr: Studies on the theory of diabetes. II. Gly- { eg cid and acetole in the normal and phlorhizinized animal, 455. Growth, influence of butter-fat on, 423; of rice plant, influence of salts on, 235; of tobacco, effect of ferments and other sub- stances on, 439. Hepsurn, J. S.: see Pennington, Hepburn, St. John, Witmer, Stafford and Burrell, 331. Hexone bases of casein, 531. ie eee Index Gluconeogenesis, chemistry of, 563. Glucose, formation of from propi-. 583 Hoaatanp, D. R.: see McCollum and Hoagland, 299, 317, 321. Hydantoin derivatives, reactions of with uric acid and phenol rea- gents of Folin and Denis, 369. Todine content of thyroid and bran- chial cleft organs, 465. Jouns, Cart O. and Emin J. Bav- MANN: Researches on purines. XIII. On 2,8-dioxy-1,6-dimethyl- purine, and 2,6-dioxy-3,4-di- methyl-5-nitropyrimidine (a-di- methylnitrouracil), 135. Karsner, Howarp T.: see Fiske and _ Kephalin, satura fatty acid of, 419. eS Kidney tissue, action of on amino- acids, 555. Kramer, B.: see Murlin, Edelmann and Kramer, 79. Leaman, Epwin P.: On the rate of absorption of cholesterol from the digestive tract of rabbits, 495. L&éprne, R.: On “‘sucre virtuel’’ and blood glycolysis, 559. Leucocytes, action of on amino- acids, 555. Levens, P. A. and G. M. Meyer: On the action of leucocytes and of kidney tissue on amino-acids, 555; ——and Donatp D. VAN Styxe: The separation of d-ala- nine and d-valine, 103;——and C.J. West: Thesaturated fatty acid of kephalin, 419; A general method for the conversion of fatty acids into their lower hom- ologues, 475; On sphingosine. II. The oxidation of sphingosine and dihydrosphingosine, 549. THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XVI, NO. 4 teed 584 Lewis, Howarp B. and Ben H. NicoLet: The reaction of some purine, pyrimidine and hydan- toin derivatives with the uric acid and phenol reagents of Folin and Denis, 369. Lewis, Rosert C.: see Mendel and Lewis, 19, 37, 55. Lipase, action of radium emanation on, 379. Liver, esterase from, 1; influence of pancreas upon acetoacetic acid formation in, 515; urea-forming function of, 399. tr, W. M.: The determi- nm of acetone, 281; Nephel- ic de’ ermination ot wanes Marsuatt, E. K., ry nnd Ts ¢ ROWNTREE: The action - dium emanation on lipase, 370. Marsu, Howarp L.: see Meigs and Marsh, 147. McCotivm, E. V. and D. R. Hoaa- LAND: Studies of the endoge- nous metabolism of the pig as modified by various factors. I. The effects of acid and basic salts, and of free mineral acids on the endogenous nitrogen metabolism, 299; Studies of the endogenous metabolism of the pig as modified by various fac- tors. IL. The influence of fat feeding on endogenous nitrogen metabolism, 317; Studies of the endogenous metabolism of the pig as modified by various fac- tors. III. The influence of ben- zoio acid on the endogenous nitrogen metabolism, $21, Meras, Epwarp B. and Howarp L. Manrsu: The comparative com- Index position of human milk and of cow’s milk, 147. MenpeEL, Larayette B. and Ros- ERT C. Lewis: The rate of elimination of nitrogen as influ- enced by diet factors. I. The influence of the texture of the diet, 19; The rate of elimina- tion of nitrogen as influenced by diet factors. II. The influence of carbohydrates and fats in the diet, 37; The rate of elimina- tion of nitrogen as influenced by diet factors. III. The influ- ence of the character of the ingested protein, 55; see also Osborne and Mendel, 423. Metabolism, after clamping abdom- inal vessels, 79; endogenous ni- trogen, effect of acid and basic salts and of free mineral acids on, 299; endogenous nitrogen, effect of benzoic acid on, 321; endogenous nitrogen, effect of fat feeding on, 317; of cold- blooded animals, 389. ethod, for converting fatty acids Santo their lower ee 475. Meyer, G. M.: see Lev ne and Mey- - er, 555; see also Van Slyke and Meyer, 197, 213, 281.) Milk, bacterial and enzymic ohaiie »: in at O°C., 331; comparativ composition of, 147. al Mi.ts, 8. Roy: see Rosenbloom a Mills, 327 Miyake, K.: The influence of salts — common in alkali soils upon the growth of the rice plant, 235. ; Mold, autolysis of, 479. i Morphine tests, non-interference of “ptomaines’’ with, 327. Murutn, J. R., Leo EpBLMANN and B. Kramer: The carbon dioxide and oxygen content of the blood after clamping the abdominal ‘aorta and inferior vena cava be- Index low the diaphragm, 79; see also Benedict and Murlin, 385. Muscle, influence of administration of creatine and creatinine on creatine content of, 169. Myers, Vicror C. and Morris 8. Fine: The influence of the ad- ministration of creatine and cre- atinine on the creatine content of muscle, 169. Nerpie, Ray E.: Polyatomic alco- hols as sources of carbon for lower fungi, 143. Nephrotoxic agents, tolerance of elasmobranch fish for, 395. NicoLet, Ben H.: see Lewis and Nicolet, 369. Nitrogen, aliphatic amino, deter- mination of, 121; amino-acid, - determination of in urine, 125, 385; amino, determination of in tissues, 187 ; —— elimination, in- acter of ingested protein on rate of, 55; influence of texture of diet on rate of, 19; ——metabo- lism, effect of asia and basic salt and of free mineral acids on, 299; effect of fat feeding on, s 317; effect of benzoic acid on,321. J. Du P. and O. M. Suepp: The effect of ferments and other substances on the growth of Burley tobacco, 439. OsporNne, THomas B. and Laray- ETTE B. Menpev: The influ- ence of butter-fat on growth, 423. B-Oxybutyric acid, determination of, 265, 293. Oxygen and carbon dioxide con- tent of blood after clamping abdominal vessels, 79. fluence of carbohydrates and fats — a on rate of, 37; influence of char-— inn ~~ eee rg, agen 585 Pancreas, influence of on acetoacetic acid formation in liver, 515. Perrce,GeorGce: The partial puri- fication of the esterase in pig’s liver, 1; The compound formed between esterase and sodium fluoride, 5 PENNINGTON, M. E., J.S. PL E. Q. Sr. Joun, E. Wirmer, M. O. Starrorp and J. I. BURRELL: Bacterial and enzymic changes in milk and cream at 0°C., 331. Phenol reagents of Folin and Denis, reactions of, with purine, pyti- midine and hydantoin d iva- tives, 369. ‘ Phlorhizinized — t of gly- cid and acetole on, 455. Pig, endoge tabolism of, 299, 317, 32 Propionie @ tion of glucose in diabetes mellitus, 375. aldehyde, effects of on sugar formation and acidosis in the diabetic organism, 563. Protein, fate of digestion products of, 187, 197, 213, 231; influence of character of on rate of nitrogen elimination, 55. Proteins, nature of free amino groups in, 539. ‘*Ptomaines,’’ non-interference of with morphine tests, 327. Purine derivatives, reactions of with uric acid and phenol reagents of Folin and Denis, 369. Purines, researches on, 135. Pyrimidine derivatives, reactions of with uric acid and phenol rea- gents of Folin and Denis, 369. Rabbits, rate of absorption of chol- esterol by, 495. Radium emanation, action of on lipase, 379. P : = 586 Rice plant, influence of salts on growth of, 235. ; Rincer, A. I. and E. M. FRANKEL: The chemistry of gluconeogene- sis. VI. The effects of acetal- dehyde and propylaldehyde on sugar formation and acidosis in the diabetic organism, 563. RosENBLOoM, JAacop and S. Roy Mitts: The non-interference of ‘‘ptomaines,’’ with certain tests for morphine, 327. Rowntree, L. G.: see Marshall and Rowntree, 379. Salts, acid and basic, effects of on endogenous nitrogen metabo- lism, 299; of alkali soils, influ- ence of upon growth of rice plant, 235. Suarrer, Paruie A. and W. McKim Marriorr: The determination of oxybutyric acid, 265. Suepp, O. M.: see Oosthuizen and Shedd, 439. Sodium fluoride and esterase, com- pound between, 5. Sphingosine, oxidation of, 549. Srarrorp, M. O.: see Pennington, Hepburn, St. John, Witmer, Staf- ford and Burrell, 331. Sr. Joun, E. Q.: See Pennington, Hepburn, St. John, Witmer, Staf- ford and Burrell, 331. “Sucre virtuel’’ and blood glycoly- sis, 559. Sugar formation in diabetes, effects of acetaldehyde and propylalde- hyde on, 563. Tasuino, Surro: Carbon dioxide ap- paratus III. Another special apparatus for the estimation of very minute quantities of carbon dioxide, 485. Texture of diet, influence of on rate of nitrogen elimination, 19. Index Thyroid, iodine content of, 465. Tissues, absorption of amino-acids by from blood, 197; action of on amino-acids, 555 ; amino-acids in, effects of feeding and fasting on, 231; determination of amino ni- trogen in, 187; determination of 8-oxybutyrie acid in, 293. Tobacco, effect of ferments and other substances on growth of, 439. Urea formation in liver, 399. Uric acid reagent of Folin and Den- is, reactions of, with purine, pyrimidine and hydantoin de- rivatives, 369. Urine, determination of amino-acid nitrogen in, 125, 385; of fish, * 389. d-Valine, separation of from d-ala- nine, 103. Van Stryke, Donatp D.: The gas- ometric determination of ali- phatic amino nitrogen in minute quantities, 121; Improved meth- ods in the gadoretric deter- mination of free and conjugated amino-acid nitrogen in the urine, 125; The fate of protein diges- tion products in the body. II. Determination of amino nitro-— gen in the tissues, 187; The hex- one bases of casein, 531; —— and Freperick J. BrrceHarp: The nature of the free amino groups in proteins, 5389; ——-and Gustave M. Meyer: ‘The fate of protein digestion products in the body. III. The absorp- tion of amino-acids from the blood by the tissues, 197; The fate of protein digestion prod- ucts in the body, IV. The locus of chemical transforma- tion of absorbed amino-acids, | i ee Index ‘587 213; The fate of protein diges- tion products in the body. V. The effects of feeding and fast- ing on the amino-acid content of the tissues, 231; see also Le- vene and Van Slyke, 103. WAKEMAN, ALFRED J.: see Osborne and Mendel, 423. West, C. J.: see Levene and West, 419, 475, 549, Wirmer, E.: see Pennington, Hep- burn, St. John, Witmer, Staf- ford and Burrell, 331. Wirzemann, E. J.: see Greer, Witze- mann and Woodyatt, 455. Woopyarr, R. T.: see Greer, Witze- mann and Woodyatt, 455. * nub otf Ow owee QP The Journal of biological 501 chemistry ~ Wal6 cop.2 Biological & Medical | Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY ay k 1 oa? ei: orgs ey ratte i “ oe ; - Wales >, ft ae wt a 7 ; ¥ Ey - {ik Writ) Teeth ix Af pent , f $y Z 1s BAS se w ae wis Pr ey nat ‘ [ =~ = eat quiet He)