oe eteae® <9 Sma on ee areas RRS ae ee ~<» 5° ee Aas oe epg ms 9 a ee Nine a ea pn ak ai, Call ete ase ee el ae Le eee - 5445, . ae ial } hin Tae ee Le ‘Oty be a vy | ay on io ms \ Dee he hy aN | | Ma ; ie { n lie iA ih | PR Ee 40) TAs ee La Ta a es ae Ni dale ( aL, AA, Lesser) ata te fa) | i i i ah y a) n me i ee Pas f Na Higy.) Pn hee \ " | A " MY ” rie ie a Fil air : a THE AMERICAN JOURNAL OF PHYSIOLOGY. EDITED FOR The American Physiological Society BY H. P. BOWDITCH, M.D., BOSTON FREDERIC S. LEE, Pu.D., NEW YORK R. H. CHITTENDEN, PH.D., NEW HAVEN W. P. LOMBARD, M.D., ANN ARBOR W. H. HOWELL, M.D., BALTIMORE G. N. STEWART, M.D., CHICAGO W. T. PORTER, M.D., BOSTON of (4s Co j AMERICAN JOURNAL THE bt tors CO) leaits Ys VOLUME XII. BOSTON, U.S.A; CINN AND COMPANY 1905 A <> AK) Wy 18 i a ne re i) Ve 7 v. \) Ht G : ~ Je ' ; ‘| ( i Copyright, 1904, 1905 By GINN AND COMPANY Gniversity Press JouN Witson AND SON, CAMBRIDGE, U. SEA = ne ©) CONTENTS. No. I, SEPTEMBER I, 1904. ON THE COMBINED ACTION OF PROTEOLYTIC ENzyMES. By P. A. eeurne anid I. B. SWORey 2 ss 2 = = THE EFFECT OF ALCOHOL AND ALCOHOLIC FLUIDS UPON THE EXCRE- TION OF URic AcID IN MAN. By S.P. Beebe . . «= -» THE ELIMINATION OF ENDOGENOUS URIc Acip. By Elbert W. Rock- PETIA dee SY he. as a bate sy. tae THE RHYTHM PRODUCED IN THE RESTING HEART OF MOLLUSCS BY THE STIMULATION OF THE CARDIO-ACCELERATOR NERVES. Sy oD NE a Oe a ae, a ee et SP Te ae or : THE NERVOUS ORIGIN OF THE HEART-BEAT IN LIMULUS, AND THE NERVOUS NATURE OF CO-ORDINATION OR CONDUCTION IN THE SMUAAUC ILE t. +f (CARLSON +.) | a iidsl ~ a4. bed aD Bene glee Ca ToNuS RHYTHMS IN NORMAL HUMAN MUSCLE AND IN THE GASTROC- NEMIUS OF THE Cat. By Thomas Andrew Storey . ON THE INTERMEDIARY METABOLISM OF THE PURIN-BODIES: THE PRODUCTION OF ALLANTOIN IN THE ANIMAL Bopy. Sy Lafayette B. Mendel and Benjamin White... . . CP igs THE CHEMICAL COMPOSITION OF SOME GORGONIAN CoRALs. By Frank PE <5 i cao Peja a, nga A Tae pee see (ee) ae ahs (5 STRUCTURAL CHANGES OF OVA IN ANISOTONIC SOLUTIONS AND SAPONIN. DG POEL SL LUIOIE wie) eee, a. ew THE EFFECT OF PARTIAL STARVATION ON THE BRAIN OF THE WHITE Bene MEISE FAME 5 sy Wie 85s EY eat Aa coke ON THE REDUCING ACTION OF THE ANIMAL ORGANISM UNDER THE IRPLUENGCE.OF COULD, By \GAVeperer | wee we THE EFFECT OF THE INTRAVENOUS INJECTION OF FORMALDEHYDE AND CALCIUM CHLORIDE ON THE HA&MOLYTIC POWER OF SERUM. Sie CPOE EL], ae Ee See eS eae ear aa PAGE 13 85 95 99 116 128 139 V1 Contents. No. II, OcroBer 1, 1904. ON RHEOTROPISM. I.— RHEOTROPISM IN FISHES. By £. P. Lyon. . THE IDENTITY OF SO-CALLED UREINE (Moor). By H. D. Haskins . THE CHEMISTRY OF MALIGNANT GROWTHS. II].— THE INORGANIC CONSTITUENTS OF TUMORS: By 3S; LGB ceven. 1.) INHIBITION OF THE ACTION OF PHYSOSTIGMIN BY CALCIUM CHLORIDE. By Samuel A. Matthews and Orville H. Brown. . . . « . ON THE ORIGIN AND PRECURSORS OF URINARY INDICAN. By Frank P: Onderhtll oes 6 ea ei OS eee ETHER-LAKING: A CONTRIBUTION TO THE STUDY OF LAKING AGENTS THAT DISSOLVE LECITHIN AND CHOLESTERIN. By S. Peskind . . THE INFLUENCE OF CHLOROFORM ON INTRAVITAL STAINING WITH METHYLENE-BLUE. By C. A. Herter and A. N. Richards ... . HYDROLYSIS OF SPLEEN-NUCLEIC ACID BY DILUTE MINERAL ACID. SL ar Oe Ss | aT hk ere) A STUDY OF THE EFFECTS OF CERTAIN STIMULI, SINGLE AND COM- BINED, UPON PARAMcECIUM. By Elizabeth W. Towle . ... . THE FATE OF STRYCHNINE IN THE INTESTINE OF THE RABBIT. By Robert A. Hatcher 6s 3 ea es es a No. III, NovemMBer 1, 1904. DIFFERENCES IN ELECTRICAL POTENTIAL IN DEVELOPING EGGs. Sy Lad. Flyde eae, et Se se THE AUTOLYSIS OF ANIMAL ORGANS. II. — HYDROLYSIS OF FRESH AND SELF-DIGESTED GLANDS. By P. A. Levene . ~ . eee ON THE SWELLING OF ORGANIC TISSUES. — RESEARCHES ON THE CORNEA? : By G: BUM ete - 0 ss ae ee ON DECAPSULATION OF THE KIDNEY. Sy /saac Levin . . . . «ss No. IV, DECEMBER I, 1904. CHANGES IN THE EXCRETION OF CARBON DIOXIDE RESULTING FROM Bicycuinc. By G. O. Higley and W.P. Bowen. . <2 ON THE ABSORPTION AND UTILIZATION OF PROTEIDS WITHOUT INTER- VENTION OF THE ALIMENTARY DIGESTIVE PROCESSES. Sy Lafayette B. Mendel and Elbert W. Rockwood. . .« s+ « . «. —e THE PRODUCTION OF CHOLIN FROM LECITHIN AND BRAIN-TISSUE. By Isador H. Coridt pe ee ee, 3 rr PAGE 173 176 207 213 220 237 241 276 297 304 311 336 553 Contents. FURTHER EXPERIMENTS ON THE H#MOLYSINOGENIC AND AGGLUTIN- INOGENIC ACTION OF LAKED CorpPusCcLEsS. By G. WV. Stewart . FURTHER PROOF OF ION ACTION IN PHYSIOLOGIC PRocESSES. By C. Hugh Neilson and Orville H. Brown THE PASSAGE OF DIFFERENT FOOD-STUFFS FROM THE STOMACH AND THROUGH THE SMALL INTESTINE. By W. B. Cannon No. V, JANUARY 2, I905. THE TOXIC AND ANTI-TOXIC ACTION OF SALTS. By A. P. Mathews THE QUANTITATIVE ESTIMATION OF CARBAMATES, Sy /. J. &. Macleod and H. D. Haskins . . THE INFLUENCE OF FEVER ON THE REDUCING ACTION OF THE ANIMAL ‘ORGANISM. By C. A. Herter THE PRODUCTION OF FAT FROM PROTEID BY THE BACILLUS PYOCYA- NEUS. By S. P. Beebe and B. H. Buxton FURTHER EVIDENCE OF THE NERVOUS ORIGIN OF THE HEART-BEAT IN LimuLus. By A. /. Carlson PAGE 374 387 419 466 471 THE “American Journal of Physiology. VOU. XII. SEPTEMBER 1, 1904. NOW TE ON Tee COMBINED ACTION OP PROTEOLYTIC ENZYMES.! By Pi A. LEVENE anpek. Bo STOOKEY. [from the Department of Physiological Chemistry of the Pathological Institute of the New York State Hospitals.] HE question of interaction of enzymes of different origin has attracted considerable attention during recent years. The ob- servation of Cohnheim? on the action of pancreatic extracts upon the glycolytic power of other organs was the most important in that direction. Similar observations were made in the laboratory of Sal- kowski by Arnheim and Rosenbaum.* In May, 1903, at the meeting of the American Association of Pathologists and Bacteriologists, we communicated some results of our investigations “‘On the Digestion and Self-Digestion of Tissues and Tissue-Extracts.”* It was noted that the proteolytic action of fresh extracts of the liver and of the spleen is more powerful than the action of one of them on the heated extract of the other. Theoreti- cally there may be several explanations; the higher activity under the conditions of our first experiments might be ascribed to the greater concentration of the proteolytic enzymes, or to the action of the extract of one organ on the zymogen of the other. Halpern® in Salkowski’s laboratory made some observations very 1 Read at the annual meeting of the Association of American Pathologists and Bacteriologists, April 1, 1904. 2 COHNHEIM: Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 337. 8 ARNHEIM and ROSENBAUM: Jézd., 1903, xl, p. 220. * LEVENE and Stookey: Journal of medical research, 1903, x, p. 217. 5 HALPERN: Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 377- I 2 | BP. A. Levene and L. B. Stookey. similar to ours. His object was the study of the action of the extract of the liver on the proteolytic power of the pancreas. Halpern’s conclusion was that the liver may aid the pancreas in its proteolytic action. The conditions of the experiments, however, were such that no decisive conclusion could be reached. Previous to these experiments, this subject attracted considerable attention in connection with the question of the influence of the spleen on pancreatic secretion. A critical review on the subject was given by Mendel and Rettger.!_ Most of their experiments had been done previous to the discovery of proteolytic enzymes in animal or- gans, and were discontinued after the presence of such enzymes had been demonstrated in all animal tissues. The present paper represents the results of observations made on the combined action of spleen and pancreas, and of spleen and liver. Extracts of the organs were used in some of the experiments, and minced glands were employed in the latter part of the investigation. It was found that when spleen and pancreas were allowed to act simultaneously on a foreign proteid,— egg albumin or casein,— the quantity of products of digestion resulting from their action was greater than the total sum of products obtained by the digestion of equal quantities of spleen and pancreas acting separately. It was thus not a summation of the action of two enzymes, but an increase in the digesting energy of one or both. However, from the results of some of the experiments the increased action might be attributed to the pancreas. Thus when three parts of pancreas and one part of spleen were allowed to act on a foreign proteid, the quantity of diges- tion-products was greater than in the experiments where pancreas and spleen were taken in equal proportions or in proportion of one part of pancreas to three of spleen. Another experiment seems to add some weight to this supposition. Fresh pancreas was allowed to stand under antiseptic precautions for about forty-eight hours at room-temperature. At the end of that time, all the zymogen of the gland presumably is transformed into enzyme. Fresh spleen was added, and their combined action on foreign proteid was not notice- ably greater than an equal portion of pancreas similarly treated and an equal quantity of spleen acting separately. Thus, these experi- ments seem to corroborate the views of Schiff, Herzen, and Mendel and Rettger, namely, that the spleen facilitates the transformation of the pancreatic zymogen into the active enzyme. 1 MENDEL and RETTGER: This journal, 1902. vii, p. 387. On the Combined Action of Proteolytic Enzymes. 3 It was impossible to detect a similar action on the part of the spleen upon the proteolytic power of the liver. The products of digestion obtained by the combined action of the liver and of the spleen were not greater in quantity than those obtained by the action of the spleen and the liver separately. EXPERIMENTAL PART. Extracts. — Fresh organs were freed from adhering tissues, minced, taken up in 0.25 per cent of sodium carbonate solution (pancreas), or in physiological salt solution (liver, spleen), and allowed to stand over night in a refrigerator. Chloroform and toluol were used as an- tiseptics in all experiments. At the end of that time, the extracts were filtered, also in a refrigerator, and the filtrates obtained in this manner were used for the experiments. In the other experiments, fresh glands were minced, taken up in 0.25 per cent sodium carbonate, or in physiological salt solution, and thus were used for the investigations. In order to estimate the quantity of products of digestion, a part of the mixture was saturated with zinc sulphate, and another part was treated with phosphotungstic acid. In the filtrates nitrogen de- terminations were made. The zinc sulphate filtrate was prepared in the following manner: The mixture was made acid, saturated with zinc sulphate, and brought to a definite volume by means of a satu- rated solution of zinc sulphate. It was then allowed to stand over night, filtered, and an aliquot part of the filtrate was used for analy- sis. The phosphotungstic filtrate was prepared in an analogous manner. Each experiment consisted of the principal one, in which the two enzymes were allowed to act simultaneously, and of controls in which the proteid was subjected to the action of only one gland or extract. The results of our experiments are given in the following tables: P. A. Levene and L. B. Stookey. ; ius ‘ f * 9}evuasouresvo wn Iso | svo | 60 | 690 +2 eis be 4010) ca, | ey ney ee ke ae pete | 2 na : : * 9}euasoutas 100 | -8TO. )-er0 | =eao +2 » | S06 SP lt SRO SOT pes reeali'g. opp ake. eae t ? i * 9yeudsoutaseo Ww 100 | 600 | 600 | STO +2 a COM eres |r SOOO Menea eee. ade? 80°0 ITO +7 vr z10°0 $8 €20'0 09 OOr pie ae fee HSeOlOUeT, q c0°0 620°0 +7 ONE $00°0 OI 9100 O£ 0's ; as uaalds aT OS SS 2 9S > ay gaedara Rs rerehe| 190 690 80'T 6£T +2 ee 9¢'0 aM Ir'0 SZ st uaetds Jo 39¥1}x9 aUI[eS a Z se svaioued Jo 4ov13x9 dUL[eS : ; ; : se rie 50 9 Os iA * uluinqle 33q ) 43'0 S30 90'T 82'l +e 40 260 os ee svasourd Jo yovsjxa auyeg § | ? 1s0 130 42 S°8¢ +t'0 Z99 910 Os my svaroued Jo 4ov13x9 oUITeS 2 ; : ee unlunqye 33 soo | soo | 940 | +0 +2 Ay ght | destiadoudeuseetnes ! g 65°0 260 +E O12 0s‘0 CE LLO OS 4 uaa[ds Jo yo¥13x9 aUT[eS vy wei3 wed sueid suid queso 10d wei3 queo 10d med "2°D sues “OVEVTY | 9ve.OTG | oye.yy | “over antes Fe A uz | AM uz ae posal ‘SUNOT] aise NI NOLL -tve | Art ‘dX ee see wt N sar | EY MUN | EMMY UZ MEN | AS | Avag ‘SHONVASENS 40 ‘NOILSHSIGC, YWaLAY 10 HLONYT ‘NOLLSHDIG ANOAAT at GIAisl\Ale, 2 2c Enzymes. Proteolyt ad Action of 272€ On the Comb 86'T Som L0'0 +10 140 SET cr'0 800 071 Srl +2 be +7 +E +2 +é +2 +é +Z +2 ce SL 90°0 +0°0 c0'0 +0'0 800°0 S100 STt0'0 +0'0 c0'0 é20'0 uf 60°0 - 90°0 oa £0'0 90°0 7 200 sa $0'0 FOL $0'0 OL 900 as $0'0 ey $s0'0 OFZ O9T 0S 09T 0S O9T OZT oor O9T O91 Oot 09 OL 0°02 O'F Oo! OOor 07@ O'Or OF 002 0% OOr OF 0'02 0:02 0°02 OF OST O's O'F OOr 0OOL ayeuasourased UINIpOS Sc de a Seolouury peg ee ee) aks) ayeuasoulased WINIPOS ae SP OTOUe ee ee EROS TOS tS ees ee ISEONOURGT . . . . . . usge{ds ajyeuasouresed wn1pos ' oa a uaelds ayeuasOUlased UINIPOS . . . . . . usaidg ajyeuasoulased WNIpOS Ap eS BOO ery a}eudsouresed WINIPOS 2 a emISEOTOuery ayeuasourased wntIpos 5 eee i ise oad DEFOE 03 OO RY GKS ayeussourased wWINIpoOS os we ° . . uaafds Joe eae ee SV OIOUL ET a]BuasoUurIIsed WINIPOS . . . . . * yoo jdg FN RS Se SR aIOUE ET q % S . Stookey B vene and L pe By Pad v6'0 607 0 wed mers 9¢9°0 $60°0 9b+'O 6110 98E'0 7770 8c 0 wed +66'0 LEO 9¢9'0 SFL'O FEL'0 We0 8LTO med +Z +7 ; Sc’ se 8t'0 6LT 4900 | ¥S@ | 960°0 Selle || ISHLtG 0'+9 est 690°0 £€? queso 10d weaid queo 10d STO Oot OO OOT S es Cnt Met G Seectecl pac pele ele ‘+ | *svaroued jo yov13x9 8OD%eN %S0 ae * + © + waserds Jo 3981}xo ouT[eS svarourd jo yovsyxo §Q9°eN %S'0 SD aH VaH IS eiehel svaioued jo yovlyxa % 970 * uaaids jo JovI}X9 OUI[RS iiss. ep svaioued jo yovixa % 670 * usajds Jo JoRs} XO BUT[ey svoioued jo yoviyxo §Q9%eN TATA 0 * + + + wsepds Jo 3981} x9 BUI["S J GF DF jahieoanvo pict iste uaa[ds Jo yOv1}xX9 BUI[LS § svaroued Jo yovsjyxe §QO%"N %SZ'0 svaiourd jo yousyxa ®OD*eN % S70 uaatds Jo 3Ov1}xX9 JUI[TeS “OyeAVY *OJeI1V[Y M uZ ae. Y M 87k UZ ut prajord USIAIOF JO N uN “NOILSHSIC, VWALAV SmnNOH NI NOLL -SHSIC] AO HLONAY "NL INL ‘aye AA UN ‘oyeyYy UZ Ul | "NOLESHDI(, ANOMA Tl ATaVL “IOA “SHONVISHOAS 7 wc Enzymes. Proteolyt of On O med Acti On the Combi 8h 0 9&¢6'0 61 0 £600 0670 1930 PIO 610°0 8L¢°0 ces'0 SIl0 ést‘O 0260 SL+'0 SLL'0 9ST‘ be +2 +2 +2 +7 6SE vST STO so'0 120 60°0 STO +L0°0 ce0 Oro £60°0 9¢°0 sto 9¢'0 FILO 2 Sauna ody - + + waalds Jo 39eI}x9 sUI[eS svainurd jo yovrjxa §O9%eN %Sz0 SS rng) saa ! usads Jo yovsjxo ouryes svorourd jo Jovajx9 &§OD%eN YZ 08 2 eeu GT ERs Go a tupelds yogornxs outyes svaiourd jo yovijxe §O9%eN %5770 ove eS st reg aca ) svaroued Jo yovsyxa §Q9%"N %6z0 5 es svoisued jo jov1}x0 8§Q ON %Sz'0 usatds Jo yov.yxa auT[eS svasroued jo yov13x0 8§OO%eN 4570 * + + * waatds Jo 3oRI}xe oUTTeS * + + * wsealds Jo 19R1}x9 oUTTeS wna ue ‘ svasrourd jo yovizxa ®§OO%eN % S50 oe Ne cae ws uaards JO JORAJXI OUT[VS *svoroued jo yorsyxo 8§O9%eN %S'0 - + + + wealds jo Jov3xe UTES *svoroued jo yoryxa ®§Oo%en % See eee eS ee UATEOU G] eaaecdic fe i ) Suserds) yo aEnixe Sete ) P Stookey. B vene and L Dae a a. wes ured ‘ayer | 97810 M UZ ut prajoid uUDSIOIOJ JO N ‘NOMSADICQ] WAIAV 0L0'0 ese0 8920 O+0'0 1S0'0 med ayelq[y NM 92T'0 $s0'0 wed £9 queo ted 97 81U Uz ON ‘sunOoH NI NOL STOIC] 10 HLONGT ANAL ‘a7ye1y|y MA UN | 8qB1Y UZ UN ‘NOILSADIC] ANOAAT panityuor—JTT ATAVL * a}euasoUuTIsed WINIPOS . . . . . . uaaldg * a}euasoUuTaseo WINIpOS Cot) sAneeiae Bop eames uaaids oOf 38 SINoY Qf J9IFe svaioUeT * + + 9}euaSouTased WNIpOS oOt 1¥ Sinoy Sp 1ayye svaioueg seaioueg uaalds ( j | | “SHONVISHNS 2 zon of Proteolytic Enzymes. On the Combined Act + ayeussoulased wWNIpos ) 9900 | 9610 | szto | oszo0 +2 = +£0°0 cose || EO a aes tD sa Dg nanan woaids | 2 : ; ee eae f * gyeuasouresevd wnIpos 0900 | 1gt0 | Seto | zZsz0 +2 re 8z0'0 "| goo | OoT ¢ OS Se et ne oe vas Y a oo eee cee f * ayeuasoutaseo wnipos 0800 | e410 | ISTO | sszo +2 z 6200 Sip a) SPOOL O9E | OI We eee tupac lea Gio epee er 0900 | ebro | OITO | €Iz0 4 ee | EON) 3 BeOS Sane Fe aes ae i 6100 | zr0'0 | 920°0 | 38900 +2 PS a OLN O Mapes ss ble (Oe bay i ime us use Cee a 3 1z0'0 | 440°0 | L200 | ssToO +2 Diy |p "ee AY “SEO0 OORT 1 80 a ar aad ea Sea p 600'0 | ¥20'0 | 8zo0 | 0s0°0 4 SS ecOOO | a | C100) | OF : Wee ti ceaide 2 e200 | es00 +2 9 zoo | 6% | +200 | 09 Or ae ia Seer Vm h (2 900 | Ts0°0 +2 oe | 6000 | os | +200 | 09 ES | OR ON 8 oe ee meee mes wed weg med queo red wes queo 10d wes *o°D suei3 aqeQ[Y | aqVTY | ow [y | ‘aeTy Fits. re ca ee ie ie Now NL ; “SUNOF] ae ul prajo.id xe pees 3 ares SAS) | CANINE ; 7 oy uSta10F jo N ul N- ee oye IVY M ul N 97e1}/G UZ ul N qO -NVAQ STONY ISdns son HLONAT TOK ‘NOILSHOIC] YALAY ‘NOLLSADI(, AVOAT Th Tavs ‘qUON | OUON 8000 | 8200 ‘QUON | OON “OUON eumoON 3000 | 9100 9000 | 0100 £000 | 6100 weg wed ‘ayeaqyy | 972g M UUs, O+0'0 L500 éc0'0 L£0'0 S100 ++0'0 9100 c100 é10'0 weg aye Y M $900 OLl0 c£0'0 ¢c0'0 0Z0'0 wed ‘ay elqy, uz P. A. Levene and L. B. Stookey. ut prajoid USI910J JON ‘NOILSADIC, WALAY 10 46 +2 +7 +2 +2 +7 +7 +7 ‘SUNOP] NI NOLL -SaDIq Rte) HLONAT be £¢0'0 ots ee £600 50 es S100 ae aC 120'0 Be a6 600°0 “ie +e +20°0 are ve 300°0 Be v9 600°0 TL FS queso 10d O+0'0 O+0'0 020'0 0£0°0 010°0 0£0'0 0100 0100 OZT OcI OcT OZI val OcT OcI ARE we ‘ae A UE N | ‘99e1q1y UZ Ul N ‘NOLLSHDIC] WAOIA ‘panuuuor —TIJ AIAVL - Ine InINe? IHBined ARES ene) meq onl men = mm WwW . ayvuasoulases wntpos DEO Tea ees SUet ules © Sieg ayeuasourased WINIpOS S28 OF ONE RES milligrams c.c milligrams 60 28.5 6+ 21 25 * Meal hour. Lffect of Alcohol upon the Excretion of Uric Acid. 29 Experiment 11.— Hourly excretion of uric acid_as affected by alcohol. 150 c.c. of port wine containing 15 per cent by volume of alcohol were taken with the test meal on the alcohol day, Subject E. TAB EH XC. SUMMARY OF EXPERIMENT ll. CONTROL Day. ALCOHOL DAY. Volume. Uric acid. Volume. Uric acid. milligrams Oe milligrams c.c. 26 18 3 30 18 26 21 40 21 26 26 21 RESIDUE FOR THE 24 Hours. Control day. Alcohol day. Wollavant 3 6 Pa be ao eLOlICero Wolume mit us sees we Re eens OD ONC:Ce Ue Belg Ss Bg oo Bg OIL jooveqancy Witcracid Sis meee tne ct Ons Motalubicacidua week «nooo moms, Potal unc acdsee. 1a) Olemems: * Meal hour. 30 ws. Ps Beebe. AVERAGE RESULTS FROM EXPERIMENTS 4, 5, 8, 9, IO, II. The following figures given in Table XII were obtained by averag- ing the results of the six experiments on the influence of alcoholic fluids upon the hourly excretion of uric acid. TABLE XII. ALCOHOL Day. CONTROL Day. Hour. Uric acid. Uric acid. milligrams milligrams 2 18.9 19.0 10 IS 7 16.3 11 Wl 16.4 WE: 16.1 14.4 12.8 15.6 19.3 23.6 16.2 LOS Zoid 28.3 see? 28.8 24.8 Zo RESIDUE OF 24 HouRs. Control day. Alcohol day. Uric acid. . 442 mgms. Uric acid. . 338 mgms, * Meal hour. The results show that an increase of uric acid excretion is in most cases one of the effects of taking alcohol. This increase reaches its maximum five hours after the meal and alcohol have been taken. Other investigators have found that the maximum increase as a result of taking food occurs at the same time, so that an additional argument is afforded for regarding the increase as caused by the alcohol as due to a disturbed metabolism of the uric acid precursors found in the food. Effect of Alcohol upon the Excretion of Uric Acid. 31 Personal idiosyncrasy is, however, of some importance in this behavior toward alcohol, as it is in metabolic processes in general. The experiment given on the following page is the only one in the whole series that was carried out for this paper in which an increase of uric acid excretion did not follow the taking of alcohol. In both the experiments on Subject C, a smaller quantity of uric acid was ex- creted immediately following the meal on the alcohol day than on the control day; and the total quantity for twenty-four hours was smaller on the alcohol day. It will be observed, however, that even in this case the amount of uric acid excreted between 6 p.m. and 8 p.m. the following morning was considerably greater on the alcohol day. The same plan was observed in this experiment as in the others, and the only explanation for such a divergence from the results obtained under similar condi- tions with other subjects is the personal idiosyncrasy of Subject C. 22 S. P. Beebe. Lixperiment 12. — Hourly period experiment on the excretion of uric acid as affected by alcohol. On the alcohol day, 700 c.c. of India Pale Ale containing 8 per cent by volume of alcohol were taken with the test meal, Subject C (Table XIII). TABLE XIII: SUMMARY OF EXPERIMENT 12. CONTROL Day. ALCOHOL Day. Volume. Uric acid. Volume. Uric acid. G58, milligrams Gs milligrams 24 13 20 21 21 23 21 17 11 15 25 26 26 27 24 RESIDUE FOR THE 24 Hours. Control day. Alcohol day. Woltnres i) 6: ere. coun O5O0lcre: Volume Whar eee! 9 oo 6 9 6 9 SILe loner: Uric acid Totaliinielacid sao oo mems: Total uric acid. * Meal hour. Effect of Alcohol upon the Excretion of Uric Acid. 33 LEixperiment 18.— Subject C. On the alcohol day 300 c.c. of port wine con- taining 15 per cent by volume of alcohol were taken with the test meal (Table XIV). TA BICE XLV SUMMARY OF EXPERIMENT 13. CONTROL Day. ALCOHOL Day. Volume. Uric acid. Volume. Uric acid. ie:cs milligrams Cs milligrams 62 27 46 28 16 28 15 13 22 13 14 13 10 iil * Meal hour. Two metabolism experiments were carried out on different subjects, A and F, in which alcohol was taken in the form of ale. The same plan was followed as in the previous metabolism experiments, great care being taken to have a uniform diet throughout. The results are shown in Table XV, for Experiment 14, and Table XVI, for Experiment 15. They show in each case a marked increase of uric acid excretion during the alcohol period. The experiments differ in no essential way from those previously carried out, in which alcohol in other forms was used, and they simply afford additional evidence of its action. Experiment 14. — Metabolism experiment, Subject A. Alcohol in the form of an ale containing 8.44 per cent by volume of alcohol was used. The same general plan was followed as in the previous metabolism experiments. 4 S. P. Beebe. TABLE XV. SUMMARY OF EXPERIMENT 14. Date. Volume. Nitrogen. Uric acid. Ale. c.c. grams gram c.c. Feb. 23 1060 11.40 0.354 eas 1240 13.41 0.432 25 1370 14.64 0.516 f~ 26 1000 14.65 0.528 Sewell 1200 14.78 0.516 S28 1700 14.11 0.625 1050 epee! 1400 13:94 0.768 1050 73 1460 S719 0.612 1050 eens) 1260 14.04 0.738 1050 Soest 1320 14.15 0.558 ee 4S 1120 14.04 0.564 ray ko 1000 14.16 0.540 DaILy AVERAGES. Nitrogen. Uric acid. grams gram Fore period . 14.65 0.520 Alcohol period . 13.82 0.685 After period 14.12 0.554 RESULTS OBTAINED BY ANALYSIS OF COMPOSITE SAMPLES. Total nitrogen Urea nitrogen Uric acid Total phosphates . Total sulphates . Combined sulphates Fore period. Alcohol period. grams 14.460 12.770 0.522 1.950 2.700 0.214 grams 13.190 11.870 0.645 2.175 3.118 0.214 After period. grams 13.620 12.260 0.516 2.220 2.402 0.201 Lifect of Alcohol upon the Excretion of Urie Acid. 35 Experiment 15.— Alcohol in the form of ale containing 8.44 per cent by volume of alcohol was used in this experiment, Subject F. TABLE XVI. SUMMARY OF EXPERIMENT 15. Volume. Nitrogen. Uric acid. grams gram c.c. 1080 10.56 0.378 820 Tee 0.390 1040 12.75 0.432 1500 13.08 0.516 1100 1278 0.490 860 12.45 0.402 COMPOSITE SAMPLES. Fore period. Alcohol period. grams , grams INGtLOREIN ane Teuneiera cs 11.410 12.740 (WIRE SIROM hee ogo. Ae Be 10.700 11.700 COA CCIARe time arree fete f 0.393 0.498 Total phosphates) . . . . . 1.905 2.400 Total sulphates beth o 6b 2.239 2.809 CONCLUSIONS. Some of the metabolism experiments give evidence of the proteid- sparing effect of alcohol ; for instance, in Experiment 13 the average daily nitrogen of the fore period was 14.65 grams; of the alcohol period, 13.82 grams, of the after period, 14.12 grams. It will be re- called, however, that the proteid-sparing effect appears with a person unaccustomed to alcohol only after the alcohol has been taken for some days; and the conditions of the writer’s experiments prevented the appearance of this effect. But Experiment 13 was carried out upon a subject (A) who, though unaccustomed to alcohol in the beginning, had served in a number of experiments, and had thus reached a condition of partial immunity to the toxic effects of the alcohol. 36 S. P. Beebe. After a consideration of these experiments, it hardly seems possible to doubt that alcohol, even in what is considered by the most conserv- ative as a moderate amount, causes an increase in the excretion of uric acid. And this effect is seen almost immediately after taking the alcohol. The following points indicate that the effect is due to a toxic effect on the liver, thereby interfering with the oxidation of the uric acid derived from its precursors in the food: 1. Alcohol taken without food causes no increase. See Exper- iment 3. 2. In Experiment 2, the diet contained much less purin food than in Experiment 1, and there was a smaller increase in excretion. 3. The maximum increase occurs at the same time after a meal as it does when purin food but no alcohol is taken. 4. The purin bases are affected to the same degree as the uric acid. 5. Alcohol is rapidly absorbed and passes at once to the liver, the organ which has most to do with the metabolism of proteid cleavage products. There is no evidence that the alcohol has merely hastened the ex- cretion of urates normally present in the blood; the increased excre- tion means that a larger quantity has been in circulation, and although it is classed by Von Noorden ' among the substances easily excreted, still most physiologists would consider the presence in the blood of this larger quantity as undesirable. Certainly in pathological condi- tions it might be harmful. If we accept the origin of the increased quantity of uric acid to be in the impaired oxidative powers of the liver, the results of these ex- periments will have greater significance than can be attributed to uric acid alone. For the impaired function would affect other processes which are normally accomplished by that organ, and the possibilities for entrance into the general circulation of toxic substances, of intes- tinal putrefaction, for instance, would be increased. The liver per- forms a large number of oxidations and syntheses designed to keep toxic substances from reaching the body tissues, and if alcohol, in the moderate quantity which caused the increase in uric acid excretion, impairs its power in this respect, the prevalent ideas regarding the harmlessness of moderate drinking need revision. 1 Von NooRDEN: Diseases of Metabolism and Nutrition, nephritis, 1904, p- 44. Lffect of Alcohol upon the Excretion of Uric Acid. 37 Alcohol is a food in the sense that when used in small quantities the energy from its oxidation may be used for some of the body needs; ‘but since, at the same time, it interferes with the normal activities of a most important organ, its food-value may be overbalanced by its toxic effect. Salt water may be used in a steam-boiler, and the steam from its evaporation may transmit the energy of the fuel to the revolving wheels, but its corrosive action on the steel forbids its use, like alcohol, except in emergencies. The writer is much indebted to Professor Chittenden and Professor Mendel for many helpful suggestions during the course of the work. THE ELIMINATION OF ENDOGENOUS UORIC, ACI. By ELBERT W. ROCKWOOD. [From the Laboratory of Physiological Chemistry, State University of Iowa.| HE intimate relationship between the so-called purin constitu- ents of the diet and the output of uric acid is too firmly estab- lished at the present time to require detailed discussion! It is likewise well known that the elimination of uric acid does not cease during hunger or after the ingestion of foods which are free from purin components. It is obvious that in these cases the uric acid must have its origin in tissue constituents; the purin derivatives of disintegrating cells rich in nucleic acid compounds at once suggest themselves as precursors of the eliminated substance. Burian and Schur? have attempted to distinguish between exzdogenous and exog- enous urinary purin constituents, the latter term being applied to that fraction of the total output which has its origin in the purin bodies of the foods, while the endogenous component is referred to tissue metabolism. As the result of their extensive investigations, Burian and Schur come to the conclusion that the endogenous com- ponent of the eliminated urinary purin compounds is constant for each individual, that it is, in other words, an individual factor. Mares * had earlier concluded that the elimination of uric acid reaches a constant for every individual, although the individual differ- ences may vary widely. His publications date from a period when the genetic relation between the purin compounds and uric acid was little understood. Schreiber and Waldvogel* had also assumed a 1 The enormous literature on the subject has been reviewed in monographs by SCHREIBER: Die Harnsaure, Stuttgart, 1899; WALKER HALL: The Purin Bodies of Food Stuffs, 1903; WIENER: Ergebnisse der Physiologie, 1902, i, Part 1, p. 355; Lbzd., 1903, ii, Part 1, p. 377; BuRIAN and SCuHurR: Archiv fiir die gesammte Physiologie, 1900, Ixxx, p. 241. 2? BuRIAN and ScuuR: Loe. cét., also Archiv fiir die gesammte Physiologie, Ixxxvii, p. 239. 3 MARES: Archives slaves de biologie, 1888, iii, p. 207. 4 SCHREIBER and WALDVOGEL: Archiv fiir experimentelle Pathologie und Pharmacie, 1899, xlii, p. 69. 38 The Elimination of Endogenous Uric Acid. 39 constant endogenous output of uric acid, as the result of observations on starving men. The criticism has, however, properly been offered that the data obtained during inanition — when a continued breaking down of cellular tissue may reasonably be assumed to take place — are scarcely to be accepted as ‘“ normal” or “ physiological ” values. The conclusions of Burian and Schur with reference to the constancy of the endogenous purin output are based on a large number of ob- servations upon men living upon a strictly purin-free diet. Sivén! likewise found a remarkable constancy in his output of uric acid during periods in which the purin-free diet was widely varied in quantity and composition. Loewi? has assumed that the endogenous output of uric acid is dependent in a measure on the food ingested, — a conclusion which has, in turn, been subjected to rigorous criticism.? Finally Kaufmann and Mohr favor the theory of the individual con- stancy of uric acid, publishing experimental data in confirmation. They believe, in distinction from Burian and Schur, that the endoge- nous nuclein metabolism can be diminished by increased ingestion of non-nitrogenous food, — ‘durch Calorieniiberfiitterung,” as they express it. “‘ The sources of endogenous purins are probably numerous, and the quantities derived from each may vary with the hourly activities and daily needs. Although our present knowledge upon the point is somewhat inadequate, we may be sure that in pathological conditions alterations in any one of the factors may lead to diminution or increase of endogenous purins. So far as experimental results can suggest normal action, one portion of the total endogenous purins is broken down to urea, and the remainder excreted as uric acid. Abnormal endogenous purin metabolism may, therefore, consist in an increased production with excessive or diminished destruction. Hence arises the difficulty of any correct inference from the results of endogenous purin elimination. Constancy of endogenous purin excretion points to normal metabolism and the maintenance of the several factors con- cerned. Variations in the endogenous urinary purin of the same indi- vidual upon a fixed diet indicates altered relations of the contributory 1 SIVEN: Skandinavisches Archiv fiir Physiologie, 1901, XI, ps 032: ? LoEw!: Archiv fiir experimentelle Pathologie und Pharmacie, 1900, xliv, p. 1; Archiv fiir die gesammte Physiologie, 1902, Ixxxviii, p. 296. 3 Cf. BURIAN and ScHuR; also WIENER: Loc. cit. 4 KAUFMANN and Monr: Deutsches Archiv fiir klinische Medizin, 1902, Ixxiv, Pp- 141, 348. 40 Elbert W. Rockwood. functions.” + The preceding quotation from Walker Hall, influenced largely by the work of Burian, indicates the importance of determining definitely the question regarding the constancy of endogenous uric acid output and its dependence upon the individual and his daily dis- position. The following experiments were undertaken as a contribu- tion in this direction. The selection of a purin-free diet was obviously of primary importance. It is well known tbat foods of animal origin, as a class, contain much more of the purin compounds than do the vegetable foods. Among the latter we find, from the analyses of Burian and Schur (doc. cit.), and of Hall (Zoc. cz¢.), that rice, eggs, milk, butter, cheese, and preparations of fine wheat, such as bread, contain none or but very minute quantities. Most of the older ex- periments on uric acid formation have a lessened value because of a non-recognition of these facts. In many instances where some sub- stance was studied as to its effect upon the production or elimination of uric acid, too little attention has been paid to providing a basal diet which in itself would not serve as a source of the acid. The plan followed in the present trials was to have the subject select from the purin-free foods such a diet as he could endure and, having proved this by trial, to take the same kind and amount of food and at the same times of day as long as the test lasted. Trials were made at different times with the same subject, using the same and also differ- ent foods; two subjects were tested with a common diet, and the influence of other factors, such as labor, increased food-materials, or water were studied. The analyses of the urine were conducted as follows: nitrogen was determined by the Kjeldahl process; uric acid by Folin’s method; phosphoric acid (P,O;) by titration with uranium acetate. EXPERIMENTS. Subject A, First Series.— The subject was a healthy male of 42 years, weighing about 55 kilos. During the first nine days the daily diet consisted of: iti en Ree iss vale eed, Oo. dads vexe: SOT OLCE 2 © (uses: penal ti Raat ce 35 gms. OGEIII & kee ta 6 a oe SO SUS AT eis mits cet ts mete and Quincke, dependent upon protoplasmic streaming. 1 Hype, I. H.: Zeitschrift fiir wissenschaftliche Zodlogie, 1894, lviii, p. 532. 2 Logs, J.: Journal of morphology, 1892, p. 253. 3 Loe, J.: This journal, 1900, iii, p. 435. 4 LoEs, J.: This journal, 1902, vi, p. 432. 5 BUTSCHLI, O.: Untersuchungen iiber mikroscopische Schaume und das Protoplasm, 1892. 241 242 Ida FH. Flyde. This requires as Quincke has shown, a definite degree of viscosity. If the viscosity is too great or too small, no protoplasmic streaming is possible. Since the viscosity is altered by the presence of electrolytes, the streaming of the protoplasm must also be influenced by the presence of these. We learn from D’Arsonval! that changes in surface tension also produce protoplasmic streaming, and he shows that on the basis of Lippmann’s observations electrical charges must lead to changes in surface tension. Loeb also believes that there is a possibility that part of the chemical energy in the development of the egg is con- verted into some form of electrical energy, that the ions formed in metabolism play a réle in the dynamics of life-phenomena, and that these ions or electrical charges may be responsible for such physical manifestations as cell development and segmentation. Besides these, other investigators who have studied the mechanics of mitosis agree that a decrease in surface tension over a definite area of the egg, and a flowing state of the protoplasm are two of the factors upon which segmentation depends. We may assume that these conditions are in turn produced by certain definite physico- chemical changes occurring among the nuclear and cytoplasmic ele- ments of the egg, and also between the egg itself and its surrounding medium. Of these changes, the most efficient are the actions of the ions through their electrical charges, the enzymes, and a certain amount of water. It would seem therefore that in the concentrated sea-water, the absorption of water from the egg’s contents instituted a condition in the egg which interfered with the normal actions of its ions, and probably also its enzymes, so that segmentation of the cytoplasm did not proceed. I believe that physiological or metabolic changes dependent upon certain definite physical interactions of electrolytes and colloids must occur throughout the life-history of the developing egg, and are accompanied by differences of electrical potential, which, if demon- strable, would give information of certain changes progressing in the egg throughout its ontogeny. The differences of potential might be employed, for instance, as a measure to ascertain the time in the development of the egg when the phases of heightened and lowered activities occur, the regions in which they take place, the relation of the electrical organization in the egg to that in the embryo and 1 D’ARSONVAL: Archives de Physiologie, 1889, p. 460. Electrical Potential in Developing Eggs. 243 adult, as well as the effect produced by various external influences or solutions upon the different phases and the time of their occurrence. I determined, therefore, to ascertain if a difference of electrical poten- tial in the egg could be detected. Several kinds of eggs were tested at the Naples Zodlogical station, with a very sensitive capillary elec- trometer, constructed especially for this purpose. I had the satis- faction of learning that an electrical difference of potential could be detected. My time was however occupied with other work, so that the study on electrical polarity was postponed. During the summer of 1902, while at Wood’s Hole, I resumed the investigation begun six years previously, but did not publish the re- sults at once, because I decided to investigate the egg of the Toad- fish, which matures early in June. Since] first began this investigation many important contributions bearing upon the physical and chemical changes in protoplasm, especially regarding its colloids and ions have enriched this field of physiology. These achievements have not only strengthened the belief in the chemical and electrical phenomena accompanying development but have offered results with the aid of which the existence of these phenomena can be better understood and explained. I now endeavored to ascertain whether in maturating and fertilized eges of different animals the differences of electrical potential were related to known embryological stages, and whether there existed an electrical polarity related to the astral radiations, spindle fibres, and polarity of the egg which Driesch/ maintains exists in every cell, and upon which segmentation and organization depend. I desired also to determine whether, in fertilized eggs, rhythms of differences of elec- trical potential might be discovered which would coincide with those ‘special rhythms discovered by Mrs. Andrews? in starfish and Echinus . eggs. In the eggs of these forms Mrs. Andrews found that the time required for staining or fixing reagents to act varied rhythmically, and that these rhythms are sympathetic with rhythms of resistances to pressure. Then, too, I wished to learn whether rhythms of differ- ences of electrical potential are related to the periods of resistance, to lack of oxygen, presence of potassium cyanide, and production of carbon dioxide which Lyon? found in Arbacia eggs. These periods may be related to those of susceptibility to mechanical agitation 1 Driescu, H.: Archiv fiir Entwickelungsmechanik, 1897, iv, p. 79. 2 ANDREWS, G. F.: Journal of morphology ; supplement, 1897, xii, pp. 30-57. 8 Lyon, E. P.: This journal, 1902, vii, p. 56; 1924, xi, p. 58. 244 Ida Ff. Ftyde. found by Scott! in unfertilized eggs of Amphitrite, and to the stages in the egg of the starfish highly susceptible to artificial fertilization, which Delage? believes occur, at a time between the breaking down of the germinal vesicle and the appearance of the first polar body. Another question which might be worth considering is, whether the electrical potential is increased or decreased by changing the environ- ment of the egg, either by changes in temperature or ions in the surrounding medium, and whether such changes in difference of potential bear a definite relation to the susceptibility to artificial parthenogenesis. METHODS. The difference of potential existing between two parts of the egg was determined with a D’Arsonval galvanometer, as well as with the most delicate capillary electrometers of the modified Lippmann, Por- ter,2and Lyon* types. Purified mercury, chemically pure sulphuric acid, and glass parts that were thoroughly cleansed, were used in the construction of the electrometer. The movements of the meniscus in the capillary were measured with an ocular micrometer scale. Each division indicated approxi- mately 0.00001 volt. The nonpolarizable electrodes were composed of zinc, zinc sulphate, clay, and tiny brushes. The clay and brushes were moistened with % sodium chloride solution, isotonic with the fluid contents of the turtle and frog’s body or egg, or an m 2 sodium chloride solution, when working with fishes’ eggs. The electrodes were tested before and after each reading, in order to keep them isoelectric for a considerable time. Each part composing the elec- trometer and connected parts received the most careful attention. The apparatus stood on a firm stone foundation, so that all external vibrations were eliminated, and the temperature of the room was constantly taken into consideration. The eggs and brush electrodes were kept at practically the same degree of moisture throughout the observations. 1 Scott, J. W.: Biological bulletin, 1903, v, p. 35. 2 DELAGE, Y.: Archives de zodlogie expérimentale et générale 1901, ix, p. 88. 8 Used in the physiological laboratory of the Harvard Medical School. * Used in the physiological laboratory of the University of Chicago. Electrical Potential in Developing Eges. 245 MATERIAL. Eggs from different types of animals were investigated. Those of invertebrates and of most fishes were too small to be advantageously employed. Nevertheless, besides some unsatisfactory results with eggs from the toadfish, a series of observations were obtained with greatest difficulty and care, from the eggs taken from Fundulus. A few observations were made on toads’ eggs and more on eggs from turtles. The records of a few experiments conducted on an uncertain species of toad’s eggs, several hours after fertilization, indicated a difference of electrical potential of a mean of 0.00002 volt, between the poles of an axis, at an angle to an axis directly through the pigmented and unpigmented regions of the egg. In tadpoles from eight to ten days old the difference of potential was about 0.0002 volt, and in a direc- tion from the tail to the head of the embryo. If the first cleavage plane corresponds to the median plane of the toad’s body, the direc- tion of the current in the egg and embryo is the same in reference to the first cleavage. The results obtained from the toad and toadfish were too meagre to be of value. DIFFERENCES IN ELECTRICAL POTENTIAL IN THE TURTLE’S EGG. (CHRYSAMYS PICTA.) The egg was. taken after decapitation from the ovary or oviduct. Adhering fluid or connective tissue was carefully removed, and the egg placed on filter paper moistened with % salt solution. The egg was touched only with camel’s hair pevenes or delicate paraffine- covered egg lifters. Eggs were studied at different stages of their maturity. They ranged from five to twenty millimetres in diame- ter. They are somewhat ellipsoidal in form. The germinal disc, with one or more brownish micropyles, is recognized at one pole by its lighter yellow color. The eggs are fertilized in the ovary, and the segmentation nucleus lies directly under the germinal disc during maturation, but after fertilization it is further from the surface. The egg is surrounded, in the late segmentation stages, by a closely adher- ing glary mucilaginous membrane which cannot be removed without injury to the egg. Therefore only the eggs in the maturation and first segmentation stages proved available. 246 Ida Fl. FLyde. A. The first object was to determine if a difference of potential existed between the animal and vegetative poles, and if so, the direction of the current in the egg. For this purpose the egg was placed with its respective poles in contact with the brushes of the non-polarizable electrodes. As usual, the scale reading was first obtained with both brushes in contact, to show whether the elec- ABICR el: (A). — SHOWING THE DIFFERENCE OF POTENTIAL BETWEEN THE BLASTODISC AND VEGETATIVE POLE OF THE MATURE TURTLE’S EGG. Diameter | Scale readings Beles paper in in 0.00003 ne : r blastodise toward the centimetres Volt. é se vegetative pole. Eggs reversed. Current reversed. ee 1.6 1.8 i From the vegetative pole toward the blastodisc. 1.6 “é “cc (B).— Megan ReEsuL’s FOLLOWING PUNCTURE TO THE VEGETATIVE POLE OF THE TURTLE’S EGG. Direction of the current | Eggs uninjured. in the egg from the blastodisc to the vege- tative pole. Vegetative pole to the | Eggs injured and blastodisc. not reversed. trodes were isoelectric. When several readings had been secured with the egg in one position, the egg was reversed, and the extent and direction of the movements of the meniscus of the mercury, or of the scale readings, were compared with ‘those of the first readings. Table I is a record of a few of many observations of the differences of potential existing between the animal and vegetative poles of the ® Electrical Potential in Developing Eggs. 247 turtle’s egg. It shows that the more mature eggs have a greater difference of potential than have the smaller less developed ones, Though as a rule, the difference of potential varies within small limits between those of the same size, nevertheless, the least current is found in the smallest eggs. With but few exceptions the direction of the current in the egg was from the animal toward the vegetative pole. The records indicate that as maturation of the egg progresses, metabolism increases, and that its seat of action is greater at the animal pole. B. From Experiments 7 and 25, Table I, it will be observed that the direction of the current in these eggs is the reverse of that observed in the other eggs. This indicates that at the time the read- RABLE, TT: DIFFERENCES OF POTENTIAL BETWEEN THE POLES OF THE MAIN AXIS OF THE BLASTODISC IN THE TURTLE’S Ecc. : . Egg diameter Scale reading BOE gS in cms. in 0.00003 volt. 1.0 ing was taken, there was a greater difference of potential at the vegetative pole. To ascertain whether this may be due to injury, several eggs were slightly punctured at the vegetative pole. The mean result obtained from these eggs, as stated in B, Table I, shows not only that the current was reversed after injury to this pole, but that it was in excess over that existing in the egg before it was punctured. The current of injury which was produced may be the result of factors similar to those that cause the demarcation current in nerve-muscle preparations. C. It was next of interest to determine if differences of electrical potential existed between the poles of the main axis in the blastodisc. If the egg is placed on a glass stand so that the blastodisc is upper- « 248 Ida FH. FLyde. most, and the tiny non-polarizable electrodes are placed at the same level at different, opposite points of the disc, it is found that there are two such points that exhibit the greatest difference of potential. These poles bear a definite relation to the long axis of the egg, and may correspond to the future head and tail end of the body. They may be considered the poles of the main axis. Accordingly, an axis midway and perpendicular to the main axis, would point to the right and left sides of the body. A few results of many experiments conducted for the purpose of ascertaining whether there is a fixed polarity in the blastodisc, are tabulated in Table II. It was learned that the same strength of cur- rent or difference of potential was registered, but reversed in direction, when the electrodes were placed at reversed poles of the main axis TABLE III. MEAN RESULTS OF SEVERAL EXPERIMENTS TO DETERMINE THE [)IFFERENCE OF POTENTIAL EXISTING AT DIFFERENT LEVELS BELOW THE BLASTODISC OF THE MATURATING TURTLE’S EGG. Wxpexiente Gis an ene Axial distance below Scale reading in I ; : the blastodisc. 0.00003 volt. One-fourth Half-way Three-fourths Half-way of the blastodisc, or if the egg was reversed. This proved that there is a fixed polarity in the blastodisc of the egg, and that there is a determinable constant current flowing in a definite direction in it, Whether this current is from the tail toward the head end of the adult, or vice versa, can be proved only by comparing the current in the axis in the animal area throughout the different phases of devel- opment. When the electrodes are placed at the poles of an axis perpendicular to the main one, as a rule no difference of potential is detectable, though in a few instances a slight, fixed current was observed. D. Differences of electrical potential between the poles of an axis at different distances below the blastodisc of the maturating turtle’s egg were next investigated. When the electrodes are placed Electrical Potential in Developing Eggs. 249 at the poles of an axis, parallel to the blastodisc, and below it, that is, perpendicular to an axis extending through the animal and vege- tative poles, it is proved that the difference of potential decreases as the electrodes are shifted further away from the blastodisc area. Half- way down, there is one diameter at which the meniscus of the mer- cury registered no difference of potential between the poles. In some eggs the difference of potential was zero half-way down ; but three-fourths down it was above zero, as tabulated in Table III. Pos- sibly this was because the blastodisc is exequatric, and the electrodes were placed at unequal distances from the pole of the highest potential. The experiments on the maturating turtle’s egg proved the pres- ence of a difference of electrical potential between the blastodisc and the opposite pole of the egg. This difference increases with the development of the egg, and is probably the result of chemical or physico-chemical activity in the blastodisc area predominating over that of the vegetative pole. From the fact that a difference of poten- tial exists between the poles of the blastodisc area, producing a cur- rent in a definite direction as regards both the blastodisc and the egg as a whole, it is evident that a fixed polarity exists in the blasto- disc. It is possibly due to this polarity that a difference of potential is detectable at different levels below the blastodisc pole, and per- pendicular to the main axis of the egg. II. CHANGES IN ELECTRICAL POTENTIAL IN THE EGG OF FUNDULUS AFTER FERTILIZATION. Eggs from Fundulus were obtained by stripping the fish. The eggs were fertilized in sea-water, in which they were kept under observation throughout the experiment. The appearance of external changes, as well as the time of their occurrence, were noted; it was therefore possible to ascertain whether development between the brush electrodes was at all times pro- ceeding normally. The first cleavage appeared in about one and one- half hours. The second about one hour later, and the third about forty-five minutes after the second, depending upon the temperature of the water. The eggs are from two to two and one-half millimetres in diameter, and the germinal disc is easily recognized by its whitish color. The brush electrodes and the clay were moistened with mm 3 sodium chloride solution isotonic with that of sea-water, and the 250 [da 1. Flyde. eggs were washed in m 3 salt solution, before being placed between the brushes. Each division of the scale in the ocular micrometer indicated 0.00001 volt; the resistance of the non-polarizable elec- trodes was determined to be 10000 ohms. It required many patient efforts before a reading of the movements of the mercury meniscus could be recorded. Only four experiments TAB IEE EV: DIFFERENCES OF ELECTRICAL POTENTIAL BETWEEN THE ANIMAL AND VEGETATIVE POLES IN FERTILIZED EGGS OF FUNDULUS. JK. ‘Al ime in Scale divi- Time in Scale divi- Obsesaee minutes sion in én the minutes sion in 6a the after fertili-}| 0.00001 ’ after fertili-| 000001 ‘ zati , ess: : egg. zation. volt. zation. volt. Observation 15 =l|9 180 4-cell stage. 20 Segmentation. 185 Nucleus formed. 190 195 Third furrow. First segmenta- §-cell stage. tion. Furrow. * + means that the current in the egg is from the blastodisc toward the opposite pole. t+ — means that the current is from the opposite pole to the blastodisc. were satisfactory, of these, two are shown in Table IV. The first record (A) was secured fifteen minutes after the eggs were fertilized. The egg under observation and the electrodes were placed in a moist chamber, that is, in a glass vessel in which was placed moist filter- paper, so that evaporation from brushes and eggs was avoided. The Electrical Potential in Developing Eggs. oye)! difference of potential of the egg existing between the animal and vegetative poles was obtained at consecutive intervals of time by placing the egg in the electrometer circuit for a moment, and observ- ing the extent and direction of the movements of the meniscus of the mercury. The egg was under observation for one hour and thirty minutes. The first segmentation furrow had appeared after ninety minutes. The plus and minus sign before the number of scale divi- sions indicate the direction of the current in the egg. The plus is | to express the direction of the current from the blastodisc to the opposite pole, the minus the current in the egg, in the reverse direc- tion, that is, from the vegetative pole toward the blastodisc. It is seen that the difference of potential is not constant, and that it shifts, during the interval from the.time of fertilization until the appearance CuRVE A. CuRVE B. 6 1 | + f fess mil x | fe) 0 2 4 | «Lt | ct | | + TIME ie) 10 20 30 40 50 60 70 «680 90 100 IO 1806190 6200 210 220 Xx 5 My Xs 3 Curve A illustrates Table IV. It shows the changes in differences of electrical potential in the ege of Fundulus from fertilization to the first segmentation furrow. CURVE B illustrates the changes in the differences of electrical potential after the second to the third segmentation furrow. At F, fertilization of the egg took place; X,, forma- tion of the segmentation nucleus; Xg, first segmentation furrow appeared; Xs, third segmentation furrow appeared. The abscissa indicates time in minutes; the ordinate, the scale readings in direction and strength of electromotive force. of the segmentation furrow, from one pole to the other. For instance, fifteen minutes after fertilization, the current in the egg was from the vegetative toward the animal pole, and the difference of potential was only one scale division, or about 0.00001 volt; since the scale divi- sions were so close, it was difficult to estimate whether it was the whole or only a fraction of the division. Five minutes later a great change was indicated, the current was now from the blastodisc and the electromotive force two scale divisions. The next reading was taken ten minutes later, and showed no change in electromotive force and direction of the current. In the long interval of ten minutes, however, judging from the observations made on other eggs, some changes must have occurred. Fifteen minutes later the current was reversed, and during the succeeding twenty minutes it had 252 [da Hl. Flyde. gradually increased to its greatest strength, while during the follow- ing fifteen minutes it fell to its least. By the time the next record was secured ten minutes had passed, and in that time the first seg- mentation furrow had appeared. The current changed its direction, and in five minutes again gained its greatest electromotive force, only to again lower it to its least amount in the succeeding ten minutes. The second record (B), Table IV, shows practically the same rela- tion of increase and decrease of potential after cleavage had begun as was found in the other records, following a certain interval of the time after fertilization. A COMPARISON OF MY OBSERVATIONS WITH THOSE OBTAINED BY OTHERS. The records of the differences of electrical potential obtained from the fertilized eggs of Fundulus would indicate that from the time of fertilization until the appearance of the first cleavage, and from that time onward, pronounced physiological activities took place ac- companied by electromotive changes. Judging from the direction and extent of these changes, several distinct phases are discernible. The first begins with fertilization and extends over a period of about fifteen minutes, after which a change in the direction of the current is manifested. The second phase continues about fifteen minutes, during which time the current is again reversed. The third phase continues about thirty-five minutes. It is characterized by a gradual increase in difference of electrical potential. The fourth lasts about fifteen minutes, during which time the difference of potential falls from its greatest to its least value. The fifth is illustrated by a change in the direction of the current and a rise in the difference of electrical potential, and during this phase the first segmentation fur- row appears. The sixth is noted by a gradual fall in potential. About twenty minutes after fertilization, near the time when the first segmentation nucleus is formed, and again at the time of the appearance of the segmentation furrow, the current in the egg has the same direction, that is, the seat of greatest potential, or physico- chemical change is in the same region of the egg. The direction of the current is from the blastodisc toward the vegetative pole, before the eggs are fertilized. It changes its direction after fertilization. From the observations noted in Table IV, and curve (A), it is evi- Electrical Potential in Developing Eggs. 253 dent, that there are rhythms in regard to the changes in the direction of the current. After fertilization there is a negative current, that is, one whose direction is towards the blastodisc; it may be due to changes in chemical reactions that affect physical changes. Then a positive current sets in, arising from the blastodisc and followed by a current first negative and then positive. Moreover these rhythms are characterized by an increase and then a fall of potential. These phases as well as their causes may be variously interpreted by embryologists, microchemists, and physico-chemists. From the embryological standpoint, the different phases of electrical potential might be compared with certain phases in the egg during the in- terval between fertilization and cleavage. The first, for instance, might be compared with the period occupied by the entrance of the spermatozo6n and changes instituted thereby in the egg; the second, with the time of fusion of the male and female pronuclei and the formation of the segmentation nucleus; the third, with the prophase; the fourth, with the metaphase; the fifth, with the anaphase; and the sixth, with the telaphase. The curves of electrical potential may furthermore bear a definite relation to rhythms of viscosity and pigment wandering; to phases of resistance of the mass to pressure, and to the special chemical reactions of the astral rays, which Mrs. Andrews! observed in starfish and sea-urchins’ eggs during maturation and segmentation. The facts may also be brought into relation with the staining re- actions of chromatin as observed by Heidenhain? and Riickert? who concluded that the proportion of nucleic acid and albumin varies with periodical changes in the nucleus, and that these are related to changes in the function of the nucleus and to definite mitotic phases. Especially may certain phases of these rhythms of differences of potential be compared to those of resistance to lack of oxygen, or resistance to potassium-cyanide, or to those rhythms of production of carbon-dioxide during cleavage that were found by Lyon? to exist in the Echinoderm egg. Lyon found that the Echinoderm egg was more 1 ANDREwS, Mrs. G. F.: Journal of morphology; supplement, 1897, xii, PP: -30-57: ? HEIDENHAIN: Archiv flr mikroscopische Anatomie, 1894, xlili, p. 423. 8 RUCKERT :: Anatomischer Anzeiger, 1892, vii, p. 107. * Lyon, E. P.: This journal, 1902, vii, p. 56. 254 Lda H!. FTyde. susceptible to lack of oxygen, or less resistant to potassium cyanide about ten to fifteen minutes after fertilization, and again about the time of division. From the most susceptible stage, there is a de- crease in susceptibility to lack of oxygen, or increase in resistance to potassium-cyanide up to the time of cleavage. Similar variations had been found by him in the interval between the second and third cleavage. The first cleavage appears in Arbacia about fifty to sixty minutes after fertilization, so the rhythms of susceptibility are about thirty-five to forty-five minutes apart. The inference seems to be, that between the susceptible periods the nuclear membrane is dis- solved and the chromatin more widely distributed in the egg-sub- stance, while after each segmentation the nuclear membrane keeps the chromatin confined and at that time the egg is more susceptible to lack of oxygen. The rhythms of difference of potential may also have some rela- tion to the processes of differentiation which Scott! found existed in the fertilized eggs of Amphitrite. These processes of differen- tiation may be started into activity at certain definite susceptible periods, by mechanical agitation. The periods are at 35 to 50 and at 80 to 100 minutes ‘after the eggs are placed in sea-water, periods in which the polar bodies are extruded and the first segmentation furrow appears. Scott found that eggs subjected to mechanical agitation at these periods develop parthenogenetically in greater numbers than when agitated at any other time. In harmony with these phases of differentiation in the eggsof Amphitrite are also the periods in the starfish egg which are highly susceptible to artificial fertilization. Delage? found these periods to be between the break- ing down of the germinal vesicle and the appearance of the first polar body. It seems that phases of susceptibility to lack of oxygen or presence of potassium cyanide, and phases of differentiation or physical man- ifestations of change in the constitution of different eggs, all point to a relation to certain phases of the rhythms of differences of electrical potential found to exist in the eggs of Fundulus. In this egg the extrusion of the polar body takes place about 15 to 20 minutes after the egg is placed in sea-water. The first segmentation nucleus forms about 15 to 20 minutes after fertilization, and the segmentation furrow appears about 90 to 100 minutes after fertilization. At these 1 ScoTT, J. W.: Biological bulletin, 1903, v, p. 35. 2 DELAGE: Loc. cit. Electrical Potential in Developing Eggs. 255 s intervals of time it is seen from Table IV that the difference of electrical potential is greater at the blastodisc area. From these periods onward the difference of potential decreases, and the current is reversed. At the beginning of these phases of positive differences of potential, when the segmentation nucleus and the segmentation furrows are forming, it is probable that the viscosity or contractility of the protoplasm is increased. Moreover we learnt from Lyon’s paper that during the phases which correspond to those of positive difference of electrical potential, as indicated in Table IV, the Echinoderm egg was least resistant to lack of oxygen and potassium- cyanide, and from Scott’s observations that the Amphitrite ege was most susceptible then to mechanical agitation, since more eggs develop parthenogenetically at these phases than do when agitated at any other time. It is also during these phases that the chromatin is in its anabolic or synthetic stage, building up the complex nucleo- chromatin compound, and during this time it has its greatest affinity for oxygen. Delage on the other hand found that the starfish egg was more susceptible to artificial fertilization between the time of dissolution of the germinal vesicle and the appearance of the polar bodies. During this period the viscosity of the egg’s protoplasm is decreased and probably its irritability greatly increased. A CONSIDERATION OF SOME THEORIES OF SEGMENTATION, AND THEIR. BEARING UPON THE DIFFERENCES OF ELECTRO- MOTIVE FORCE IN EGGs. It would be interesting to ascertain the factors which cause the difference of electrical potential, and to determine their relation to those factors assumed by the prevalent physical theories to effect segmentation of the egg. A consideration of the mechanical and chemical theories which attempt to explain maturation and seg- mentation of the egg, and of the physical phenomena upon which Lippmann’s capillary electrometer is based, as well as important facts made known through the works of Bredig, Ostwald, Hardy, Van’t Hoff and other physico-chemists, regarding colloidal solutions and ion actions, should direct to a better comprehension of the physical manifestations of the processes of cell division, as well as account for the existence of differences of electrical potential in developing eggs. I shall therefore briefly review some of the theories and inves- tigations that bear on this problem. 256 Ida H. Hyde. One group of authors, in which may be placed Van Beneden,! Boveri,” Heidenhain,? and Rabl* considers the astral and spindle rays composed of contractile substances and extending from the cen- trosomes to the periphery of the cell, as the force that directs the processes of segmentation. In the other group are Biitschli,> Rhum- bler,’ Hertwig,’ and others. They support the dynamic theory, and regard the centrosomes as the dynamic or chemical centre that con- trols nuclear and cell division. The radial fibres or astrospheres, they believe, are dynamically induced formations of the centrosomes, or chemical modifications of the proximal protoplasmic elements. They are compared with lines of force, and regarded as visible ex- pressions of the chemico-physical interaction of the central body and plasma. Conkling ® inferred from his study of the segmentation of the egg of Crepidula, and V. Erlanger,’ from his observations on the segment- ing eggs of the Nematodes, that a fluid state of the protoplasm was one of the important conditions for cell division. They noticed at each segmentation period a superficial streaming of the protoplasm from the poles toward the equator of the egg. In his physical expla- nation for cell division, Biitschli!® also assumes the existence of a fluid state of the protoplasm. He is of the opinion that the astral forma- tion arose by tension action, associated with absorption by the cen- trosomes of exudations of nuclear fluid and the dissolved substances which diffused into the nuclear fluid from the plasma. The influence that the astral and spindle fibres exert at the equatorial plane, where the diffused substances from both astral phases meet, he believes must be greater than at any other part of the egg. Moreover he believes that segmentation furrows are produced by increased action 1 VAN BENEDEN: Bulletin de Académie Belgique, 1887, xiv, p. 3. * Boveri: Verhandlungen der physikalisch-medicinischen Gesellschaft zu Wiirtzburg, 1895, p. 29. 8 HEIDENHAIN: Archiv fiir Entwickelungsmechanik, 1895, i, p. 473. 4 RABL: Anatomischer Anzeiger, 1889, iv. 5 BUTSCHLI: Archiv fiir Entwickelungsmechanik, 1900, x, pp. 54-57. 6 RHUMBLER: Archiv fiir Entwickelungsmechanik, 1896 and 1897, iii and iv, p. 659. * HerTwIiG: Abhandlungen der Bayerischen Akademie, 1898, xix. 8 CONKLING: Biological Lectures, Wood’s Hole, 1899. ® V. ERLANGER: Biologisches Centralblatt, 1897, xvii, pp. 152-339. 10 BUTSCHLI: Archiv fiir Entwickelungsmechanik, goo, x, p. 52; Untersu- chungen tiber mikroscopische Schaume und das Protoplasm, 1892. Electrical Potential in Developing Eggs. 257 of surface tension at the equator, due to strengthened activity at the place of the dissolved substances, and to the influence that the asters exert there. It is well known how Biitschli’ attempted to demon- strate the action of the centrosomes by allowing warm gelatine oil- foam, in which spindle radiations had formed, to cool. The effect was a shrinkage of the air in the foam spaces, that in turn exerted a pull on the surrounding mass toward the centre of the spaces. A similar effect, he believed, was produced by the chemical and dy- namical action that exists between the nucleus, centrosomes, and protoplasm, brought about by the absorption of water by the centro- somes from the surrounding mass. The withdrawal of water causes a shrinkage in the mass, and indirectly a pull or tension in the form of spindle fibres among the cytoplasm, at the time of mitosis. Loeb? agrees with Bitschli, Conkling, and Quincke in assuming that the phenomena of cell division are dependent upon protoplasmic stream- ing, which in turn requires a definite degree of viscosity. The vis- cosity is influenced by the presence of electrolytes. Moreover a change in surface tension also produces protoplasmic streaming; while, on the basis of Lippmann’s observations, electrical changes must lead to changes in surface tension. Loeb imagines that diffu- sion currents exist at the surface of the egg which meet at the plane which separates the astral rays. These currents, he believes, lead to whirling movements that are symmetrically arranged to the equatorial plane. These whirling movements normally lead to segmentation producing increased surface tension at the equator, and decreased tension at the poles. When viscosity of the protoplasm is greatly increased through the concentration of the salt solutions, for instance, the diffusion currents and whirling movements are weak and will then not lead to segmentation. Mrs. Andrews’ observations on the Pee substances in the eggs of starfish and sea-urchins are in harmony with those published by Biutschli. Before maturation she saw contractile waves proceed to the dissipation of the nuclear membrane, preparatory to the formation of the egg nucleus. The cytoplasm then becomes more fluid, since the watery contents of the nuclear sac mingles with it. Later marked viscosity of the whole internal cytoplasm increases towards the periphery of the egg, while just before renewed division the mass } BUTSCHLT: Loc. cit. 2 Logs J.: This journal, 1902, vi, p. 432; Archiv fiir Entwickelungsmechanik, i, p. 469. 258 Ida Hl. Ftyde. of the egg would be more relaxed or fluid. During fertilization and cell division were to be seen never-ceasing series of redistribu- tion and reorganization of the foam elements, and rhythms related to the cleavage of each cell. Besides the major rhythms of viscosity, every smallest area of the cytoplasm and nuclear substance has moreover rhythms of its own varying viscosity that bear a constant relation to the phenomena of cell division and karyokinesis, and are chiefly visible as local or astral modifications. In the sea- urchin’s egg segregation of pigment granules at certain intervals of viscosity was suggestive, as correlated with structural preparations of certain areas for special physiological functions. Before each cleav- age, pigment granules were carried along in the flux of the streaming substance toward the line where the split was to take place, and appeared to be carried outward and inward from this point. In addition to all these observations, it must be added that atten- tion was called by Fol,! years ago, to the resemblance of mitotic fig- ures to those in a field of magnetic force, and later Zeigler and also Gallardo? pointed to this resemblance. Gallardo, by a very ingenious method, illustrated how the action of the centrosomes might produce the astral spindle radiations. He placed crystals of quinine sulphate in a vessel of turpentine into which were put two insulated wires connected with the poles of a condenser. The crystais oriented themselves like the lines of force in an electric field, and so were produced radiations at each pole, as well as a connecting spindle. On the basis of the various phenomena that he secured, he thought it possible to refer the force causing the mitotic figures to one ema- nating from the central bodies. He believed that the centrosomes cause radiations during segmentation, and that they act by some power similar to a magnetic or an electric force. Rhumbler,® on the other hand, assumed in his earlier work that an electric force worked in unison with surface tension in cell division, but changed his view when he learned that Roux found that an electric current exercised no definite directive influence on the segmentation of the frog’s egg. Rhumbler? believed cell division possible through the influence of the contractile power of the radial fibres, and an increase 1 Fou: Jenaische Zeitschrift, 1873, vii. 2 GALLARDO: Anales de la Sociedad cientifica Argentina, Buenos-Aires, 1896 and 1897, xlii and xliv. 3 RHUMBLER, L.: Archiv fiir Entwickelungsmechanik, 1896, iii, p. 57, and 1897, ING Bs We ec Electrical Potential in Developing Eggs. 259 of the cell membrane, produced either by growth or by expansion, the latter as a result of the fluid given off from the nuclear sac at the period of dissolution of the nuclear membrane. Roux! was of the opinion that segmentation was produced by elec- trical action, and argued that in this case an electrical current passed through a solution that contained frogs’ eggs would influence the normal process of division. As the result of a strong constant current passed through the solution containing frogs’ eggs, he ob- served a typical wandering of the pigment towards a polarized area of the egg. The outline of the polar area was on the anode side, and it is of great interest that in the blastula each cell shows its special polar field. When unsegmented eggs were subjected to weak cur- rents, slight protrusions, extraovate-like formations appeared at each pole, but they did not possess the polar field. On the other hand, when the eggs were subjected to strong currents, the protrusions disappeared. In Tritons’ eggs, the weak current produced a pig- mented polar area, and the eggs became elongated in the direction of the current, while the current caused fertilized but unsegmented Teleosts’ eggs to bulge out at each pole, so that a furrow was formed in the median area between the protrusions. The polar changes oc- curred at entrance and exit of the eggs, and only when the eggs were moist. Although Roux employed maximal currents for the frogs’ eggs they did not seem to influence normal segmentation. It is questionable whether Roux’s observations gave Rhumbler sufficient or conclusive reasons for changing his first view, and reasons for doubting that an electric force acted in unison with changes in surface tension in effecting the segmentation of eggs. His results are not decisive in proving that electrical currents are not concerned in mitotic division of the nucleus, since such high local tension might be produced in the nucleus that currents which at their entrance would not act destructively on the cell body, might not be strong enough to influence the normal division of the egg. Even the strongest current may, as Roux points out, give negative results, since through its high tension it may have the tendency to include the surface of the conductor, and thus fail to penetrate the cell nucleus. The protrusions or extraovate-like formations of unsegmented eggs acted upon by weak currents, indicate a localized lowering of the 1 Roux, W.: Gesammelte Abhandlungen tiber Entwickelungsmechanik, 1895, i and ii, pp. 545-765. 260 Ida FH. Flyde. surface tension, due to electrical influence, while the disappearance of these, when strong currents are applied, again show a change in surface tension, but this time an increase. These changes of surface tension and concomitant formation and disappearance of extraovate-like structures resemble the action of dilute and con- centrated salt solutions on the development of Echinoderm eggs. In these, the segmentation is interfered with by the action of electro- lytes that produce in dilute solutions extraovate blastomeres, while if these are now placed in concentrated solutions, the extraovate-like formations may disappear, but the normal segmentation of the egg will still be interfered with. I assume that the weak electrolytic solution changes the surface tension of the egg, the osmotic pressure and the state of its protoplasm, through its absorption of water, and the action of the electrically charged ions. The stimulating action of an electrical current is due to polarization, and since polarization in the frog’s egg is well marked, we should have expected that the egg would be stimulated with a suitable strength of current or affected in such a manner that its normal development would be influenced. It is true that the eggs of some animals are resistant to sudden changes of osmotic pressure of the media in which they live. Fundulus, for instance, as Loeb! proved, develops in distilled water as well as in concentrated sea-water, being permeable to both, and it may be that this property interferes with the influence exerted by the electrical current, for Brown? showed that these eggs are polarized with diff- culty, and that the current has no influence on their development, while Arbacia and Asterias’ eggs are susceptible to electrical cur- rents. I believe, therefore, that Rhumbler’s earlier views deserve support, and that an electrical force acts in unison with surface tension in the segmentation of eggs. A review of many of the hypotheses which account for the factors producing segmentation shows that most of them are in harmony in placing in the centrosomes and their radiations the central force for nuclear and cytoplasmic division, and in regarding them as the chief factors in the eggs’ polarity. According to other theories, a change in state, or streaming condition of the protoplasm, and an alteration in surface tension over a localized area, combined with tension action of astral rays, are the principal agents in cell division. Our next inquiry is whether during maturation and division there 1 Loren, J.: Archiv fiir die gesammte Physiologie, 1894, lv, p. 370. 2 Brown, O. H.: This journal, 1903, ix, p. 113. Electrical Potential in Developing Eggs. 261 are evidences of changes in the chemical composition and the reac- tions to stains of any of the constituents of the egg, definitely related to the varying physiological changes that accompany the segmenta- tion of the egg, and which directly or indirectly add strength to the morphological and dynamical theories of segmentation. As regards the function of the nucleus, it is the opinion, since Nussbaum’s ! experiments on infusoria, that the nucleus is the centre for anabolic and formative activity, and Korschelt’s? experiments on the ova of Dytiscus, strengthened the fact that the nucleus was the most influential in synthetic and assimilative processes in the egg. Korschelt compared its influence on the surrounding substance to ferment action. According to Kossel,? Carnoy, and Lebrun,! and others, the cell is composed of three organized proteid groups of substances. The composition of the nuclein group, for instance, varies in accordance with the different conditions of the nucleus, whether it is-at rest or active during the processes of division. The nucleinic acid moreover is found in the nucleus in different states, first, as free acid, second, combined with a base, such as a protamine and albumose, and third in combination with albumin. According to Kossel, the albumose disappears in ripe sperm, and the protamine then combines with the nucleic acid to form nuclein. From the above, and other related statements we conclude that the chemical constitution of the nucleus and the cytoplasm is mod- ified during maturation and cell division, and that these changes are related to definite functions of the constituents of the egg. We may bring these facts into relation with the staining reaction of chromatin, by recording some of the observations made on eggs during the history of the germinal vesicle and the nucleus during segmentation. : Kossel*®*and Lillienfeld® proved that the nuclein series show an affinity for basic dyes in direct proportion to the amount of nucleic acid they contain. From the fact that chromatin in the cell nucleus assumes, during different phases of maturation and division, various 1 Nusssaum: Geschichte fir Natur und Heil-Kunde, Bonn, 1884. * KoRSCHELT: Zodlogisches Jahrbuch-Anatomische Abtheilung, 1891, iv. 3 KossEL: Verhandlungen der physiologischen Gesellschaft, Berlin, 1892-1893, Xvii-xviii, pp. 5-6. * Carnoy and LEBRunN: La cellule, 1899, xii, p. 197; 1885, i, p. 202. ® KosseEL: Archiv fiir Physiologie, 1891 ; 1893, p. 181. § LILLIENFELD: Archiv fiir Physiologie, 1893, p. 395. 262 [da HL. Hyde. shades of blue-green to green, with Ehrlich’s methyl-green stain, and pure green, most intense during mitosis, Heidenhain! concludes that the proportion of nucleic acid and albumin varies with periodic changes in the nucleus, and that during mitosis the chromatin consists largely of nucleic acid. Since free nucleic acid possesses qualities which in combination with albumin disappear, the idea arose that the com- bination and dissociation of these indicate especially during mitosis important physiological processes, related to functions of the nucleus. The changes undergone by the chromatin during the growth of the egg were also studied by Riickert.2 His observations corroborate those mentioned above. He found that at an early stage the chromosomes are small and stain deeply; later they increase enor- mously in size, but lose their staining power. As the egg approaches its full size, the chromosomes decrease in size, but increase again in staining power. The variations in size of the chromosomes point, Riickert believes, to absorption of large quantities of matter, that produce combinations which do not take on the chromosome stain. When this matter leaves the chromosomes, as a result, possibly of katabolic changes, the chromosomes are left with greater staining power. Heconcludes further that increase of surface of chromatin facilitates exchange of material between chromatin and plasma. The high percentage of nucleic acid originally contained in the chromosomes gives to it its staining power, when, in combina- tion with a large amount of albuminous substances, it forms lower members of the nuclein series. The resumption of staining power is caused by a decrease of albumin and restoration of nucleic acid preparatory to division. Greatest activity in cytoplasm coincides with decrease in nucleic acid and staining ability; while suspension of activity corresponds to richness of nucleic acid and greater power of staining. The chromatin passes through cycles in the life of the cell. During vegetative activity, there is an increase of albumin and a decrease in the reproductive phase of activity. Mrs. Andrews also observed in the starfish’s egg that at different stages in the changing optical and physical conditions the astral rays show differences of reactions to chemicals as characteristic of their substance in its varying states of viscosity or contraction. A few seconds may make a difference in reaction of these structures to hardening fluids. 1 HEIDENHAIN: Archiv fiir mikroscopische Anatomie, 1894, xliii, p. 423. * RUCKERT: Anatomischer Anzeiger, 1892, iv, p. 107. Electrical Potential in Developing Eggs. 263 Other writers have advanced chemical theories. Strassburger for instance, assumes that the movements of the chromosomes may be chemotropic, and Carnoy believes that the asters are formed under the influences of specific ferments. PHysicAL CONDITIONS THAT INFLUENCE CLEAVAGE. From the theories and facts which have been cited above, it would appear that the essential conditions for cleavage are an increase of surface area over a definite region of the egg ; a flowing state of the cytoplasm; tension action of the astral radiations ; directive force of the centrosomes ; and metabolic control over the cell by the nuclear elements ; moreover it would seem that these factors are related to chemical changes especially in the chromatin elements of the egg manifested at definite phases of mitosis by reactions to special stains. Chemical changes are, however, not necessarily the fundamental cause of the physical alteration of the surface extent of the egg, or of the phase of its cytoplasmic constitutents during segmentation. It becomes necessary therefore to seek for the factors that will explain ‘these conditions. It is well known that the potential energy present at the surface of a liquid tending to occupy a smaller volume, produces a tension that is opposed by a force which tends to increase the size of the liquid surface. Of the forces that may alter surface tension, the first in importance is undoubtedly the energy manifested by the difference in osmotic pressure of electrolytes that may exist on the sides of the limiting surface. The effect of electrolytes on surface tension is easily demonstrated in living or artificially produced cells with thin semiper- meable membranes. The enlargement of the cell by entrance of fluid that effects dissociation of contained electrolytes and increase of their solution tension, indicates a lowering of surface tension of the cell. It is at present an established fact that the chief physiological effect of an electrolytic solution depends upon the number of its dissociated ions through the charges they bear. The works of Bugarszky and Tangl,! Hardy,” Bredig,? Loeb,! Mathews,” and others 1 BuGARSZKy and TANGL: Archiv fiir die gesammte Physiologie, 1898, 1xxi, p. 467. 2 Harpy: Proceedings of the Royal Society, tgoo, Ixvi, p. I11. 3 BreDIG: Anorganische Fermente, Igol, p. 16. 4 LoeB: This journal, 1902, vi, p. 430. 5 MATHEWS: This journal, 1904, x, p. 431. 264 Ida H. Hyde. have proved. that metals are most active in the ionic state, and that the ions act chiefly by means of their electrical charges. Mathews believes also that the effectiveness of the ions is determined by the affinity that the ion has for its charge. The increase of surface over a definite region of the egg in cleavage, due to lowering of surface tension, can also be brought about by placing unfertilized eggs, ¢.g., those of the starfish, for a time in hypoisotonic salt solutions, in which the eggs increase in size and form extraovates or dumb-bell shaped forms. On the other hand, the cleavage of the cytoplasm, and increase in surface extent, does not appear if starfishes’ eggs, or those of Scypho- medusze, and other forms are placed in strong hyperisotonic solutions. It may be said that in the former case, in the more dilute solution, the dissociations of electrolytes into ions is more complete, produc- ing an increase of electrical charges, sets free more thermoelectric energy, and also increases the stability of the colloids in the cyto- plasm. These conditions tend to lower the surface tension, and pro- duce the enlargement of the egg. In the latter case the surface tension is increased, not diminished, because the number of electric charges acting counter to the surface tension are decreased and the cytoplasm or hydrosol is rendered thereby more unstable. But a more satisfactory explanation may be drawn from researches con- ducted by Lippmann and Helmholtz,! who have demonstrated that surface tension at the plane of separation between a metal, ¢.g., mer- cury, and an electrolyte is a function of the difference of potential of this plane, and that the force with which the surface tends to dim- inish itself is determined by the natural surface tension of the mercury diminished by the force with which the electric charges of the ions tend to increase the surface. If the metal is positive it car- ries positive charges, and this surface of charges lowers surface tension, in that the like ions repel each other and thus through their repelling force increase the surface area. The surface tension and difference of potential are altered by the addition or subtraction of electrical charges, and the potential between metal and electrolyte depends upon the solution tension of the metal and the osmotic pres- sure of the ions. .The cause of difference .of potential between a metal and a solution we have learned is on the one hand, the solution tension of the metal, tending to drive ions from the metal into the 1 OstTWALD: Lehrbuch der allgemeinen Chemie, 1893, ii, p. 922. Electrical Potential in Developing Eggs. 265 solution, and the osmotic pressure of the solution, acting counter to this, tending to cause the kations already present to separate on the electrode in the metallic condition. This explanation leads to the inference that the charges accu- mulating on each side of the semipermeable egg membrane during different phases of mitosis are the active agents in lowering the sur- face tension and responsible for the difference of electrical potential. Their relative proportion and signs depend upon the strength, re- actions and nature of the electrolytes that are separated by the membrane. Their presence at the periphery, the streaming state of the proto- plasm, as well as the formation of the segmentation furrow next de- serve our consideration. Biitschli believes that segmentation is due to a lowering of surface tension at the region nearest the asters. These structures, he maintains, repel each other, because they bear electrical charges of like sign. They are acid poles that repel ions of like sign centrifugally ; this lowers surface tension over this special area, while at the same time the interaction of the ionic charges produces a streaming of the protoplasmic contents. It was observed by Mrs. Andrews, Conkling, Erlanger, and others that protoplasmic stream- ing may be distinctly seen during mitosis. The paths of movements are from the periphery toward the equatorial plane. In addition to these active manifestations, Mrs. Andrews noticed rhythms of inces- sant changes of viscosity in the protoplasm, the dissolution of the nuclear membrane and the liquid contents of the nuclear sac diffus- ing into and increasing the dilution of the cytoplasmic fluid. The question is, what are the physical explanations for these phenomena, and what is their significance? According to Hardy! and Bredig? protoplasm is a colloidal solu- tion consisting of solid particles in a fluid, the particles floating in a dilute solution of the colloid and water and possessing a double electrical layer at the limit between the suspended particles and the fluid. The viscous solution surrounding the particles resists their movements through the fluid ; moreover, they are kept apart by elec- trical charges of the same sign which they carry, the charge being of opposite sign to that of the surrounding medium. The colloidal particles may carry either positive or negative charges depending TSECAR DV as LOG (627: 2 BREDIG: Loc. cit. 266 Ida Ff. Flyde. upon the reaction of the surrounding medium, that is, the proteid particles and the fluid form an electrically homogeneous mass when the fluid is neutral. On the other hand, the proteid particles have electrical characters conferred upon them by the nature of the re- action of the fluid. The proteid particles seem to act as basic or acid particles, according to the circumstances in which they find themselves. Under the influence of an electrical current, they go to the kathode in an acid, or to the anode in an alkaline medium. If their charges are taken from them either by oppositely charged ions or an electric current, their physical nature is altered; they are con- verted from hydrosols into hydrogels. The stability of the system may also be destroyed by induction, the active agents being free ions carrying static charges. In this case the action is on the electric layers immediately around them, and the active ions are those whose electric sign is the opposite to that of the charge on the surface of the particles. By robbing them of their charges they reduce their surface and produce protoplasmic movements. Consequently salts and alkalies, as well as acids, if present in a solution with suspended colloidal particles may act on these, the result being decomposition of complex molecules. Anabolic as well as katabolic changes may thus occur as a result of electrical charges which, even in small quan- tities may produce great effects, for as Bodlander has shown, one part of hydrochloric acid in 1500000 parts water was effective in producing marked precipitating changes in suspensions of different substances, resembling catalytic action. The change in state and streaming of protoplasm is accordingly regarded as due to the interaction of the electrical elements. That is, the ions either rob compounds of their charges, or are deprived of theirs by the compounds, so that new affinities are developed and new relations produced among the fluid and solid constituents of the ege’s contents. Besides the difference in surface tension that obtains among the particles of the egg protoplasm, there must exist a difference of electrical potential and surface tension, not only between the surfaces of nuclearplasm and cell protoplasm, separated by the nuclear membrane, but also between the surface of the egg and its surrounding medium. The semipermeable membrane sepa- rating electrolytic solutions may be compared to the two parts of a Franklin condenser separating on the one side positive charged ions from the negative on the other, and thus surface tension comes into play. The nature of the membrane and its resistance to influences Electrical Potential in Developing Eggs. 267 may be affected by its surrounding electrolytes, and therefore changes in solution tension of its own component particles may arise, and secondarily, have an effect upon the surface tension as well as on the difference of electrical potential of the egg. We are led to infer that the chemical and physical energies mani- fested among the mass of heterogeneous ions, atoms, and compounds in the egg’s protoplasm and its surrounding fluids produce the vital phenomena of development. At the time of maturity a certain equilibrium of forces has been established. With the entrance of the spermatozo6n, the existing state of equilibrium is upset. The forces between the electrolytes and colloids receive a new impetus that starts activities which lead to segmentation. The combined male and female nucleus is the pole that exerts the directive force, representing, as it were, in its function an automatic centre which is kept active by the interaction of its own ions and those of its medium. It may therefore effect oxidation and reduction changes or synthetic and analytical processes. We know from chemical analysis and microchemical tests made by Kossel, Riickert, Lebrun, and others that the chromatin molecule decomposes during its active phase, setting free nucleic acid and albumins, and during the resting stage, the nucleic acid, albuminous and protamine groups combine again with each other. During the katabolic acid stage the chromatin particles therefore carry charges whose signs are opposite to those carried by them during the neutral or alkaline anabolic phase. The chromatin particles would move therefore under the influence of an electric current to different poles, wandering with its negative charge in an alkaline fluid during the acid phase to the anode and to the kathode during the alkaline phase. Inductive action would also exert its effect, producing a change in the state of the charged ions by acting on the electrically charged double layer, causing a repulsion or attraction between them and the acid chromatin particles. I would agree with Biitschli, Lillie, and others, that during the acid phase the centrosomes resemble highly concentrated electrostatically charged poles that are not in a field of force only, but also in an electrolytic field repelling the ions of like sign toward the equator, while attracting those of unlike sign. At the periphery of the egg there would exist therefore an electrical double layer of ions bearing negative charges inside of the membrane and positive charges out- side. In the meantime the interaction of the electric charges of the 268 lda Hf. Flyde. ions forming the acid pole with those attracted by them carrying charges of opposite sign, accompanied by the absorption of fluids, produces a change in the physiological function of the polar chroma- tin elements. At the same time the ions that were repelled toward the equatorial plane meet in an electrolytic medium of definite osmotic pressure and it is possible that new combinations are here also formed. As a result of the change progressing at the poles and equatorial plane, chromatin elements of each half of the plane not only repel each other, but are also attracted toward the astral poles; the latter in consequence of their physical and chemical changes, have reversed their electric surface charges. At the equatorial plane, there is produced therefore a field of stress that greatly increases the surface tension over a definite area of the egg, and thus instigates the formation of the segmentation furrow. The complex forces that have been developed, produce at first dissolution of the nuclear membrane and a less viscous and more irritable state of the cytoplasm; moreover, some of the colloidal elements of special composition arrange themselves in the field of force as precipitated fibril strands, The formation of the astral radiations are due, accord- ing to Harvey, to a pre-existing stress that fashions the foam-network of the cytoplasm, so that they are a coarse diagram of a dynamic phase of the cell’s history. He bases his statement on the results that he and also Biitschli obtained from albumin and soap films that were subjected to a stress. By fixing the film in alcohol the threads of coagulated albumin could be distinctly seen as fibres radiating from the centre of application of stress. We should expect that the changes in surface tension produced indirectly by the interaction of the centrosomes and other elements of the egg would be accompanied by changes in differences of electrical potential between the surface of the egg and its external medium, and in mesoblastic eggs between the vegetative and animal poles. The difference of electrical potential, if present in a measurable degree, should be detected therefore with a suitable galvanometer, as the different stages of mitosis and the changes in intensities of the acid and alkaline phases make their appearances. The observations on the differences of electrical potential secured from the turtle’s egg during maturation, and from Fundulus eggs after fertilization, prove that the phases of mitosis are accompanied by changes both in the character and the number of the surface charges born by the egg. During the maturation period the surface Electrical Potential in Developing Eggs. 269 of the blastodisc in the turtle’s egg is at a higher potential than any other part of the egg’s surface. Moreover, one pole of the blastodisc near the nucleus has a greater difference of electrical potential than has the opposite pole. The sign of the surface charge is the same as that existing on the surface of the blastodisc of the Fundulus egg at about twenty minutes after fertilization, and again when the seg- mentation furrow is forming. When the egg is placed in the electro- meter circuit at these periods, a current is indicated as passing from the animal pole of the egg toward the vegetative one. During the time intervening between these periods the surface bears charges of opposite signs and the direction of the current in the electrometer and the egg is reversed. According to the argument advanced above, the chromatin during this interval of time is undergoing ‘decomposition and is progressing in its katabolic acid stage. In this state it attracts oppositely charged ions and repels those carry- ing charges of like or negative sign to the periphery, thereby lowering its own surface tension, while the preponderance of these charged ions, by inductive action, attract to the outer surface of the egg electrolytes that carry positive charges. If these statements are in accordance with the facts then we should expect that if the egg were placed in contact with the non-polarizable electrodes of the electrometer during the acid phase of the chromatin, the mercury in the electrometer, which, in contact with the sulphuric acid, is positively charged, would receive additional positive charges from the surface of the egg. These charges would increase its original charge and consequently decrease the surface tension of the mercury, causing it to drop from the capillary tube of the electrometer, thus indicating a current that travels in the circuit through the egg from the vegetative toward the animal pole. The greater the number of charges on the surface of the egg, the lower is its surface tension, and the greater is the extent of the movement of the mercury in the capillary. On the other hand, when the surface charge of the egg is negative then the mercury receives negative charges which decrease its surface charges and increase its surface tension, so that now the mercury rises in the capillary, indicating a current in the egg, if it is placed in the circuit from the animal toward the vegetative pole of the egg. From Table I and also Table IV, it is seen that this is just what occurred. If this is true, then we may say that the change in differences of electrical potential is a reliable index of the changes in surface tension, streaming state of the protoplasm, and other 270 Ida F1. Flyde. special chemical and physical alterations going on in the egg during its mitosis. The difference of electrical potential is difficult, if at all possible, of detection in eggs that possess a thickened egg membrane, or one complicated by accessory structures or secretions that could interfere with the ready interchange of ions between the interior and exterior of the egg. The interchange of electrolytes is also impeded by the more viscous, less irritable state of the egg’s protoplasm, due to absorption of the cytoplasmic fluid, or to diffusion produced by a high external osmotic pressuré. The conditions that control the segmentation of the egg may be determined by mass action or by the interaction of enzymes, both of which are dependent upon the electrical or physical character of the complex molecules or com- pounds that are prepotent at any given time in the egg. Whether two substances combine or remain inactive at a definite temperature, it is well known, depends upon the concentration of the substance. It is possible that at a certain concentration they are completely dissociated, while at any other only partly so. According to the law of Guldberg and Waage, the tendency and rapidity for chemical reaction of a substance is determined by its mass action, that is, by the concentration of the reacting substance, or the substance acted upon, or the reaction product. The greater the amount of reaction substance that. has been formed, the greater is the tendency to reverse the action. It has also been demonstrated that the reaction is re- versible in all cases in which the mass action is discernible, that is, in which the reaction is incomplete, in contradistinction with the usual complete reaction, as, for instance, in saponification of an ester by an alkali, and in certain catalytic actions, e. g., the catalysis of ethy] acetate, water, and acid which is opposed by the catalysis of a mixture of equimolecular quantities of the reaction products. The reversible action of enzymes also depends upon the presence of the reaction products. This fact has been demonstrated, since Croft Hill’s experiments, by others, and lately by Loevenhart and Kastle.’ The behavior of the reversed chemical and catalytic actions may perhaps be inferred from the behavior of the secondary or reversed movements of suspended proteid particles that were subjected to an electrical current. They were, as Hardy? observed, negative to the water, and became positive to it after a longer or shorter sojourn in the neighborhood of the anode. This change may be referred to the 1 LOEVENHART: This journal, 1902, vi, p. 331- 2? HarbDy: Journal of physiology, 1899, xxiv, p. 288. Electrical Potential in Developing Eggs. Py i electrolytic action of the current upon the electrolytes present. Owing to this, an electronegative particle which moved to the anode would there reach an acid region in which its character would change from electronegative to electropositive. This view is supported by the observation that the reverse movement is more pronounced, the higher the concentration of the electrolytes, and therefore the greater the concentration of acid or alkali at the poles. We assume that the molecules of a substance in solution exist under a definite pressure, and that every substance will pass into solution until the osmotic partial pressure of the molecules in the solution is equal to the solution tension of the substance. If a metal is put into one of its salts, the solution tension may be greater than the osmotic pressure, and metallic ions with their positive charges will pass into solution, which then becomes positive, the metal negative. At the plane of contact is the electric double layer. The positive charged ion in the solution and the negative charged metal attract one another and develop a difference of potential. The solution tension tends to force more ions in solution, while the electrostatic attraction of the double layer is opposed to this. When these forces are equal, the equilibrium is established. If the solution tension is less than the osmotic pressure of the metallic ions in solution, then the ions separate from the solution. Palmaer! has demonstrated this by placing mercurous nitrate in a vessel whose bottom is covered with metallic mercury. On account of the low solution tension of the mercury, some mercury ions from the solution will give up their positive charges to the mercury, which in turn will attract electrostatically negative NO, ions to form the double layer. This will continue until a certain difference of poten- tial has been reached, when equilibrium will be established. If adrop of mercury now falls into the solution, some mercury ions will sep- arate upon it, charge it positively, and it will attract some negative NO, ions, and drag them down to the mercury. But when the drop unites with the mercury at the bottom, it will contain an excess of positive electricity, and the same number of mercury ions will pass into solution as negative NO, ions were carried down from the top. The solution tension and osmotic pressure represent two forces that are ready to act at the moment the one or the other obtains the ascendency. The solution tension and the osmotic pressure of metals or complex compounds in a solution may be influenced by an elec- * PALMAER: Zeitschrift fiir physiologische Chemie, 1899, xxviii, p. 287. 272 | lda Hl. Hyde. trical current, temperature, nature of the electrolytes present, and by the absorption or diffusion of liquids. It is probable that during the development of the egg, processes of mass action, enzyme or catalytic reactions, and differences of elec- tromotive force, all play an active part. They are controlled by the solution tension, and the osmotic pressure of the colloidal and elec- trolytic constituents of the egg. Of these elements, those of the chromatin group arethe most influential. A certain state of equilibrium is reached when the ovum is ripe. When the sperm enters, it in- troduces definite ions that bring about an unstable condition and start new reactions that tend toward a new state of equilibrium. The combinations which the chromatin elements form depend upon the concentration of the electrolytes, the presence of reaction sub- stances or reaction products, and enzymes in the cytoplasm. They carry charges of definite signs and number, and during certain phases in the development of the egg pass through reversed action periods characterized as anabolic or neutral, and katabolic or acid phases. The chromatin particles are the most prepotent, and exert, through the action of their charges, the directive force in the phe- nomena of segmentation. The views expressed in this paper regarding the physical factors that influence mitosis are more or less held by others, especially Biitschli, Gallardo, Loeb, and R. S. Lillie! When this work was undertaken, I had the idea that segmentation was the result of special physical reactions herein mentioned. I communicated my views during the summer of 1902, while at work at Wood’s Hole, to Dr. F. R. Lillie and Dr. E. B. Wilson. Though these views are no longer entirely new, I publish them in their imperfect form with the results of my work, hoping that additional facts will be forthcoming that will establish the theory of segmentation more satisfactorily. CONCLUSIONS. In considering what has been said above in regard to surface ten- sion and its effect upon mitosis, I conclude that the segmentation and the normal course in the development of the Scyphomedusa egg, placed in strong salt solutions, did not proceed because of the changes instituted in the egg by the more concentrated medium. The latter must have produced a difference in osmotic pressure between the 1 LILLIE, R. S.: Biological bulletin, 1903, iv, p. 175. Electrical Potential in Developing Eggs. ay: salts in the egg and those in the water surrounding it, a withdrawal of water from the egg’s contents, a lowered dissociation of ions, and an increase in the viscosity of the cytoplasm. Possibly in the more concentrated state of the electrolytic solution in the egg, certain colloid elements are precipitated. This is assumed from the in- vestigations conducted especially by Schulze,! Linder and Picton,” Hardy,’ Bredig, and Pauli* who have studied the precipitation of col- loids by electrolytes. Therefore, in consequence of the concentrated medium, there occurred a decrease in the surface tension of colloids in the egg, and an increase in the surface tension of the egg itself. These phenomena would depend upon electrical action of ions that greatly alter the surface tension between colloidal particles and their fluid medium, as well as the differences of electrical potential that exist between the colloids and their surrounding electrolytic solution in the egg on the one hand, and on the other between that of the surface of the egg and its external medium. This statement is in accordance with Bredig’s views which are based on the investigations of Quincke, Coehn, and others. They believe that suspensions and colloidal hydrosols possess electrical differences of potential toward the surrounding medium, and from the works of Lippmann, Ostwald, Rothmund, and others that the surface tension of two contiguous mediums is a function of their potential difference and that this, as well as the surface tension, may be greatly altered by the addition of certain ions. In the maturating turtle’s egg, a difference of electrical potential which increases as development progresses exists between the animal and vegetative poles, and also between the poles of the longer axis in the blastodisc. This fixed polarity is related to a definite axis of the egg as well as to one in the blastodisc. Differences in electrical potential exist between the animal and vegetative pole in the fertilized egg of Fundulus. During the phases of segmentation these potential differences appear in periods of rhythmical sequence that are characterized by currents flowing for a definite time in one direction, gradually increasing to a certain limit and then decreasing, followed by a reversal of the current, which also increases gradually and then decreases. About twenty minutes after SCHULZE: Journal fur praktische Chemie, 1883, xxvii, p. 320. LINDER and PicTon: Journal of the Chemical Society, 1895, Ixvii, p. 63. Harpy, W. B.: Proceedings of the Royal Society, 1900, Ixvi, p. 115 ; Journal of physiology, 1899, xxiv, pp. 158-297. 4 PauLi: Archiv fiir die gesammte Physiologie, 1899, Ixxviii, p. 315. 1 2 3 274 [da FI. Alyde. fertilization, and at the time when the segmentation furrow appears the electrical potential is greater at the animal pole, and at a cer- tain time between these intervals it is least over this region. The periodic variations of electromotive force bear a definite relation to the resting alkaline anabolic, and active acid katabolic phases of the chromatin mass; in other words, to the different phases that are discernible during mitosis. They may also bear a relation to the rhythms of viscosity, resistance to pressure, and fixatives, as observed by Mrs. Andrews in the living protoplasm of the starfish egg, to the periods of resistance to oxygen and potassium cyanide, the produc- tion of carbon dioxide, ascertained by Lyon to obtain in the eggs of Arbacia, and to the intervals susceptible to mechanical agitation found by Scott to exist in the Amphitrite egg, moreover to those phases susceptible to artificial fertilization ascertained by Delage to occur during segmentation in Arbacia eggs. The difference of elec- trical potential registered in the electrometer, when the egg is in the electrometer circuit, indicates that during definite intervals, as mito- sis progresses, the surface of the egg carries charges varying in number and sign. They exist on the external surface of the egg as a result of electrostatic force, and bound one side of the Helmholtz electric double layer which exists between two fluid media, external and internal to the membrane. The electrically charged ions internal and external to the bounding surface are of opposite sign. The forces that produce these surface charges emanate from the centro- somes, and as a result of the interaction of electrical or physical ele- ments of the chromatin, and also between it and the constituent particles of the surrounding colloidal and electrolytic solutions. Metabolic activities accompanied by the decomposition of complex molecules,.or the construction of these, produce during the katabolic phase an acid chromatin mass that liberates and attracts charged ions. As a consequence of the interaction of ionic forces in the egg and its environment, chemical and physical changes result in the constitution of the egg’s protoplasm. It is possible that changes in osmotic pressure, solution tension, mass, and enzyme action, as well as surface tension, obtain during the egg’s history. These produce alterations in the viscosity of the cytoplasm, absorption of liquid by the chromatin or cytoplasmic particles, and a shrinkage in the one or the other during definite phases of cleavage, the index for the variations of changes being the difference of electrical potential. Electrical Potential in Developing Eggs. 275 When a certain state of equilibrium is attained by the chromatin mass, a reaction started between it and its products sets in. It takes the form on the one hand of synthetic, on the other, of analytical processes ; this is accomplished either by the building up of more complex chromatin molecules of a definite chemical reaction, or by the dissolution of the complex to a more simple compound of opposite chemical reaction. These alterations and interactions of ions and their associated energy elements are accompanied by physical and chemical changes of the egg’s substances. These changes give ex- pression to their dynamical or tension action in the form of astral and spindle radiations, dissolution and degeneration of the nuclear mem- brane, and chromatin molecules, and to the formation of cleavage furrows. THE AUTOLYSIS OF ANIMAL ORGANS.—II. HYDROLY-— SIS ‘OF FRESH AND: SELF=DIGESTED GLa: BY «By aA. LEVENE. [from the Department of Physiological Chemistry, Pathological Institute of the New York State Hospitals.) N hydrolysis of proteid material by mineral acids various crystalline products result. The same or nearly the same products appear on digestion of proteids by proteolytic enzymes of the gastric and pancreatic glands, and nearly the same substances on prolonged autolysis of animal glands and tissues. On the ground of these findings a general conclusion was drawn that most tissues contain an enzyme acting insthe same manner as trypsin. Upon closer observation, however, it was noted that among the end- products of autolysis of animal tissues, substances occur which could not be obtained by cleavage of the tissue constituents by means of mineral acids, while some substances usually appearing on hydrolysis by means of mineral acids are found in small quantities, or not at all, among the products of autolysis. Thus Emerson! found, on self-digestion of the pancreas, oxyphenylethylamin and cadaverin. Kutscher could find only one hexon base — lysin on the autolysis of the thymus; Dakin isolated only one purin base after prolonged autolysis of the kidney. Lawrow? found, on self-digestion of the gastric wall, putrescin and cadaverin instead of lysin and arginin. The end-products of the autolysis of the liver contain only one hexon base,—lysin; and those of the testes, only one basic sub- stance, — hypoxanthine. Several explanations of this difference are possible: the enzymes of the various organs may be unable to cause a complete cleavage of their tissues, in which case the products of self-digestion, if heated with strong mineral acids, should form the substances which failed to appear on autolysis. The other explanation may be that the same 1 EMERSON: Hofmeister’s Beitrage, 1901, vi, p. 501. * Lawrow: Zeitschrift fiir physiologische Chemie, 1901, xxxiii, p. 312. 276 The Autolysts of Animal Organs. 277 substances are formed during autolysis as on cleavage with mineral acids, but the substances undergo further decomposition, perhaps through the action of other than proteolytic enzymes. It is possible also that all the end-products suffer a quantitative change, either be- ing decomposed or synthetized into more complex substances. That surviving tissues are capable of this double function was demonstrated by Wiener in regard to uric acid. The difference may also be explained by the supposition that tissues contain, besides the proteids already isolated and analyzed, substances which as yet have not been studied. Thus, it seemed important in order to investigate the nature of the chemical reactions occurring in surviving tissues to compare the end- products of autolysis not with cleavage products of pure chemical substance, as individual proteids or nucleins, but with the cleavage products of the entire organ or tissue. This was the plan of the present work. The investigation, as yet, is not completed, and the results obtained thus far are communicated at present partly for the reason that while the work already was in progress, the very important investigations of Kossel! on Argynase, and of Walter Jones? on the Self-digestion of Nucleo-proteids have appeared. Their observations coincide with those made in course of this work and explain some of the findings. The great part of tissues, and of protoplasm in general, consists of simple and of combined proteids. Proteid at large is a very complex molecule containing radicles with very different chemical properties. From the physiological point of view they all may be classified in the following groups: (1) nitrogenous acids (monoaminoacids) ; (2) urea group (arginin, lysatinin) ; (3) uric acid group (purin and pyrimidin bases); (4) chromogenic group (pyrol-derivatives); (5) sulphur group (cystin); (6) basic group (ammonia, lysin, ornithin) ; and (7) carbohydrate group. ; The present communication represents only the results of the analysis of the basic constituents of the pancreas, spleen, and liver. The results of similar investigations of muscle will follow soon. The analysis of the amino-acids is in progress also. 1 KossEL: Zeitschrift fiir physiologische Chemie, 1go4, xli, p 321; 1904, xlii, Pp: hSt. 2 JonES: /ézd., 1904, xli, p. 101; 1904, xlii, p. 38. 278 P. A. Levene. PANCREAS, Purin bases.—The literature on the subject of transformation of the nuclein bases (purin and pyrimidin bases) in course of autolysis is reviewed by Jones in his recent publication, and therefore it will suffice to communicate here the results of this investigation. Four bases of the purin group had been identified in this gland; namely, adenin, guanin, xanthin, and hypoxanthin. In the analysis of the fresh glands guanin was found in pre- dominating quantity, adenin followed, and xanthin and hypoxanthin were not detected. If present, their quantity surely was very insignificant. In the digested gland, on the contrary, little guanin was found. Adenin was totally absent, while xanthin and hypoxanthin were present in very appreciable quantity. The total sum of purin bases was markedly diminished in the digested glands. This observation is in harmony with that of other observers. Thus, Kutscher analyzing the purin bases of self-digested pancreas observed that the usual four bases were present among the products of digestion, though seemingly in smaller quantity than from the fresh gland. Burian and Hall noted the transformation of imid-purins into oxypurins in the pancreas. The difference in the results of the present experiment and of that of Kutscher may be explained by the fact that extracts were employed in the older work, while the entire gland was used in the present investigation. Experimental part.—- Five pounds of fresh glands were prepared in the usual manner, taken up in a 5 per cent solution of sulphuric acid and heated 12 hours, with return condenser, in a boiling water-bath. The product was cooled, filtered, treated with Hopkins’ mercuric sulphate solution, and allowed to stand twenty-four hours. The precipitate was filtered on a suction funnel and washed with a 5 per cent solution of sulphuric acid. The precipitate was suspended in water and decomposed by sulphuretted hydrogen, The filtrate from mercuric sulphide was concentrated to remove the hydrogen sulphide, neutralized with ammonia and concentrated to a small volume. A precipitate formed consisting largely of guanin. It redissolved in boiling diluted sulphuric acid. The solution was decolorized by means of charcoal, filtered, and allowed to stand over night. Guanin sulphate crystallized. The sulphate was redissolved in_ boiling diluted sulphuric acid, transformed into the free base by means of The Autolysis of Animal Organs. 279 ammonia, and filtered while hot. The free base was again trans- formed into the sulphate, and that again was converted into the free base and analyzed. 0.116 gm. of the substance was used for a Kjeldahl’s nitrogen estimation. 36.50 c.c. of 4 H,SO, were required. WoneC HN-O: Calculated : Found: N = 46.36 per cent. 45.95 per cent. The filtrate from the crude guanin and all the mother liquids of guanin sulphate were combined, and the purin bases contained in them were removed by means of an ammoniacal solution of silver chloride. This precipitate was filtered on suction and funnel, and decomposed by means of hydrochloric acid. The filtrate from the silver chloride was treated with a saturated solution of sodium picrate. The picrate was filtered while the mother liquid still remained warm. It was then recrystallized out of water, filtered on suction, washed with alcohol and ether, dried in toluo] bath, and analyzed. 0.1170 gm, of the substance gave on combustion 0.1550 gm. CO, and 0.0247 gm. H,O. : Bor C-HN-C.H,(NO,).Or- Calculated: Found: C = 36.26 per cent. 36.12 per cent. ro 219 2.34 S From the filtrate of adenin picrate, the picric acid was removed by means of sulphuric acid and ether. The purin bases remaining in solution were removed by means of an ammoniacal solution of silver chloride. The precipitate obtained in this manner was very small, and it was considered best to continue the further separation of the bases by Kossel’s older method. The salt insoluble in cold nitric acid of sp. gr. 1.1 did not have the composition of hypoxanthin silver nitrate. 0.1472 gm. gave 0.0495 of Ag. Por CLE NiO; AgNO) eFor C/ELN.O, AgNO;: Calculated : Found: Ag = 35.29 per cent. 33.65 per cent. 33.70 per cent. The mother liquid of the insoluble silver nitrate was rendered alkaline by means of ammonia, and only a very small precipitate formed. This precipitate might have been xanthin. 280 P. A. Levene. In a second experiment 10 lbs. of the gland were treated in the same manner. Guanin and adenin were obtained. The filtrate from adenium picrate was acidulated with hydrochloric acid, and allowed to stand over night. A small crystalline deposit was found. This was dissolved again in hot water, cooled, and acidulated with nitric acid, and shaken in separatory funnel with ether. The nitric acid solution, free from picric acid, was allowed to stand over night. Again a crystalline deposit giving a positive xanthin test formed. Therefore the substance might have been xanthin. The remaining mother liquid of adenin picrate was acidulated with hydrochloric acid, and extracted with ether. When all the picric acid was removed, the solution was concentrated under diminished pressure, rendered alkaline by means of ammonia, and treated with an ammoniacal solution of silver chloride. The precipitate thus formed was filtered on suction funnel. The silver salts were decom- posed by means of hydrochloric acid. The excess of acid was re- moved by repeated evaporation of the solution under diminished pressure. The residue was twice taken up with alcohol and evapo- rated to dryness. The residue finally was treated with warm distilled water. The insoluble part did not give the nitrate characteristic of xanthin. The soluble part was treated with picric acid. A slight precipitate resembling adenin picrate formed. On standing, no other precipitates formed. Thus, if hypoxanthin was present, its quantity was very small. Digested glands. — 5 lbs. of the glands were taken up in 0.25 per cent sodium carbonate solution and allowed to digest ten weeks, toluol and chloroform being added to prevent bacterial growth. The products of digestion were concentrated to the original weight and decomposed with a 5 per cent solution of sulphuric acid in the same manner as with the fresh glands. The purin bases were obtained also by the same process as was employed with the fresh glands. The precipitate forming on concentration with ammonia gave a very small yield of guanin sulphate. This was transformed into the free base. It had the following composition. 0.0765 gm. of the substance gave 31 c.c. of nitrogen (over 50 per cent KOH rat’ — 24-5. C..and p= 760 m. For CH NEG: Calculated : Found: N = 46.36 per cent. 46.71 per cent. The Autolysts of Animal Organs. 281 The mother liquid of guanin sulphate-was rendered alkaline with ammonia, acidulated with acetic acid, and filtered while hot. On cooling, a precipitate containing 39 per cent of nitrogen formed. It was dissolved in normal sodium hydrate solution. On addition of an excess of nitric acid (2 parts of acid to 3 parts of water), a precipitate formed. The precipitate was filtered on suction funnel, dissolved in ammonia water, precipitated by acetic acid, washed with alcohol and ether, and dried in xylol bath. It had the following composition : 0.1310 gm. of the substance gave on combustion 42.5 c.c. of nitrogen (over 50 per cent KOH) at t*?=26° C. and p=760 m. or C.F N,O;: Calculated : Found: N = 36.84 per cent. 36.98 per cent. The mother liquid from the crude guanin was treated with an ammoniacal solution of silver chloride. The silver salts thus ob- tained were decomposed with hydrochloric acid, and the filtrate from silver chloride was treated with a saturated solution of sodium picrate. No precipitate formed immediately. On standing forty-eight hours, a heavy crystalline precipitate was deposited. It was once re- crystallized out of hot water, washed with alcohol and ether, dried in toluol bath, and analyzed. 0.1280 gm. of the substance gave on combustion 30.0 c.c. of mirmorven at t°—28° C. and p= 762. For C.f1,N,OC,H,(NO,)OH: Calculated : Found: N = 26.85 per cent. 26.71 per cent. The mother liquid of this precipitate was acidulated with hydro- chloric acid, the picric acid extracted with ether, and the solution rendered alkaline by means of ammonia. The purin bases still present were precipitated by means of an ammoniacal solution of silver chloride. A very small gelatinous precipitate formed. It was treated by the method of Kossel. The part insoluble in nitric acid of sp. gr. 1.1 had the following composition : 0.3030 gm. of the substance gave 0.1060 gm. of Ag. For €-HN,O; Ag NO3;: Calculated : Found: Ag = 35.29 per cent. 34.98 per cent. 282 P. A. Levene. The filtrate from hypoxanthin silver nitrate gave on neutralization with ammonia only a very small precipitate. Before reviewing the comparative yields of the bases, it may be well to bear in mind that it is impossible to obtain the individual bases in a sufficient degree of purity without loss of the substance, and that the quantities obtained in this work are regarded only as approximating those actually present in the tissues. The comparative yields calculated for 5 lbs. of the gland were as follows: Fresh gland: Digested gland: Adenin: 4.0 gms. ert Guanin : ZO 0.200 gm. Hypoxanthin : trace Wes) Xanthin : trace 1.2 4§ Pyrimidin bases. — Kutscher found hystidin on self-digestion of the gland, and the author isolated thymin, uracil, and cytosin from the cleavage products of the pancreas-nucleic acid. On self-digestion of the gland uracil also was obtained, the presence of cytosin was made probable, but thymin could not be detected. Further, it was estab- lished that pancreatic extract does not possess the power of trans- forming thymin into uracil. Thus, the difference in the pyrimidin bases obtained on autolysis of the gland, and on cleavage of the pan- creas-nucleic acid, remained without explanation. It was also very surprising that on acid hydrolysis of the gland, the pyrimidin consisted chiefly of uracil and of cytosin, thymin not being detected at all../ This occurrence could be explained best by the assumption that there is more than one nucleic acid present in the pancreas, one yielding on cleavage the three known pyrimidin bases, the other containing no thymin, but uracil either as the only base of that group, or ina preponderating quantity. Of course there is also a possibility that in the presence of the large quantity of pro- teid material, and of other tissue constituents, the nucleic acids are affected by boiling mineral acid in a different manner than are free nucleic acids. This supposition, however, has to be abandoned for the reason that on acid hydrolysis of the spleen only thymin and cytosin and no uracil at all could be detected. Of course the possibility is not excluded that both uracil and cytosin are secondary products, and that the gland contains more . of the mother substance than does the free nucleic acid. Doctor Stookey and myself are engaged in the investigation of the nucleic The Autolysts of Animal Organs. 283 acid of the pancreas, with a view to ascertain the presence in the gland of more than one nucleic acid. On hydrolysis of the fermented gland only uracil could be identified ; also cytosin was absent. It thus seems probable that cytosin is trans- formed into uracil in course of self-digestion of the pancreas. Experimental part.— The glands were heated with strong hydro- chloric acid over direct flame with return condenser for twelve hours. The melanin formed during this cleavage was removed by filtration, and the filtrate condensed under diminished pressure. The residue was re-dissolved and reconcentrated repeatedly. The remaining hydro- chloric acid was removed by means of lead carbonate, and the excess of lead by sulphuretted hydrogen. The filtrate from lead sulphide was concentrated under diminished pressure until crystals of tyrosin began to appear. The tyrosin was removed by filtration, and from the re- maining liquid the pyrimidin bases were removed by silver and barytic water following the methods of Kossel-Jones and Kossel-Kutscher. The precipitate containing the silver salts of the bases was taken up in sulphuric acid, the silver and sulphuric acid removed in the usual manner, and the remaining liquid concentrated under diminished pressure toa very small volume. On standing, a crystalline sediment formed. It consisted chiefly of uracil and of some cytosin. The mother liquid was treated with a saturated picric acid solution, and allowed to stand over night. A deposit of cytosin picrate was thus formed. The uracil precipitate was recrystallized out of a hot 2 per cent solution of sulphuric acid. The substance thus formed had the following composition: 0.1600 gm. of the substance were used for a Kjeldahl nitrogen esti- mation. It required 28.5 c.c. #4 H,SO, to neutralize the ammonia thus formed. Por C,H, N,O; : Calculated : Found: N = 25.05 per cent. 24.94 per cent. The mother liquid of the uracil was neutralized with barium hydrate, filtered, and from the filtrate cytosin precipitated by means of a con- centrated solution of picric acid. The two picric acid precipitates were combined and transformed into the sulphate in the usual manner. Typical crystals of the basic salt of cytosin appeared. This sulphate was then transformed into the platinic chloride double salt, and as such analyzed. 284 P. A. Levene. It had the following composition : 0.144 gm. of the substance gave 0.0440 gm. of Pt. For (C,H;N,0),, PtCl,, 2HCI: Calculated : Found: Pt = 30.87 per cent. 30.55 per cent. The mother liquid of the cytosin picrate was treated with sul- phuric acid and ether to remove picric acid, and then with phospho- tungstic acid to precipitate hystidin. The phosphotungstic precipitate was decomposed in the usual manner, and the solution obtained in this way evaporated to dryness under diminished pressure. The residue was dissolved in strong hydrochloric acid, and placed over sulphuric acid in a vacuum desiccator. Hystidin, however, did not crystallize. The solution was then taken up with absolute alcohol, and allowed to stand, and crystalline precipitate of the hystidin hydro- chloride then was formed. The yield, however, was quite small. In the second experiment the gland was taken up in 0.5 per cent sodium carbonate solution, and subjected to autolysis for two months. At the end of that time the sodium carbonate was neutralized with acetic acid, and the product of autolysis concentrated until the weight of the mass was reduced to the original five pounds. It was then treated like the gland in the previous experiment. Only uracil was obtained. The substance employed for analysis had the following composition: 0.1420 gm. of the substance was employed for a Kjeldahl] nitrogen estimation. The Gunning modification was used. It required 25.2 c.c. of 45 H,SO, for neutralization. Fone, Et NOs: Calculated: | Found : N = 25.05 per cent. 24.84 per cent. The neutralized mother liquid from uracil, gave no precipitate with a concentrated solution of picric acid, thus showing the absence of cytosin. The picric acid was then removed in the usual manner and the solution thus obtained was treated with a solution of phospho- tungstic acid to remove hystidin. A very small precipitate resulted. The precipitate was decomposed in the usual manner. The yield of hystidin from this precipitate was very insignificant. The comparative yields of the bases are shown in the following table. 5 lbs. of the gland were used in such experiment. The Autolysts of Animal Organs. 285 Fresh gland: Fermented gland: Uracil : 0.460 gm. 0.600 gm. Cytosin : 40) CO Hystidin : trace trace Hexon bases. — Arginin and lysin were found by Kutscher among the products of self-digestion of the gland. Thetwo bases were found also on acid hydrolysis of the gland. However, on acid hydrolysis of the fermented gland neither of these bases could be detected. On their place tetramethylene-diamin was identified. An analogous ob- servation was made by Lawrow. Among the end-products of self- digestion of the gastric wall he failed to discover any arginin or lysin, but did find tetramethylene-diamin and pentamethylene-diamin. The difference in the observations of Kutscher and those of the author again may be explained by the fact that intracellular enzymes cannot be easily extracted, and that therefore the tissue itself is more active than any extract of it. Experimental part. — The filtrate from pyrimidin fraction was treated according to Kossel for removing arginin. The substance was identi- fied as its silver nitrate salt. 0.1905 gm. of the substance gave 0.0520 gm. of Ag. For G.ti,,.N,0,, HNO;, AgNO, : Calculated: Found: Ag = 26.54 per cent. 26.75 per cent. The filtrate from arginin served for obtaining lysin. The base was precipitated by means of phosphotungstic acid, the acid removed in the usual manner, and lysin precipitated by means of an alcoholic solution of picric acid. The salt was recrystallized out of water once and dried in toluol bath and analyzed. 0.1282 gm. of the substance gave on combustion 0.1797 gm. of CO, and 0.0570 gm. of H,O. Porn Chi N,O;,C,H,(NO,).,0OH- Calculated : Found: C = 38.40 per cent. 38.31 per cent. E53 i" 5.00 “a The digested glands were treated in the same manner, but the attempt to obtain arginin from the corresponding fraction proved futile. The lysin fraction was evaporated to a very small volume. A small part of it tested with an alcoholic solution of picric acid gave no precipitate; it, however, formed a crystalline picrate on the addi- 286 P. A. Levene. tion of a concentrated solution of sodium picrate. Washed with alcohol and ether, and dried in toluol bath, it had the following composition : 0.1190 gm. of the substance gave on combustion: 0.1538 gm. CO,, 0.0415 gm. of H,O. 0.1513 gm. of the substance gave on combustion 270 c.c. of nitro- gen (over 50 per cent KOH) at t= 24.0 C. and p= 755 mam. For G,H,.Ns, 2 C,H,CNO;).08H : Calculated: Found: € = 35.16 per cent., 35.24 per cent. H = 3.30 “ 3:87 7S N=2051 “ AU) The comparative yields of the bases are shown in the following table. Five pounds of the gland were used in each experiment. Fresh gland: Fermented gland: Arginin: 6.4 gms. 0.0 gm. Lysin: “ies 0.0 “ Putrescin : OO Oe AMMONIA. It was found that the fermented glands yield on hydrolysis with hydrochloric more ammonia than the fresh ones. Experimental part.— The hydrochloric acid solution of the glands used for the estimation of the pyrimidin and hexon bases was con- centrated to a definite volume. 10 c.c. of the solution were diluted to 50c.c. 5 c.c. of this solution were used for a Kjeldahl nitrogen estimation. 20 c.c. of the same solution were partly neutralized with a solution of caustic potash, then rendered alkaline with magnesium oxide and distilled under diminished pressure at 4o° C. until the dis- tillate ceased to contain ammonia. The comparative yields calculated for five pounds of the glands were as follows: Fresh gland: Fermented gland: Total nitrogen : 48.44 gms. 57.68 gms. Nitrogen as ammonia: 6.06 “ 10.06 “ Nitrogen ammonia in percentage of the total: 12.34 per cent. 17.44 per cent. The Autolysts of Animal Organs. 287 SPLEEN. Purin bases. — As stated already, while this investigation was in progress the very interesting work of Jones appeared. Jones came to the conclusion that on self-digestion of the spleen guanin remains unaltered, adenin is transformed into hypoxanthin, and that no xanthin can be detected. The digestion was continued five days (only the soluble part of the gland being used for experiment),— and during that time no diminution in the total quantity of purin bases was observed. In the present investigation the entire gland was used for the experiment, and the digestion lasted in one experiment twelve, and in the other fifteen weeks. The results were very similar to those obtained on digestion of the pancreas. The total amount of purin bases was greatly diminished in the digested organs. Guanin and adenin apparently disappeared, while hypoxanthin was isolated in a considerable quantity and xanthin in a very small quantity. However, the fresh glands contained besides guanin and adenin, also hypoxanthin and xanthin, and the quantity of hypoxanthin was not much lower than in the digested gland. Still the assumption that hypoxanthin and xanthin are intermediate products seems quite justifiable, especially in view of the observation of Jones, and those of the writer regarding the pancreas. The source of the hypoxanthin and xanthin in the fresh gland still remains to be ascertained. The nucleic acid obtained by the writer’s second process does not seem to contain any of these two bases; but whether they are derived from some other nucleic acid not yet isolated, or whether they are oxidation products of adenin and guanin, will be the subject of a special investigation. : Experimental part.— 10 lbs. of the gland were prepared in the usual manner and decomposed with a 5 per cent solution of sulphuric acid. The further treatment was the same as that described in the pancreas experiment. The crude guanin was dissolved in dilute sulphuric acid, decolorized with charcoal, and filtered. The sulphate was converted into the free base, and this into the sulphate. For analysis, the sulphate again was transformed into the free base. It was washed with alcohol and ether, and dried in xylol bath. 0.1122 gm. of the substance gave 45 c.c. of nitrogen (over 50 per: cent KOM) at t=23 €. and p— 765. mim. 288 P. A. Levene. For C,H,N,O: Calculated: Found: N = 46.36 per cent. 46.68 per cent. The mother liquid from guanin sulphate was rendered alkaline by ammonia, then acidulated with acetic acid, boiled, and filtered. The clear filtrate was allowed to cool. The precipitate this formed was washed with alcohol and ether, dried in xylol bath, and analyzed. 0.1130 gm. of the substance was employed for a Kjeldahl nitrogen estimation. 30.20 {% H,SOs were required to neutralize the ammonia thus formed. For G.H,N,0,: Porn@er, N;O; : Calculated: Found: N = 36.84 per cent. 33.70 per cent. 37.36 per cent. The substance thus could be either xanthin or a mixture of methylxanthin with guanin. Part of it was then dissolved in a normal sodium hydrate solution, and an excess of nitric acid (two parts of strong acid to three parts of water) was added. On standing, a crystalline deposit formed, which showed that the substance was xanthin. The filtrate from crude guanin was treated with ammoniacal silver chloride, and the silver salts thus obtained were decomposed with hydrochloric acid. The hot filtrate from silver chloride was treated with a solution of sodium picrate, and the adenin picrate was removed on suction funnel while the mother liquid was warm. After it was once recrystallized out of water, washed with alcohol and ether, and dried in toluol bath, the substance had the following composition : 0.1330 gm. gave on combustion 0.1755 gm. of CO, and 0.0315 gm. or HO. Bor C,H,N, CoE ( NOs) ,Or: Calculated : Found: C = 36.26 per cent. 35.99 per cent. ie — 2220 2.63 ss The filtrate from adenin picrate gave on concentration another precipitate. To this precipitate sufficient boiling water was added to effect solution, On standing, the solution turned into a crystalline mass, consisting of very long (some reaching 2 cm.), light yellow crystals. They were filtered on suction funnel, washed with alcohol and ether, dried in xylol bath and analyzed. The Autolysts of Animal Organs. 289 0.1630 gm. of the substance gave on combustion 33.5 c.c. of nitrogen (over 50 per cent KOH) at t° = 24° C. and p= 760 mm. The substance was recrystallized and analyzed. 0.135 gm. of the substance gave on combustion 28.0 c.c. of nitrogen (over 50 per cent KOH) at t° = 25° C. and p = 760 mm. Mor. Get, N,O,¢€,H,(NO,),0H» For C,H,N,0,C,H,CNO,),0H: Calculated : Found: N= 26.85 per cent: 23.96 per cent. 23.72 per cent. 2 SO ee Whether this substance was an impure hypoxanthine or para- xanthine will be established by further experiment. There is not sufficient material to do so at present. The filtrate from this precipitate was acidulated by hydrochloric acid and extracted with ether. The solution free from picric acid was rendered ammoniacal and treated with an ammoniacal solution of silver chloride, and the precipitate thus formed filtered on suction funnel, and decomposed with hydrochloric acid. The filtrate from silver chloride was concentrated to dryness under diminished pres- sure, the residue taken up in warm water and again concentrated to dryness. This operation was repeated several times, and finally the residue was taken up in alcohol and evaporated to dryness under diminished pressure. The water insoluble part of the residue was dissolved in normal sodium hydrate solution and to the solution an excess of nitric acid (two parts of strong nitric to three parts of water) was added. On standing, a crystalline sediment formed. It gave a very pronounced ‘“xanthin” reaction, and thus may be regarded as xanthin. The water soluble part of the hydrochloric acid residue was treated with sodium picrate. The precipitate was dissolved in hot water. On cooling, a precipitate formed, which was removed by filtration. It consisted apparently of adenin picrate, and on standing, a second precipitate of hypoxanthin picrate appears. Washed and dried in the usual manner, it had the following composition: 0.1820 gm. of the substance gave 42 c.c. of nitrogen (over 50 per ¢ent KOH), at.t°=22.5° C. and p= 762 mm. For. C.H,N,0,. €,H,Q@NO,),0H: Calculated : Found: N = 26.85 per cent. 26.86 per cent. Digested glands.— 5 lbs. of spleen were prepared in the usual manner and allowed to digest in 0.2 per cent solution of acetic 290 P. A. Levene. acid for twelve weeks. The products of digestion were concentrated to the original weight and decomposed with a 5 per cent solution of sulphuric acid. The product was treated in the usual manner. Neither guanin nor adenin picrate was obtained. In place of adenin picrate, on standing forty-eight hours a picrate appeared resembling that of hypoxanthin. It was recrystallized out of water, washed with alcohol and ether, dried in toluol bath, and analyzed. 0.148 gm. of the substance gave on combustion 34.0 c.c. nitrogen (over 50 percent KOH) at t° =26.0° C. and p= 762. Por C,HN,O,7 Cb (i@> Orr: Calculated : Found: N = 26.85 per cent. 26.30 per cent. The filtrate from this precipitate was acidulated with hydrochloric acid, extracted with ether, rendered ammoniacal, and then treated with an ammoniacal solution of silver chloride. The precipitate thus obtained was very small. Further separation of the bases was accomplished by Kossel’s method. . The silver salt insoluble in nitric acid of sp. gr. 1.1 had the following composition: 0.1510 gm. of the substance gave 0.0526 gm. of Ag. For'C,H,N,O. -AgNO;: Calculated : Found: Ag = 35.29 per cent. 34.83 per cent. The filtrate from this precipitate gave on neutralization with ammonia only a very small flocculent precipitate, which might have been xanthin. In the second experiment 15 lbs. of the gland were allowed to digest fifteen weeks. They were treated in the same manner as in the first experiment, and the results were practically the same. Only a very small precipitate appeared in place of crude guanin, so that it was considered useless to purify. No adenin picrate was obtained at all. Hypoxanthin was removed as a picrate, the mother liquid of this salt was treated with hydrochloric acid and ether to remove picric acid, and the remaining purin bases were removed by an ammoniacal solution of silver chloride. The precipitate thus obtained was decomposed by hydrochloric acid and the filtrate from silver chloride was concen- trated under diminished pressure. The residue was treated as described in previous experiments to remove the excess of hydro- ‘ The Autolysts of Antmal Organs. 291 chloric acid. The water insoluble part was dissolved in a normal sodium hydrate solution and an excess of nitric acid (two parts of acid to three parts of water). On standing a crystalline deposit, giving a pronounced xanthin test, appeared. This substance, there- fore, might have been xanthin. The comparative yields, calculated for 5 ]bs. of the gland, were as follows: Fresh gland : Digested gland: Adenin: 1.85 gms. 0.0 gm. Guanin: TOFS ONO Hypoxanthin : 050° * es Xanthin : Oem ONS OMe Pyrimidin bases. — Concerning the pyrimidin bases of the gland it is known from the writer’s analysis of the spleen-nucleic acid, that all three bases occur, and that thymin predominates over the other two. Jones, who recently analyzed the pyrimidin bases of the self- digested spleen, detected only uracil. In the present investigation, thymin and cytosin were obtained from the fresh gland, and thymin and uracil from the digested gland. Thus the assumption of Jones that cytosin on digestion of the gland is transformed into uracil seems justifiable. Experimental part.— The procedure in obtaining the pyrimidin bases was identical with that described in the pancreas experiment. The crude bases were dissolved in a 2 per cent solution of sulphuric acid and decolorized with charcoal. ‘The first base to crystallize had the typical appearance of thymin, and the following composition : 0.1300 gm. of the substance was employed for a Kjeldahl nitro- gen estimation. 20.5 c.c. 7g H,SO, were required to neutralize the ammonia thus formed. For Cl, NO, Calculated : Found: N = 22.22 per cent. 22.07 per cent. From the mother liquid of thymin, cytosin picrate was obtained in the usual manner. It had the following composition: 0.1462 gm. of the substance gave on combustion 0.1880 gm. CO, and 0.0310 gm. H,O. For CG, rN,O CC. Ho(N©>).OH - Calculated : Found: C = 35.29 per cent. 35.07 per cent. H = 2.35 es 2.36 292 P. A. Levene. From the mother liquid of thymin and cytosin, histidin dichloride was obtained in the same manner as in the pancreas experiments. The yield was very smail. Digested glands were treated in the manner described in the pan- creas éxperiment. Thecrude bases were dissolved in a hot 2 per cent solution of sulphuric acid, decolorized and allowed to cool. The sub- stance that crystallized first had the appearance and the composition of thymin. 0.1240 gm. of the substance was used for a Kjeldahl nitrogen estimation. 20.10 c.c. 74 H,5O, were required to neutralize the ammonia thus formed. . For Cli N.O.: Calculated: Found : IN = 22°22, percent: 22.69 per cent. The mother liquid on standing in vacuum desiccator gave another precipitate resembling uracil. It was twice recrystallized out of I per cent solution of sulphuric acid. For analysis it was dried in xylol bath. Its composition was as follows: 0.1285 gm. of the substance was employed for a Kjeldahl nitrogen estimation: 23:30 ¢.c. of 4 H,SO, were required: Por CEN, OO: Calculated : Found: N = 25.05 per cent. 25.38 per cent. No appreciable quantity of cytosin picrate could be isolated. The phosphotungstic precipitate supposed to contain histidin was so small that further treatment was abandoned. The comparative yields of the bases calculated for 5 Ibs. of the gland were as follows: Fresh gland: Digested gland: Thymin: 0.400 gm. 0.380 gm. Uracil: eeagere 0450535 Cytosin: 0.300 “ OCS = Histidin : trace trace Hexon bases. — Both hexon bases, arginin and lysin, were obtained on acid hydrolysis of the fresh as well as of the digested glands. However, the yields were markedly smaller in the digested glands. The process employed for obtaining the bases were the same as described in the pancreas experiment. The Autolysis of Animal Organs. 293 Arginin was identified as silver nitrate. The analysis of the silver nitrate obtained from the fresh glands gave the following results : 0.1990 gm. of the substance gave 0.535 Ag. The analysis of the substance obtained from the digested glands gave the following results: 0.2200 gm. of the substance gave 0.0595 of Ag. For C,H,,N,0,, HNO,, and AgNO,: Calculated : Found: Fresh gland: Digested gland: Ag = 26.54 per cent. 26.83 per cent. 27.07 per cent. Lysin was identified as its picrate. 0.1540 gm. of the substance obtained from the fresh gland gave on combustion 0.2170 gm. CO, and 0.0700 gm. H,O. Hor C,t1,,N,0,, €.H,(NO,),0H: Calculated : Found: C = 38 40 per cent. 38.46 per cent. H = 4.53 ¥ 5.03 ‘ 0.1590 gm. of the substance obtained from the digested glands gave 26.5 c.c. nitrogen (over 50 per cent KOH) at t°'=26.6° C. and — 755 mim. For €,1,,N,0,, C,H,(NO,),0H: Calculated : Found: N = 18.67 per cent. 18.92 per cent. The comparative yields calculated for 5 pounds of the gland were as follows: Fresh gland: Digested gland: Arginin : 3.2 gms. 1.5 gms. Lysin : ZO 2 LIVER. Purin bases. — In a general way the purin bases undergo the same changes on digestion that they do in the organs. The difference may be only quantitative. Thus, a diminution in the total quantity of bases was noted. Guanin was isolated from the digested gland in a very small quantity, and adenin was obtained, though in comparatively small quantity. In this respect the process in the liver differs from that in the pancreas and spleen. After ten weeks of digestion of 294 Pw, Levene pancreas, adenin could not be detected. Hypoxanthin was isolated from the fresh and from the digested gland, while xanthin could be demonstrated in very small quantity in the fresh gland. Experimental part. — 10 lbs. of liver were treated in exactly the same manner as described in the experiments on the spleen. Guanin was analyzed in form of sulphate. 0.1230 gm. of the substance was employed for a Kjeldahl nitrogen estimation. 28.0 c.c. of {4 H,SO, were required. For (C, eN.© ) le SOp 2a, 0 Calculated: Found: N= 32-2 percent: 31.88 per cent. Adenin was identified as the picrate. 0.1325 gm. of the substance gave 36 c.c. nitrogen (over 50 per cent KOH) at t?=26" C. and pa-75,5mina. For C,H: N;CHa(NO;):OH: Calculated : Found: N = 30.71 per cent. 30.95 per cent. Hypoxanthin was isolated in the form of the picrate. 0.140 gm. of the substance gave on combustion 33 c.c. nitrogen (over 50 per cent K@H)/at t '=27 C. and p=76oimmn For €: HNO; Gar CN @s). Or: Calculated : Found: N = 26.85 per cent. 26.90 per cent. The picrate was decomposed by nitric acid and ether, and on con- centration the nitrate of hypoxanthin crystallized. Xanthin was isolated in a small quantity in form of its insoluble nitrate. Digested glands. — 5 lbs. of the gland were employed for the ex- periment. The yield of purin bases was very small. The crude guanin was once crystallized out of sulphuric acid and then converted into the free base. On analysis it gave the following results: 0.1200 gm. of the substance was employed for a Kjeldahl nitrogen estimation. 33.90 c.c. of 44 H,SO, were required. For €-HieN Or Por 'C_ELN,O, : Calculated : Calculated: Found: N = 46.36 per cent. 36.84 per cent. 37.6 per cent. Thus showing that the substance consisted chiefly of xanthin. The Autolysts of Animal Organs. 295 From the filtrate of the guanin fraction adenin and hypoxanthin were isolated in form of picrates. Both were in small quantities. The adenin picrate was once recrystallized and had a MP.= 282°C. Hypoxanthin gave on analysis the following results: 0.113 gm. of the substance gave 28 c.c. nitrogen (over 50 per cent Ot )sat t:—26.5° C. p=758 mm. For C.H,N,O, C;HjCNQ,),0H: Calculated : Found: N = 26.85 per cent. 27.32 per cent. The comparative yields calculated for 5 lbs. of the glands were as follows: Fresh gland: Digested gland: Guanin: 0.755 gms. trace Adenin: ES ro 0.37 gm. (of the crude subst.) Hypoxanthin : Hal a 0.300 “ Xanthin: very small quantity . 0.300 “ Pyrimidin bases. — The yield of these bases was very small. How- ever, a precipitate consisting of the scales typical of thymin was obtained from the fresh gland. From the mother liquid little of cytosin picrate was isolated. It had a melting point of 275° C. Attempts to isolate the bases from the digested gland were futile. Hexon bases.— Both bases were obtained from the fresh and digested gland. However, the yield from the digested glands was smaller, thus showing that also in the liver the hexon bases undergo further decomposition. The discovery of arginase by Kossel explains the diminution in the yield of arginin. Arginin was identified in the form of its silver nitrate. 0.1875 gm. of the substance obtained from the fresh gland gave 0.052 gm. of Ag. 0.2460 gm. of the substance obtained from the digested gland gave 0.0655 gm. Ag. For C,H .N,0O;, HNO, AgNo,: Calculated : Found: Fresh gland: Digested gland : Ag = 26.54 per cent. 27.12 per cent. _ 26.66 per cent. Lysin was identified in the form of its picrate. 0.1330 gm. of the substance obtained from the fresh gland gave 22.0 c.c. nitrogen (over 50 per cent KOH) at t° = 26.5° C.and p=760 mm. 296 P. A. Levene. 0.2020 gm. of the substance obtained from the digested gland gave 33-0 c.c. of nitrogen (over 50 per cent KOH) at t° = 23° C. and p= 762 mm. For C,H,,N,0,;0 42,0 N@,0r: Calculated: Found: Fresh gland: Digested gland: N = 18.67 per cent. 18.94 per cent. 19.09 per cent. The comparative yield calculated for 5 lbs. of the gland were as follows : Fresh gland: Digested gland: Arginin: 10.8 gms. 5.2 gms. Lysin: IPA? 2 Oe I wish to express my indebtedness to Prof. R.H. Chittenden, under whose control this investigation was carried out. The expenses of the wayk were defrayed by a grant from the Car- negie Institution of Washington. ON THE SWELLING OF ORGANIC TISSUES. — RESEARCHES ON. THE CORNEA. By, G. BULLOT: H. LEBER, in 1873, found that, when the endothelium covering the posterior surface of the cornea is removed, the eyeball being kept entire, the corneal stroma swells considerably and becomes opaque. The swelling is due to the corneal stroma absorbing the aqueous humor filling the anterior chamber of the eye. Normally, this imbibition cannot take place on account of the impermeability of the living endothelium to aqueous humor. Similarly to the corneal endothelium, the epithelium covering the anterior surface of the cornea is also impermeable to water, as Th. Leber and others have shown. But its removal is not followed by any absorption of liquid by the stroma. The cornea of the entire eyeball, with its epithelium scraped off, remains as thin and transparent as itis normally, whether the eyeball is left in situ with the eyelids open or sutured, or is enu- cleated and transplanted into the peritoneal cavity. (L. Lor and G. Bullot.) The present paper discusses the reason for this difference in the reaction. I. The phenomenon is not due to a difference in the power of im- bibition of the conjunctival layers on each side of the stroma. When the cornea of the rabbit is detached and placed in distilled water, fresh water, or sodium chloride solutions, it swells considerably, whether only the anterior surface or the posterior surface alone has been scraped. Ii. It is not due to the quantitative difference which exists in the salt contents of the aqueous humor on one side, and the tears, con- junctival secretion, or peritoneal liquid on the other. S. Ringer, Fr. Hofmeister, J. Loeb pointed out that the degree of swelling of organic tissues or substances is affected by the presence of salts. S. Ringer particularly observed that the dry tissue of the marine alga Laminaria swells much more in sodium chloride solutions than in calcium chloride solutions. Now, from what is known of the physio- logical antagonistic action of sodium and calcium salts, the question 250) 298 G. Bullot. suggests itself whether a similar antagonism exists here in the mere physical process of swelling when those salts are brought together in the same liquid as is the case for organic liquids. The antagonistic action would be obtained, as usual, when a certain relation exists between the quantities of the salts. On the other hand, it is known that the aqueous humor has not the same quantitative salt composi- tion as the other liquids under consideration. It might therefore be that the aqueous humor contains too little calcium salt to prevent the swelling due to sodium chloride, while the tears, conjunctival secretion, and peritoneal liquid possess a sufficient quantity of calcium salts to check the swelling. Such, however, is not the case. At concentrations equal to those of the sodium and calcium salts in organic liquids, or at higher or lower concentrations, there is, as a rule, no antagonism between calcium and sodium, although, at concentrations of the same order of magnitude as the one at which they exist in organic liquids, the calcium salts, when present alone, make the cornea swell much less than the sodium salts alone. Many experiments made either on fresh pieces of the detached cornea or on small strips cut from the dry cornea give the same neg- ative result. Only when the quantity of sodium chloride is very small can its action on the swelling be checked by the calcium salts. Data. — If pieces of the scraped fresh cornea be placed in various liquids, at a temperature of from 15 to 25° C., it will be observed that the swelling is greatest in distilled water. The thickness attained reaches 3 mm., which is ten times greater than the thickness of the normal cornea. In sodium chloride solutions from +575 to 2 m the degree of swelling is distinctly less.1 The thickness of about 2 mm., in yg7%yo, decreases very slowly with an increase in the concentration of sodium chloride so that, in 2 # sodium chloride, it is equal to 1.5 mm. (Identical results are obtained with solutions of potassium chloride, lithium chloride, am- monium chloride, sodium bromide, sodium iodide, sodium fluoride, sodium nitrate, sodium sulphate, sodium carbonate.) * 1 This behavior of the cornea differs from that of gelatine, which was found by Fr. Hofmeister to swell more in sodium chloride solutions than in distilled water. 2 To test the action of all these salts, only narrow strips cut from the dry stretched cornea were employed. Their use is to be desired, on account of the quickness of the swelling and the number of trials which can be performed with a single cornea. Such pieces cannot always be used because the degree of swelling is less than with pieces of the fresh cornea (2 mm. in distilled water). The same relation exists, however, between the degree of swelling in distilled water and in the various sodium chloride solutions. On the Swelling of Organic Tissues. 299 With calcium chloride solutions from yo%5y to m, it is observed that the cornea swells the least in +7, ; the swelling is much less than in any of the sodium chloride solutions. With higher or lower concentrations, the degree of swelling gradually increases, but the increase is greater towards m than towards y,%y59- (Exactly similar results are reached with barium chloride, strontium chloride, and magnesium chloride solutions, with regard to the shape of the curve plotted with the figures of the concentration, and the degree of swelling, and with regard to the absolute degree of the swelling.) * In mixtures of sodium chloride and calcium chloride an antagonistic action of calcium chloride may be clearly seen only when the composi- tion is about ;>> NaCl and ;35 CaCl. In this case, the degree of swelling is as small as in solutions of calcium chloride alone. When the mixture contains 7%, NaCl, which is about the concentration of sodium chloride in the organic liquids, the antagonistic action of calcium chloride completely disappears. III. The difference we are discussing is due principally to the action of the intraocular pressure which does not prevent imbibition through the posterior surface, both forces acting in the same direction, but which opposes imbibition through the anterior surface. When the entire eyeballs of the rabbit, with the epithelium scraped off, are suspended in various solutions of sodium chloride from qi m to 785 m, it is observed, after a certain number of hours that those solutions which produce hypertension of the eyeball, on ac- count of their hypotonicity compared with the internal liquids of the eye, leave the cornea thin and nearly transparent. On the contrary, in the hypertonic solutions, z. e. solutions above dis m,in which a hypotension of the eyeball is determined, the cornea swells consider- ably. This result cannot be attributed to the direct action of the solution on the cornea since we saw, in the case of the detached cornea, that the swelling is considerable in all concentrations of sodium chloride and less in the high concentrations than in the low ones. In distilled water, in which a very strong hypertension is produced, the cornea after a certain number of hours is as thin as in the normal state. Later, however, when through the injurious action of the distilled water, the endothelium of the posterior surface of the cornea is killed, the cornea swells through imbibition of aqueous humor. In fresh water, which has no such injurious action on the endothelium, and in which also a strong hypertension is pro- 1 Also studied with strips of the dry cornea. 300 G. Bullot. duced, the cornea remains of normal thinness even after two days. Results quite similar to those obtained with sodium chloride are reached with various concentrations of cane-sugar, which has proved to be of no influence on the swelling of the detached cornea. (What- ever may be the concentration of cane-sugar from 77 to m, the degree of swelling of the detached cornea is identical with that in distilled water. ) On the normal eyeball in situ in the living animal, the intraocular pressure is quite sufficient to prevent the imbibition of the scraped stroma through the anterior surface. The same is true when the enucleated eyeball is placed in the peritoneal cavity of the living animal. But when the enucleated eyeball is suspended in blood serum or white of egg, the imbibition through the anterior surface is also checked, at least partially, in spite of a decrease in the intra- ocular pressure, which is readily determined under those conditions. Here a second factor must be considered. IV. This second factor which, in case the eye has been enucleated and placed in albuminoid liquids, intervenes independently of the intraocular pressure to produce the difference in the degree of imbi- bition of the corneal stroma through the anterior and posterior surface, is the action of colloidal substances present in albuminoid liquids and nearly completely absent in the aqueous humor. E. H. Starling has found, using blood serum, that the albuminoid substances of blood serum have an osmotic pressure of about 35 mm. Hg. As there is about seven per cent of albuminoid substances in blood serum, the pressure would be about 5 mm. Hg for each per cent. Therefore, it is to be expected that when blood serum is brought into contact with a tissue capable of absorbing water, it will give water to that tissue only if the force of imbibition which causes the water to go into the absorbing tissue is greater than the osmotic pressure of the serum. And conversely, when the tissue is already completely swelled, the blood serum will absorb water from it and diminish the degree of swelling if the osmotic pressure is greater than the force which holds the water in the swollen tissue. The first point is proved by the fact that, when the scraped eyeball is placed in blood serum, although the intraocular pressure decreases, the cornea swells very slowly and very moderately (1 mm.), much less than if it had been placed in a sodium chloride solution possess- ing the same decreasing effect on the intraocular pressure. This cannot be due to. the action of the calcium salts of the serum, since it On the Swelling of Organic Tissues. 301 | has been said above that, at the concentration at which sodium chloride exists in blood serum, the calcium salts have no longer any antagonistic action on them, as far as the process of swelling is con- cerned. When white of egg is used instead of blood serum, similar results are obtained, but the degree of swelling is still less. With blood serum as well as with white of egg, the cornea, although swollen, remains transparent. Such a state of transparency is permanent in blood serum. In white of egg, it is followed, after a certain number of hours, by the appearance of a moderate opacity which is not accompanied by any further increase in the degree of swelling. This leads to the conclusion that the albuminoid solution goes into the corneal stroma, since, in salt solutions free from colloidal substances, the swelling is always accompanied by opacity except in case of suffi- ciently concentrated solutions of salts of bivalent metals. If, instead of the entire eyeball, pieces of the scraped cornea are placed in blood serum or white of egg they swell, but much less than in equivalent solutions of sodium chloride (1 mm.) and less in white of egg than in blood serum. The second point, the subtraction of water by albuminoid solutions from the already swollen cornea is made evident in the following manner. Three entire eyeballs, with the corneal epithelium scraped off, are first placed in a 4 m solution of cane-sugar, which is hyper- tonic compared with the liquids of the eye and the albuminoid solutions to be used afterwards. After two or three hours the corneas are completely opaque; the hypotension of the eyeballs is considerable. From one of them, the cornea is then detached and is found to be strongly swollen with a liquid rich in cane-sugar. This gives an indication concerning the degree of swelling of the cornea of the two other eyeballs at that moment. One of them is then trans- ferred into blood serum or white of egg. After three or four hours in case of white of egg, five or six hours in case of blood serum, the cornea becomes again completely transparent and, being detached, shows itself to be again nearly as thin as normal. The cornea of the third eyeball which was kept in cane-sugar being now detached, shows the same degree of opacity and of swelling as the first one. This proves that white of egg and blood serum have the power of subtracting water from the swollen cornea, even when the liquid of imbibition is hypertonic compared with those colloidal liquids. In case the second eyeball should have been kept longer in white of egg or blood serum, the cornea would again have swelled moderately, 302 G. Bullot. more in blood serum than in white of egg, keeping its transparency in blood serum, becoming gradually opaque in white of egg, so that identically the same state of affairs would exist as when the eye is directly placed in serum or white of egg without passing through cane-sugar. Here, as above, the state of permanent transparency of the cornea in the case where blood serum is used, warrants the conclusion that some of the colloidal substance has entered this membrane. The first subtraction of water from the swollen cornea, under the influence of albuminoid liquids, cannot be attributed to the osmotic action of the salts of the blood serum, as the cornea has imbibed a hypertonic solution. Nor could this be due to inequality in the initial rate of osmosis of cane-sugar and of the salts of the blood which, after a certain time, would make the blood serum more con- centrated than the liquid imbibing the stroma and enable it to absorb water from the cornea through osmosis (Lazarus Barlow). It could not be so, because when the swollen cornea is placed in any concentration of sodium chloride or calcium chloride, even if the imbibing liquid is distilled water, the swelling diminishes only to a small extent. It is therefore necessary to attribute thé subtraction of water to the action of the colloidal substance of those two liquids. If, instead of the entire eyeball, pieces of the detached scraped cornea are placed first in } # cane-sugar and, after swelling, in white of egg or blood serum, a similar subtraction of water is observed. When the fragments are again placed in cane-sugar, the cornea again swells, to decrease in thickness a second time when replaced in white of egg, and this to the same extent as before. The subtraction of water from a hypertonic salt solution through a permeable mem- brane by a colloidal solution has long been known. Lately E. H. Starling attributed it to the osmotic pressure of the colloidal sub- stance which is compelled to remain on one side of the membrane, while for the salts, as they can pass through the membrane, an equilibrium is soon reached on both sides. Thus the colloidal sub- stance, by a weak but continuous action, gradually subtracts the water placed on the other side. The same occurs here, with this only difference that, instead of the mere hypertonic crystalloid solu- tion, a swollen tissue imbibed with the solution is present. After a certain time, the colloidal substance penetrating slowly into the cornea, a certain degree of swelling is again possible. On the Swelling of Organic Tissues. 303 The subtraction of water from the previously swollen corneal stroma by an albuminoid liquid suggests the question of the struggle for water between colloids of different kinds. This question will be studied in a further paper. My thanks are due to Mr. H. Hus for his valuable assistance. ON DECAPSULATION OF THE KIDNER By ISAAC LEVIN. HE réle of the renal capsule for the function of the kidney is scarcely ascertained, less so is its réle in the economy of the entire organism. There exists also very little knowledge regarding the effect of the removal of the capsule on the function of the kidney. Notwithstanding this, decapsulation of kidneys has been recommended and practised for curative purposes. Thus R. Harrison, J. Israel, and A. Pousson reported cases where they performed nephrotomy on patients suffering from colicky pains, hematuria, and albuminuria, in the expectation of finding tuberculosis, calculus, or pus in the kidney. At the operation neither of these conditions was found, and still the symptoms, including even albuminuria, were cured. In some of these kidneys they discovered various inflammatory pro- cesses, and the albumen as well as casts found in the urine before the operation, disappeared after it. All this led the authors to believe that certain cases of nephritis may be improved or even cured by a unilateral nephrotomy. They explain the beneficial effect of the operation by the relief of the organ from congestive swelling and increased tension. G. Edebohls'came to the same conclusion from his experience with a different class of cases. He noticed in his operations for floating kidney, that the albumen and casts sometimes present before the operation, disappeared after it. His operation for floating kidney consists in partial decapsulation and anchoring of the kidney. He then took up cases of simple Bright’s disease, decapsulated the kidney, and saw his patients improve or recover. He goes much further in his conclusions than the former authors, and thinks that every case of Bright’s disease can be cured by decapsulation of the kidneys. He explains the beneficial effect of his operation by subsequent adhesive inflammation, which establishes a collateral circulation. 1 Read at the annual meeting of the Association of American Pathologists and Bacteriologists, April 1, 1904. 304 On Decapsulation of the Kidney. ‘spuogas Aq411y} ‘soyNuIU xIS Jo [BAIO}UI UY sjudsaidar aur] [wodaa yIS ay, ‘aovds Jo yox| Jo osnvd0q poonpo.da. jou ov potiad sty} Sup sp109o1 ay], “Spuosas A7xXIs Jo [vAIBIUT Uk aYVOIPUT sau, TeoNIEA YIP pue ‘pe ‘pz IST SY, ‘SpuodIs UT oUII} dy} SAAIS oUL[ ySOMOT BY} foanssaid oT1dYdsouNe ay} st auly paryy oy, “Araj4e prosvo oy ut osnssaid poojq ay} Jo 9AANO ay} JAMO] 4XOU Oy} | ADUPLY aq} FO 9AM QUINOA ayy st Suroesy ysowssddn oyy, ‘Aaupry oy) uo ureusipe jo yaya ay} Surmoys poser pasuojoid wv Jo suonsog —*T AINA 306 Lsaac Levin. Decapsulation of the kidney is now the generally accepted opera- tive interference in Bright’s disease. We see, then, that a number of modern surgeons consider the removal of the fibrous capsule of the kidney beneficial in certain diseases, while the function of the capsule is practically unknown. The question may evidently arise whether the benefit of the decapsulation, admitting that such actually exists, may not be more than counterbalanced by the injury done to the organism by depriving the kidney of its covering. I therefore undertook to examine into the possible function . the fibrous capsule of the kidney. If we compare the fibrous coverings of the liver, spleen, pancreas, and all other parenchymatous organs on FiGurE 2.— Normal kidney. Adrenalin injection. one hand, and those of the kidney on the other, we cannot fail to notice a great difference between them. While the former is very thin and firmly adherent, practically an integral part of the organ, the latter is a strong fibrous covering, easily detached from the kidney. After such detachment of the capsule, or decapsulation, the kidney immediately swells. All this makes it already, a priori, very probable that the capsule of the kidney is functionally more impor- tant than the capsules of the other parenchymatous organs. The question arises how to discover the influence of the capsule on the kidney. I chose the oncometric method of investigation, which On Decapsulation of the Kidney. 307 records the minutest changes in the size of the kidney; and here I thought, a priori, the influence of the capsule would be seen. The oncometer is in principle a plethysmograph especially modified for use upon the kidney. The apparatus which is used in the physiological laboratory of Columbia University is a bivalved, kidney-shaped metal box, hinged at the back, with a clasp in front, while a grooved notch at the centre of the rims of the valves prevents pressure on the vessels and nerves of the pedicle of the kidney when the organ is enclosed in the oncometer. In the interior of the box two pieces of thin soft rubber are so fastened to each half that a layer of water may be placed between them and the metal walls of the box. There is thus formed in each half of the box a soft water-pad, on which the kidney rests. When the kidney, freed from fat and surrounding aioe FIGURE 3. — Decapsulated kidney. Vagus stimulated. connective tissue, but with the blood-vessels and nerves entering at the hilus entirely uninjured, is laid in one half of the oncometer, and the other half is shut down upon it and tightly fastened, then each expansion or contraction of the kidney changes the quantity of water in the upper half of the oncometer. A short metal tube in the centre of the upper valve, to which is attached a piece of soft rubber tubing, allows the water in the upper half of the oncometer to rise or fall in the tube. This movement in the water column is communicated to a column of air farther in the tubing, which in turn communicates with the air of a tambour, which records the oncometric tracing. The different agents producing a shrinking of the kidney, and consequently a fall of the oncometric tracing, can be divided into two classes. The agents of the first class produce that effect by 308 Lsaac Levin. actively contracting the blood-vessels of the kidney simultaneously with the rise of the general blood pressure. The most powerful agent of this class is adrenalin. To the second class of agents belongs the stimulation of the vagus nerve. The shrinking of the kidney caused by this stimulation is not due to the contraction of the kidney vessels, but to the diminished supply of blood to the kidney through the weakening of the heart action and consequent fall in general blood pressure. While the agents of the first class produce simultaneously a fall of the kidney tracing, and rise of the carotis tracing, after the stimulation of the vagus both tracings fall. In my experiments I tried both adrenalin and stimulation of the vagus. I" yay F h my grantee te ry m. Wynnytivyivvntnas tard N uh ul Se ee Sa ee a es a Os a eC CC FIGURE 4. — Normal kidney. Vagus stimulated. The following is the course of my experiments: One of the kidneys of a dog was decapsulated through the lumbar or abdominal incision. Twenty-four or forty-eight hours later the abdomen of the same dog was opened, one of the kidneys placed in the box and connected with the recording apparatus. The carotis was also connected with a sphygmograph in order to watch the effect of the agents on the general circulation. The normal pulsation was recorded for some time on the kidney and carotis, then either adren- alin injected or the vagus stimulated. When the influence of the agents on both the kidney and carotis disappeared and the pulsation of both returned to the normal, the first kidney was removed from the box, the other placed into it, and the same experiment repeated. I performed this experiment on eighteen dogs with nearly identical results, On Decapsulation of the Kidney. 309 Comparing the tracings of the normal kidneys with those of the decapsulated ones, one notices that in the former, immediately after the injection of the adrenalin or the stimulation of the vagus, the tracing falls, then continues for some time on the same level, but always shows pulsation and returns to the old level, mostly even before the tracing of the carotis becomes normal. In the decap- sulated kidney the tracing also falls immediately after the injection, then for a considerable length of time continues as a straight line, showing an absolute cessation of pulsation in the kidney, and returns to normal much later than the carotid blood pressure. The analysis of the result of the experiments justifies the following conclusion: Any stimulus which, either by contracting the general blood pres- sure or by weakening the action of the heart, diminishes the size of the kidney, exerts a much stronger influence on a decapsulated kidney than on a normal one, and this influence also lasts longer on the former. This fact can be explained by the assumption that the capsule acts like an elastic covering. On one hand it prevents an undue overfilling of the kidney with blood, on the other hand it does not allow the kidney to remain contracted and bloodless for a long time. The other possible explanation, the theory that the decapsu- lation injures the nervous apparatus of the kidney and _ thereby changes the action of these agents is not tenable, because the vaso- motor nerves of the kidney come from the splanchnics, and the division of the latter (the only effect decapsulation can produce on the nerve) increases the size of the kidney; consequently the action of our agents ought to have been less marked on the decapsulated kidney. The research is as yet not finished. I expect to repeat it on kidneys with longer periods of decapsulation and on diseased kidneys, but the results already obtained seem to warrant the assumption that the capsule may have a much more serious influence on the function of the kidney in health and disease than one would think a priori; that it may be a powerful protection against noxious agents. In conclusion I desire to express my gratitude to Professors J. G. Curtis and F. S. Lee, in whose laboratory this work was done. or od rat a y ret , Aad ne i a : ; 4 7 a te u ad , / y : pS TR STA ti fae een De) ans ie ] O) A f iD nee 1 Oe ae my E. a aA ‘ a eae ve . (te ¢s% fs i Ne . a ie % ed . > x : ’ ( : / | . ' 7 a * 1 . ' ' * ~ 4 my 4 CHANGES IN’ THE EXCRETION OF ‘CARBON. DIOXIDE RESULTING FROM BICYCLING. By .G. .0;, HIGLEY ano W.-P. BOWEN. [From the Physiological Laboratory of the University of Michigan.]| I. INTRODUCTION. N a series of studies of the effects of muscular work that have been made in this laboratory, the purpose has been to secure results that will be of service in building up a more complete scien- tific basis for physical education. The value of results for this pur- pose depends toa large extent upon the completeness with which the effects of the work are recorded. It is not necessary that every study should take account of all or even many of the effects pro- duced by the work, but it is essential that the effects which are chosen as the subjects of investigation shall be followed continuously as long as they last, and recorded with accuracy. Here lies the weak- ness of much that has been done to discover the influence of mus- cular work upon the organism. A great number of these studies consist of only two brief tests. Such researches often bring out striking results which serve as valuable hints at the direction in which the truth lies, but their lack of completeness renders them useless as a working basis for the deduction of practical principles. Muscular work must be studied in the thorough manner followed by clinicians in studying disease. Here each change resulting from the disease is carefully traced through its entire course, thus giving an exact picture of the condition of the patient at every stage. Sim- ilarly, in order to know how muscular work may be used most effi- ciently and with least danger, we want not so much a demonstration of the greatness of its maximum effect as a full and definite knowl- edge of the course of each effect from its first appearance until nor- mal conditions reappear. With this purpose in view, emphasis has been laid in this group of researches upon making each series of records as complete as possible. Whenever it was feasible, contin- uous graphic records were made, since they give the most complete 311 312 G. O. Higley and W. P. Bowen. and reliable information obtainable. When graphic records could not be made, the observations were taken as frequently as was neces- sary to determine the general course of the changes in question, and followed clear through as in case of graphic records. We believe that it is only upon the basis of the knowledge gained by a vast amount of such careful and thorough study that the problems of physical education can ultimately be solved. The idea of securing a continuous record of the output of carbon dioxide was first suggested by a difficulty in explaining the changes in pulse rate resulting from muscular work. With all vigorous work we find, as has been previously stated,! two well marked stages of increase in pulse rate, which are often separated by a period of uni- form rate. First there is an immediate and rapid rise called the primary rise, and later a more gradual secondary rise. The primary rise, for reasons stated elsewhere,” must be due to nervous regulation of the heart. The secondary rise is not well understood, but is pos- sibly due to waste products. Since carbon dioxide is one of the chief waste products of muscular metabolism, it was thought that a graphic record of the changes in its output, placed alongside of the curve of pulse rate'in question, might throw light on the problem, and could not fail to be of interest. , The problem before us, then, was that of determining accurately the changes in excretion of carbon dioxide simultaneously with changes in the rate of the pulse, during successive periods of rest, vigorous muscular work and recovery. METHODS. Before going on to devise a method for ourselves we made a care- ful study of existing respiration methods. The methods already in use may be classed as: (1) Respiratory chamber methods and (II) Mask or mouth-piece methods. I. Respiratory chamber methods. — These methods are all modifica- tions of that of Regnault and Reiset.? As originally constructed, this apparatus was intended exclusively for experiments upon small 1 Bowen: Contributions to Medical. Research, University of Michigan, 1903, pp- 462-493. 2 BowEN: Jézd., p. 475. 3 REGNAULT and REISET: Annales de chemie et physique, Paris, (849, xxvi, p- 299. Changes in the Excretion of Carbon Dioxide. 313 animals. It was later improved by Hoppe-Seyler! and adapted to scientific experiments upon larger animals, such as sheep and dogs. A little later it was greatly improved by Pettenkofer? and from that time until the present has been extensively used in’ experiments upon man. It was further improved by Tigerstedt,? was given a capacity of a hundred cubic metres and has since been employed by Johans- son,* Atwater and Rosa,® and others. Some of the good points of this apparatus are as follows: First, it admits of experiments of indefinite length. Second, it admits of making experiments upon eighteen or more persons at once, thus enabling the experimenter to obtain average values. Third, in its most complete form as employed by Atwater, it performs the work both of a respiration apparatus and of a calorimeter, giving results which are comparable in accuracy to those obtained by the use of the combustion calorimeter and the combustion furnace. It is evident, however, that notwithstanding the many good features of this apparatus it is not at all adapted to the solution of such a problem as lay before us, since no known method for the sampling and analysis of the expired air would per- mit of a sufficiently accurate determination of the amount of carbon dioxide excreted during the sudden changes accompanying the be- ginning and the ending of vigorous muscular work. Since this re- search was begun there has appeared a new method by Jacquet.® This is also a respiratory chamber method but the chamber is greatly reduced in size, thus permitting a determination both of the carbon dioxide excreted and of the oxygen absorbed. According to this method, however, the subject is obliged to recline, and the apparatus is therefore quite unsuited to the performance of bicycle work such as was contemplated in our research. II. Mouth-piece methods.— 1. Method of Speck.’ According to this method the subject, with closed nostrils, breathes through a mouth- piece, air drawn from a calibrated spirometer, the expired air being collected in a second spirometer. At the close of the experiment a 1 HopPE-SEYLER: Archiv fiir die gesammte Physiologie, 1876, xxi, p. 18. * PETTENKOFER: Annalen der Chemie und Pharmacie, 1862-63, Supplement Band ii, p. 17. ; 8 TIGERSTEDT: Skandinavisches Archiv fiir Physiologie, 1895, vi, p. I. 4 JOHANSSON : /ézd., 1901, xi, p. 273. ° ATWATER: United States Department of Agriculture bulletin, 63, 1899. § JACQUET: Verhandlungen der naturwissenschaftlichen Gesellschaft in Basel, 1903, xii, p. 18. 7 SPECK: Physiologie des menschlichen Athmens, 1892, p. 95. 314 G. O. Higley and W. P. Bowen. sample of air is drawn from the expiration spirometer and its percen- tage of carbon dioxide and of oxygen determined by absorption with barium hydroxide and pyrogallol respectively. This method seemed unsuited to our purpose since (1) the spirometers would need to be inconveniently large in order to contain air necessary for a work experiment lasting an hour or more; and (2) only the total carbon dioxide excreted could be determined while it was necessary for our purpose to determine the character of all changes. Moreover, as it seems to us, respiration with closed nostrils, through a mouth-piece, can scarcely be called normal. Some experiments made by ourselves with one of these mouth-pieces satis- fied us that the resulting dryness of the mouth and throat renders the method quite objectionable. 2. Method of Geppert and Zuntz1 According to this method the expired air is forced through a carefully calibrated gas-meter and its volume accurately measured. Samples of the air are taken by means of a special sampling device which is operated by the gas-meter itself. A number of tubes with capillary at the upper end are filled to the tip with acid water. These are connected to a lowering device which is driven by a belt running over a pulley on the main axis of the gas-meter. As the air passes through the gas-meter the pulley re- volves, the levelling tube connected with the collecting apparatus is gradually lowered and the collecting tube is thus filled with air, whose composition has been found to represent quite accurately that of the air passing through the meter. These samples of air are then ana- lyzed for carbon dioxide and oxygen, and there is thus obtained both the carbon dioxide excreted and the oxygen absorbed, giving, of course, the respiratory quotient. This is an excellent method and to it we are indebted for a large amount of our knowledge respecting respiration. However, by this method; it would be necessary either to multiply greatly the number of absorption tubes and thus greatly complicate the gas analysis, or else the samples would be taken through such long intervals of time that changes in rate could not be determined at all; the objection urged against the mouth-piece of Speck also holds in respect to that of Geppert and Zuntz. This method, therefore, was rejected as not being suitable to our purpose. 1 GEPPERT and ZuntTz: Archiv fiir die gesammte Physiologie, 1888, xlii, p. 196; ZUNTzZ and SCHUMBURG: Physiologie des Marches, p. 209. Changes in the Excretion of Carbon Dioxide. 315 3. Method of Hanriot and Richet+ The method of these investi- gators is beautiful in principle. The outside air is drawn through an accurately calibrated gas-meter, is then inspired by the subject and expired through a second gas-meter. It now passes through an ab- sorption apparatus charged with concentrated potassium hydroxide solution which dissolves the carbon dioxide, after which it passes through a third gas-meter. If v represents the volume of inspired air, v, that of the expired air, and v, that of expired air deprived of car- bon dioxide, it is evident that v,; minus v, represents the volume of carbon dioxide excreted, and v minus v, represents the volume of oxygen absorbed. This apparatus, however, offers the objection that the subject is obliged to force the air through three gas-meters thus throwing a large amount of extra work upon the lungs. Further- more the accuracy of the method depends upon the uniform tempera- ture of the liquid in the gas-meters and upon uniform composition of the expired air. Some experiments made by ourselves tended to show that this apparatus also was quite unsuited to our purpose, and it was therefore rejected. The physiologist, the physicist, and the engineer have made exten- sive use of the graphic method. It was not until recently, however, that the chemist seriously turned his attention toward a similar application of this method to a study of the course of chemical reac- tions. In 1899 Ostwald,? while engaged in an investigation of the remarkable behavior of chromium toward acids, was led by the great expenditure of time required to note at frequent intervals the indica- tions of a gas-burette, to devise an apparatus which should record automatically the rate of evolution of hydrogen. The gas was caused to flow from the generator through a long capillary tube, a pressure thus being produced approximately proportional to the rate of evolution of the gas. This pressure was caused to actuate a light lever by means of an ordinary tambour, and the curve of rate of solu- tion of the metal was thus recorded upon a strip of paper. With this apparatus (the original chemograph) Ostwald demonstrated the periodic character of the chemical action; the effect upon the reac- tion of changes in temperature and of concentration of acid; the effect of numerous reagents on the periodicity of the reaction; and the synchronism of changes in the rate of chemical action with those 1 HaAnRIoT and RicHET: Annales de chemie et de physique, 1891, pp. 22, 495. 2? OsTWALD: Zeitschrift fiir physikalische Chemie, 1900, xxx, pp. 33, 204. 316 G. O. Higley and W. P. Bowen. of the electrical tension of the metal. A consideration of Ostwald’s papers leads to the conviction that the graphic method alone could have yielded: such satisfactory results in the study of a problem of this kind. So far as the writer was aware when this research was begun, no successful attempt had ever been made to determine the rate of a chemical change by recording the movements of a balance beam. FriGuRE 1. — The respiration apparatus except mask, chemograph, and pump. Out-door air enters at O. F is the bag. J is a condenser. The drying apparatus and guard tube are shown at G. A and DY connect with the tubes of the chemograph. T is an auxiliary tube with adjustable valve, by means of which the flow of air through the main circuit may be regulated. However, in March, 1904, Professor G. N. Stewart, of Chicago Uni- versity, while examining the apparatus to be described in this paper, stated that he had some time previously demonstrated the change in weight of a dialyzer filled with cane-sugar solution and suspended from the arm of a balance in a solution of pure water. By means of a lever attached to the arm of the balance, a curve of change of weight of the dialyzer was recorded upon a drum. Professor Stewart’s paper was read at a meeting of the Chemical Society of Owens College, Manchester, England, but was published only by title. The occasion for the construction of this apparatus arose in con- nection with an attempt to determine the cause of the secondary rise Changes in the Excretion of Carbon Dioxide. 317 in the pulse rate in man during uniform muscular work. It became desirable in the progress of this research to compare the pulse curve with the curve of excretion of carbon dioxide during successive periods of rest, uniform work, and recovery. Almost at the outset it was suggested by Prof. W. P. Lombard that it might be possible by absorbing the exhaled carbon dioxide in an absorption apparatus suspended from the arm of a balance, to write a curve of carbon dioxide excreted from the lungs. Experiments were accordingly made which resulted in the foregoing apparatus. _ Mask. The subject respires through a mask! made as follows: A copper wire about 2 millimetres in diameter is so bent as to fit over the bridge of the nose and the face inclosing nose and mouth. A piece of heavy tin is then bent in the same form, that of an ovoid about I1 centimetres in length and 8 centimetres broad at the widest part. This is soldered to the wire, making a box about 4 centimetres in depth and rather closely fitting the face. The space between this edge and the face is made air-tight in the following manner: A rubber tube, such as is used on the Townshend ether in- haler, is stretched over the wired edge of this box and by inflating the rubber and closing the tube by means of a clamp, it is pos- sible to bring a cushion filled with air between the face and the wired edge of the mouth-piece. A sheet of rubber is stretched over the front of the box, and firmly cemented and wired in place. This is then pierced in the centre and through it passes a glass tube 1.2 centimetres in diameter, which is attached to the valve- chamber. Especial care was taken to make the volume of the tubes be- tween the mouth and the valves as small as possible. The mask is held firmly to the face by wide elastic bands passing around the head. Valve-chamber. (V, Fig. 1.) The valve-chamber employed is of the same general form as that used by Zuntz and Schumburg,? ex- cept that it is made of glass instead of metal, thus permitting a view of the working of the valves. It consists of a large 7 tube, 20 centimetres in length and 4 centimetres in diameter, with a side tube 1.2 centimetres in diameter, to which the mouth-piece is attached. The valve seats are of cork covered with thin sheet-rubber fastened 1 The idea of this mask was obtained from Mosso: Der Mensch auf den Hochalpen, p. 170. 2 ZUNTZ and SCHUMBURG: Loc. cit., p. 208. 318 G. O. Higley and W. P. Bowen. on with rubber cement. The openings are about 1.5 centimetres in diameter; the valves are made of thin sheet-rubber stiffened above with a disk of very thin aluminium foil attached by means of rubber cement. The out-door air enters the lower end of the valve-chamber through a wide glass tube. From the valve-chamber the air passes into the inner bag of a small football, holding when moderately distended about 3 litres. At each expiration this bag is somewhat inflated, but is deflated through the chemograph by the action of the pump during the next inspiration. There is thus a substantially uniform delivery of air to the absorption apparatus. At this point there is placed a 7 tube and shunt so that the air may be caused to pass directly to the pump without passing through the absorption apparatus. Drying tubes. The apparatus for the removal of moisture consists, essentially, of a Y tube 75 centimetres in length and 4 centimetres interior diameter, filled with coarse pumice stone saturated with con- centrated sulphuric acid. This tube is followed by a guard tube G, about 75 centimetres long, filled in the same manner. The complete- ness of the action of the preceding tube may be seen in the fact that the guard tube in no case gained more than .o1I gram, and usually Jess than 0.005 gram, during an experiment in which air saturated with water vapor and flowing at the rate of 30 litres per minute passed through the train for 30 minutes. From the guard tube the air flows through the absorption beaker of the chemograph, passing then through two guard tubes G’ and G” filled with pumice stone and sulphuric acid. The first of these tubes shows, in an ordinary work experiment, a gain of only 0.05 gram; the weight of the second remains practically unchanged. The air passes now, at the will of the operator, through a shunt tube! containing clear lime-water as a test for the presence of carbon dioxide, then through an Elster gas-meter and to the pump. Suction pump. In order to relieve the lungs of the subject from the labor involved in forcing the expired air through the tubes and gas-meter, the latter is connected to the suction side of a blower, capa- ble of drawing air through the entire apparatus at the rate of 30 litres per minute. This amount of air is sufficient for a subject engaged in moderate muscular work, but its rate is not equal to that at which air passes from the lungs during the act of expiration. The bag pre- viously mentioned is introduced into the circuit in order to permit 1 Not shown in the figure. Changes in the Excretion of Carbon Dioxide. 319 the subject to exhale freely, the air expelled at one expiration being drawn from the bag by the pump during the next inspiration. Chemograph. The dry air containing carbon dioxide passes in the following manner into an absorption apparatus upon the balance: The tube 4 (Fig. 1) is connected by means of a short piece of very thin rubber tubing, made of a surgeon’s finger-cot, with a copper tube (B, Fig. 2) 1.5 centimetres in diameter, which is firmly attached to the beam of the balance. The air enters directly opposite the central knife-edge, passes to the outer end of the beam, and down- ward opposite the lateral knife-edge through two very thin rubber connections and a glass tube (JD, Fig. 3) into the absorption appa- ratus below. From the absorption apparatus the air passes upward through similar connections to the balance-tube C, back on the op- posite side of the balance-beam to the centre, then into guard tubes, G' and G", which have already been described. (See Fig. 1.) Absorption apparatus. (C, Fig. 3.) Several forms of absorption apparatus are used in connection with the chemograph. That con- structed for use in work experiments has been most employed and will be described here. | Since there was a question of removing the carbon dioxide from . air flowing at the rate of 30 litres per minute during work, the ab- sorption apparatus is necessarily large. It consists of a beaker 20 centimetres in diameter at the top, and 50 centimetres deep, with a cover of thin copper, provided with openings two centimetres in diam- eter, into which are fitted the inlet and outlet tubes. The air passes downward into the beaker True 2A houte econ trout through a thin glass tube 2 centi- passes through those portions only metres in diameter, to within about which are designated by arrows. 2 centimetres of the bottom of the beaker, ending in an open space 3 centimetres deep and of a diameter equal to that of the beaker. (This open space was left be- cause it was thought that the carbonic acid gas would thereby be more uniformly distributed throughout the whole cross-section of absorbent placed above.) The air now rises through about 5 kilo- grams of coarse, carefully screened soda-lime and then through glass- wool covered with phosphorus pentoxide to hold back dust and the 320 G. O. Higley and W. P. Bowen. last trace of water formed in the reaction. This beaker when charged weighs about 5.5 kilograms. It is counterpoised by another beaker of the same exterior volume filled with spent soda-lime. Recording apparatus. In order to record the movements of the balance, there is attached to the end of the balance-beam a steel loop which engages the short arm of the light lever (Z, Fig. 3), made of two straws placed side by side and tipped, at the short end, with a steel wire, and at the long end with a piece of parchment paper. By means of an arrangement (/, Fig. 3) which will be clear from the gu {Seis Gs : ; nies FIGURE 3.— This figure represents an elevation of the chemograph. The air enters at 4 and takes the direction shown by the arrows. Cis the absorbing apparatus, and C’ the counterpoise. Four gram-weights are placed upon the left beaker, and the lever- points thereby deflected upwards on the drum (7). The curve of carbon dioxide absorbed in C, is written downward to the right upon the drum. figure, the fulcrum of the lever may be adjusted vertically, trans- versely, and horizontally. On the short arm of the recording lever (40 millimetres in length) there is placed a movable weight by an adjustment of which the long arm (350 millimetres in length) is made to slightly preponderate. The lever records upon the drum the movements of the balance-beam magnified nine times. Since much depends upon the accurate adjustment of the writing lever upon the paper, the kymograph is set upon a base provided with ball- bearings and with two springs working against a screw, so that the Changes tn the Excretion of Carbon Dioxide. 321 kymograph may be rotated around a vertical axis and thus the drum be quickly and accurately adjusted to the recording lever at any time. Attached to the frame of the kymograph is a vertical brass rod! to which are clamped three slender brass springs which extend horizontally and whose points may be brought in light contact with the paper on the drum. The middle one marks the level of the centre of the fulcrum of the recording lever. This marker when once ad- justed is, of course, never disturbed. The upper and lower ones draw lines marking the upper and lower limits of the excursion of the lever- point during calibration. They are readjusted from time to time as may be necessary. It has been already stated that the rubber connections of the balance were made very light in order to avoid, as far as possible, interference with the free movements of the balance. In order now to be able to write a curve of considerable length, representing, for example, a mass of five grams, it became necessary to diminish by some means the sensibility of the balance while interfering as little as possible with the uniformity of its movement. There was, there- fore, attached to the frame of the balance, about ten centimetres from the central knife-edge, a steel yoke (A, Fig. 3) passing over the beam. From this yoke there was suspended a coil, five centi- metres in length and about one centimetre in diameter, made of phosphor-bronze wire approximately half a millimetre in diameter. This coil was attached at its lower end to the upper side of the beam of the balance by means of a hook, which together with the yoke could be set at any desired distance from the central knife-edge. A set screw, with check-nut by which the upper end of the coil is attached to the yoke, admits of an adjustment of its tension at the will of the operator. Adjustment of tension. The balance is brought into equilibrium with spring disconnected. The spring is now attached to the beam and brought into a state of tension by turning the screw at the upper end of the coil. Weights are now placed upon the pan on the same side until equilibrium is again restored. In most of the work done with this apparatus the spring has had an initial tension of four grams. We have here a combination of the beam and the torsion balance. This apparatus was subjected to a series of careful tests in order to 1 Not shown in the figure. 322 G. O. Higley and W. P. Bowen. ascertain whether the records inscribed by it upon the blackened paper were of any value. Tests of the balance.1— 1. Zest with all connecting tubes removed and with a load of 5.5 kilos. In order to ascertain the accuracy with which weighings could be made on this balance, various objects were weighed upon the balance, and then to tenths of a milli- TABLE I (A). Series I. Series II. | Series III. Average. Wipper eran en nnnne Sa 15a 14.7 15.06 Second gram ~4 5. ©3)\5 15.3 15.0 15.0 INS) Aphixd ornament see 15.2 5r3 14.7 15a hourthteram/) enn) 14.6 14.6 14.7 14.63 Ib eAeN 4 6 G5 4 6 6 (14.2) (13.9) (13.3) Total'deflection 4). 60.5 60.0 59.1 TABLE I (B). Series I. Senes il, || jsenessaue Average. Wimr min 5 5 6 5 6 14.4 14.7 14.7 14.6 Secosdieram y.46 coe ier 14.8 15.0 15.03 ahirdscramy eyes ieee ee 14.8 14.8 14.5 14.7 Hourthesrame gles eee 14.7 14.6 14.5 14.6 TURN, frehey g 5 3 5 a oe (14.0) (14.5) (14.4) Total deflection ... . 59.0 58.9 58.7 gram upon a fine balance. It was found that the variation of the two weights was, in no case, more than 0.0103 gram and averaged 0.005 gram. 2. Calibration of the apparatus. The tubes and recording lever were now adjusted, and a careful test was made of the amount of vertical deflection of the end of the recording lever produced on the 1 This balance was made by RUPRECHT, and is a most satisfactory instrument. Changes in the Excretion of Carbon Dioxide. 323 blackened drum by a mass of five grams. This was done as follows: The balance was brought into equilibrium. The beam was now arrested and five gram-weights were placed upon the counterpoise beaker; this produced an angular deflection of the beam of about 1° 45’, and a vertical deflection of the recording lever-point of about 73 millimetres. After a delay of about 30 seconds to allow the lever to assume its position of rest, the screw controlling the position of the drum was carefully turned until the blackened paper was brought into light contact with the writing lever, and the kymograph was started and allowed to run until a short horizontal line had been drawn by the point of the recording lever upon the paper. The beam was now arrested, the weights removed, the beam again released, ex | u ll | >< me. 1 Ill = ry waa [= 14 ' 14 i144 Pu feu Frcures 4 A, and 4 B. — These figures show the results of calibration of the chemograph with 5 gram-weights. The distances ¥ X’ are the vertical deflections of the recording lever-point for a mass of 5 grams added to the pan. The numbers represent the vertical deflection in millimetres of the lever point for ove gram. A calibration of at least one series is made at the beginning of each experiment, and often at the close also. and the writing lever again allowed to come into a position of equilibrium. The kymograph was now started as before and a second light horizontal line drawn upon the paper. The vertical deflection of the writing point is a measure of the mass of five grams. This process was repeated many times, the weight being alternately added and removed to find out the accuracy with which the point of the writing lever returned to the same level on the drum. It was found that at the beginning of work, after the apparatus had stood for some hours, there was some irregularity at the first two or three move- ments of the beam. However, after a few minutes the movements became quite uniform. Starting, now, from the highest position of 324 G. O. Higley and W. P. Bowen. the recording lever, the gram weights were removed, one by one, the position of rest of the point of the lever being marked at each step by a short horizontal line as in the preceding case. Fig. 4 A and Table I (A) show the results of one of these calibra- tions in which are given the deflections due to one gram. Omitting the lower or fifth gram of each series in I (A) and we have the following averages: 15.06-, 15.1-, 15.1-, 14.6. The total deflection for four grams is, in the three series, 60.5, 60, and 59.1 millimetres, respectively. The extreme variation in deflection for four grams is 1.4 millimetres or 2.3 per cent, and the greatest variation from the average is 0.7 millimetre or 1.15 per cent. The greatest variation in deflection for a single gram (omitting the lower or fifth gram in each series) is 0.8 millimetre or 5.2 per cent; the greatest variation from the average is 0.36 millimetre or 2.4 per cent. Figure 4 B and Table I (B) show the results of a calibration of the same apparatus, with slightly different tension in the spring. In this case the greatest variation in deflection for 4 grams and 1 gram are 0.3 millimetre or 0.5 per cent, and 0.6 millimetre or 4 per cent, respectively, and the greatest variation from the average, 0.2 millimetre or 0.32 per cent, and 0.38 millimetre or 2.5 per cent. The results of numerous cali- brations showing that the deflection for the lower or fifth gram in- variably gives low values, the use of this portion of the arc has been discontinued. 3. The next test was carried out as follows: A small crystalliz- ing dish, previously weighed upon a fine balance, was placed upon the right beaker and the balance brought into equilibrium. Four gram- weights were now added to the left beaker and the point of the lever thereby deflected vertically about 58 millimetres upon the drum. The usual calibration with 4 grams having now been made,- the drum was started and a slow stream of mercury was allowed to flow into the crystallizing dish from a simple apparatus with capillary delivery-tube. There was thus described upon the drum a short horizontal line, followed by a more or less regular curve inclining downward toward the right. Finally, when about two grams of mercury had been allowed to flow in this manner into the crystalliz- ing dish, the addition of mercury was discontinued, and the curve caused to end in a horizontal line. The beam was now arrested and the crystallizing dish removed. The vertical deflections of the writing lever due to the successive addition and removal of four grams in the initial calibration process were now determined. The Changes in the Excretion of Carbon Dioxide. 325 average of these values is, of course, the graphical equivalent of four grams. From this there was readily obtained the modulus of the balance, viz. the number of milligrams represented by one millimetre of vertical distance upon the drum. The vertical distance between the initial and the final positions of the writing lever in the experiment with mercury was now measured, and TABLE II. SHOWING THE RESULTS OF CALIBRATION OF BALANCE WITH MERCURY. Vertical deflec- tion for one gram. millimetres Modulus? M. Initial height Final height Weight graphic- ally deter- mined (h—h’) M. millimetres millimetres Weight on fine balance. Per cent error. 11.63 11.63 11.63 11.63 14.57 14.57 14.57 55.9 26.4 58.5 34.6 66.9 0.0859 0.0859 0.0859 0.0859 0.0686 0.0686 0.0686 0.0686 0.0686 0.0686 0.0686 0.0686 82.5 55.9 83.5 58.5 88.0 66.9 87.8 59.9 88.0 2.287 2.537 2.148 2.055 2.273 2.5695 +0.014 +0.0326 —0.0144 +0.0152 +0.014 +0.0065 +0.0057 +0.001 —0.024 +0.61 +1.26 —0.66 +0.70 +0.98 +0.29 2.1624 2.0707 1.4331 2.2172 1.447 34.5 2.2237 59.9 24.3 76.0 62.3 Bohs 42.7 T9149 1.9092 Dar Vis 0.847 0.9282 +0.3 IGE Bi 14.58 14.59 2.4434 0.823 0.9396 0.953 LS?) +0.04 —2.8 76.0 73.3 +0.0114 +0.0032 +0.0213 +1.23 14.59 0.9488 +-0.40 14.59 59.3 1.177 -+1.9 1 Experiments 1-4 were made by the use of a brass spring; the following experi- ments with a phosphor-bronze spring of quite a different tension. the widely different values of the modulus. This accounts for the weight of. mercury added, obtained by multiplying the modulus by this value. Finally the crystallizing dish with its contents was reweighed upon a fine balance and the weight of the mercury thus determined compared with that obtained by the graphical method. The results of a series of such tests are shown in Table II, in which there are given: Vertical deflection in millimetres of lever-point for 326 G. O. Higley and W. P. Bowen. one gram added to the pan; Modulus J/ or the weight in grams corresponding to one millimetre deflection ; Initial height of writing lever-point “h”; Final height of lever-point “h’”; Weight of mercury as graphically determined, (h-h’) M; Weight of mercury as determined on fine balance; Error + or —; and per cent of error. It will be observed that the error was generally positive and varied in the different experiments between 0.001 and 0.0326 gram, the average error being about 0.014 gram or approximately 0.7 per cent: 4. Test with a uniform current of carbon dioxide. A current of carbon dioxide, as uniform as possible, was now caused to flow for seven minutes through the apparatus, the drum meanwhile revolving uniformly. The result was a smooth curve about 30 centimetres in length, which was a very close approximation to a straight line, measurements showing that at no point was the deviation from straight line greater than 0.7 millimetre. This experiment was repeatedly performed with practically the same result. It is evident that when the recording lever has reached the lower limit of the arc the beam may be arrested, and additional weights added without interrupting the experiment and with a loss of only a small portion of the curve. This may be repeated for hours, thus enabling the operator to determine both the course of the reaction throughout its whole extent and the total weight of gas absorbed. The method of determining the rate of the reaction during any period is as follows: With a radius equal to the length of the long arm of the recording lever, and with the proper points on the central refer- ence line as centres, arcs are drawn cutting the time line at the begin- ning and the end of the desired period, and also the curve of carbon dioxide. The vertical distance between the two intersections of the carbon dioxide curve by these arcs is the measure of the amount of that gas absorbed during the time cut off below. The modulus of the apparatus having been determined as described earlier, the rate of absorption between the desired limits is readily determined. 5. Test with a weighed quantity of carbon dioxide. A series of experiments was now carried out with carbonic acid gas. There was set up a carbon dioxide apparatus consisting of a small fractionating flask provided with a dropping funnel and delivery tube, to which was attached a drying tube filled with pumice stone and sulphuric acid. Into this flask there was brought a quantity of a saturated solution of sodium carbonate, while sulphuric acid was placed in the dropping Changes in the Excretion of Carbon Dioxide. 327 funnel. The apparatus was now carefully weighed, after which it was attached to the drying tube of the chemograph, a current of pure air free from carbon dioxide was drawn through it at the rate of half a litre per minute, and the sulphuric acid was slowly dropped upon the carbonate. The gas thus produced, diluted with seven litres per minute of purified outer air was drawn through the absorption ap- paratus, the kymograph drum meantime revolving at a uniform rate. This experiment was repeatedly tried, with the result that the maximum error was 2.8 per cent. It is evident that this form of chemograph may be used in a study of the course of many chemical reactions in which gas or vapor is evolved, since the apparatus will write a curve of Joss in weight as readily as of gain in weight. It is only necessary to place the gen- erator upon the pan of the balance, to make the usual adjustments, and allow the process to continue as long as desired. A tracing of the course of a reaction in which there is an escape of hydrogen will, perhaps, not be practicable with this apparatus, on account of the relative lightness of that gas. However, the curves of rate of loss of water, ammonia, carbon dioxide, etc., will be readily written. Some work in which the apparatus has been used in this manner has already been done, the results of which will be published later. THE EXCRETION OF CARBON DIOXIDE RESULTING FROM BICYCLING. I. General form of the curve——About twenty experiments were made to find the general course of the changes in output of carbon dioxide resulting from work. The work was done upon the stationary bicycle described in the papers already published in this series.’ It is sufficient to say here that the machine is driven by the subject at a chosen rate which is recorded continuously, and that the resistance of the machine can be adjusted at any desired amount. The graphic record of carbon dioxide excretion is taken on a slowly revolving drum along with records of respiratory movements, revolutions of the bicycle, and time in seconds. Ona loop of paper passing around two drums and moving more rapidly, the pulse is recorded along with a time curve, giving seconds. In beginning an experiment, the sub- ject puts on the breathing mask, mounts the wheel, the pneumograph and pulse tambour are put on and tested, and then the mask is con- 1 BowEN: Loc. cit., fe 465. 92m) G. O. Higley and W. P. Bowen. nected to the tubes leading to the apparatus for determining carbon dioxide. The drums are started, and the subject sits quietly on the bicycle until sufficient records have been made to show the normal rate of pulse, breathing, and output of carbon dioxide. Then at a signal from the experimenter he begins to drive the wheel in unison with a metronome placed before him, and the records continue. The balance is handled as described in the preceding section. The electric signal connected with the bicycle indicates the moment of starting and stopping, and the speed of revolution. On cessation of the work the records continue until the carbon dioxide has resumed approximately the rate before the work began, as indicated by the slant of the line described by the recording lever. The experiment then ends. A record taken in this way is shown in Fig. 5. By looking at Fig. 5 one can observe in a general way the effect of the work on the output of carbon dioxide. Beginning at the left, the ey i vl mi ii wit) ny i oo A hha Hie i na : \ FiGcureE 5.— Graphic record of carbon dioxide excretion during bicycling. Read from left to right. 2, respiratory movements. CQO,, record made by chemograph. The descending lines, as AC, are due to accumulation of carbon dioxide in absorbing beaker. The ascending lines, as CC’, are due to addition of weights by operator. T is the time, marked every ten seconds. JZ, revolutions of the bicycle crank. nearly straight slanting line A 4, drawn by the lever connected with the balance, indicates the rate of output during rest. Then, when work begins, the increased slant of 4 C indicates an increased excre- tion, and this continues to increase for about two minutes. From this time to the end of the working period the slant of the line re- mains about the same, indicating uniform output of carbon dioxide. As soon as work stops there is an immediate change in the slant D £, showing the diminished excretion. There are some irregular- ities in the line we are considering, but they are not sufficient to interfere with an accurate reading of the curve. These disturbing features result from vibrations of the balance caused by placing on of weights, and by vibrations of the building due to various causes. Changes in the Excretion of Carbon Dioxide. 329 A more accurate idea of the changes in question can be obtained from Fig. 6, which was obtained from the record of Fig. 5 by careful measurements and plotting. This figure also contains the plotted curve of pulse rate, which was recorded simultaneously on another paper. The pulse curve shows plainly a rapid primary rise, a6, a plateau, 4c, and a slow secondary rise, cd. The curve of carbon dioxide rises rather rapidly during the first two minutes, which in- cludes the period of rapid rise of pulse rate and a part of the plateau. During the remainder of the working period the output is seen to be practically constant, although the pulse rate is rising for the latter half of the time. On cessation of work the output diminishes until at the end of two minutes it has returned to practically the original rate. At the same time the pulse rate, although falling rapidly at first (de) is oscillating about a rate (ef) 20 per cent above the normal. These results are fair examples of those obtained in the series. A few ex- In Figure 6, p. 329, of Volume XII, the ordinates for carbon dioxide beginning at the base line are: grams 0).7 14 2.1 2.8 20) The excretion per minute before the beginning of work was 0.42 gram. Minutes U U4 4 C4) 5 ju u z bicycle, the ae of starting FicurE 6.— Plotted curves of pulse rate and being accurately indicated by excretion of carbon dioxide from same ex- the marker 4, The experi- periment as Fig. 5. Broken line, pulse rate ; ment continues only lone solid line, carbon dioxide. Ordinates give 5 enough Pe aebotuatanureonie a eta and grams of carbon crease in the output of carbon dioxide. After fixing the record in shellac, the point 2 on the curve of carbon dioxide, where the line first changes its direction as the result of the work, is revolved to the base line with a radius equal to the length of the long arm of the writing lever, so as to avoid error 330 G. O. Higley and W. P. Bowen. due to rotation of the lever on its axis. Now the number of seconds intervening between the beginning of work and the resulting change in output of carbon dioxide can be readily obtained from the time record, 7. To find the result desired we must deduct from the time found in the manner just described the time occupied by the passage of the exhaled air from the mouth and nostrils through the mask and con- necting tubes, and sufficient additional time to collect in the soda- lime enough carbon dioxide to overcome the inertia of the balance. The time to be deducted is found as follows: while sitting quietly upon the bicycle, with the record in progress, the subject holds his breath for several seconds. The result is shown in curve C of Fig. 7. The pneumograph curve at the top of the record shows when the breath is held. Soon the lever which records the movements of the balance ceases to fall, as indicated by its writing a level line w v. When the subject begins to breathe again the pneumograph curve shows the exact moment of the first expiration, and the time from this point to the point v, where the carbon dioxide lever first begins to fall again is the time of delay due to the apparatus. From a large number of tests this time was found to be close to 6 seconds. This delay evidently depends upon the rate at which air is drawn through the apparatus by the suction pump. For this reason all the experi- ments on latent period were made with the air moving at the uniform rate of 20 litres per minute. Making the deduction of 6 seconds in the case recorded in curve A of Fig. 7, we obtain 5 seconds as the latent period. This period varied considerably in different cases, all the way from 3 to 14 seconds. Now it is evident that none of these results is long enough to cor- respond with what we hoped to find, viz., the time required for the carbon dioxide formed in the muscles at the time of the first mus- cular contraction to reach the outside air. The carbon dioxide must first diffuse into the blood from the tissues where it is formed, then traverse the venous half of the systemic circulation, the right side of the heart, and the arterial half of the pulmonary circulation, and finally diffuse into the air of the alveoli before any of it can appear in the breath. From the latest conclusions of Stewart ! and others who are considered as authorities on the time of the circulation, we learn that from 15 to 20 seconds is the least possible time for the blood to travel 1 STEWART: Manual of physiology, Philadelphia, 1900, p. 124. Changes tn the Excretion of Carbon Dioxide. 331 this distance, to say nothing of the diffusion time. We must evidently account for the shortness of the latent period thus found. The explanation is provided by two facts which appeared in the course of these experiments, showing that the prompt increase on beginning work is due to an increase in the depth of respiration which accompanies the beginning of work. The fact that there is a prompt increase in depth of respiration when work is begun has been mentioned by Tigerstedt.1 We discovered the increase in this way: in the description of apparatus in the preceding section, men- tion was made of a light rubber bag inserted into the air circuit be- tween the mask and the absorbing beaker for the purpose of holding temporarily a part of the air expelled at each breath, and thus pre- venting excessive pressure in the lungs during expiration. With each breath exhaled the puff of air partly inflates this bag, and the B € i wecaen i Mutt Weyer TAY yr, fr, Mi m n P SS nt oT ao Ea mere ohana te a sDLere ns McesbesD Gerpcen roe eA a a a a aaa aa aan M Ficure 7.— Excretion of carbon dioxide during work and during modified respiratory movements. A, respiratory movements; CQ,, record of chemograph; 7, time in ? seconds; J/, bicycle. Curve A shows changes due to beginning work; curve B, deepened respiration followed by work; curve C, holding the breath. action of the suction pump empties it again while the next breath is being inhaled. Now, by a careful observation of the amount of infla- tion of the bag, one can detect any considerable change in the volume of successive breaths. Upon watching the bag at the time of begin- ning work it was noticed that a marked increase in the volume of each breath occurs promptly in almost aN cases as soon as the subject begins work. Upon observation of this fact the question at once suggested itself, how much effect a voluntary increase in depth of respiration would have if not associated with work at all. The answer is given by + TIGERSBEDT 942.02 C27.,;p.1170. 332 G. O. Higley and W. P. Bowen. curve B of Fig. 7. Line g g indicates the rate of output of carbon dioxide with normal breathing. At g the subject began deep inspira- tions, with the increased output indicated by the greater slant of the line g x This marked increase, which is temporary, is to be ex- plained in all probability by the greater amount of alveolar air, richer in carbon dioxide, which is exhaled during deeper respiration. The fact thus noted suggested in turn a method of attacking the original problem. Ifa subject ordinarily increases the depth of his respira- tion on beginning a piece of muscular work, and thus introduces an element that obscures the point at issue, viz., how soon the direct effects of the work show themselves, why not voluntarily increase the breathing before beginning the work, and thus separate the two effects ? Several experiments were made in this way and the results are quite satisfactory. Curve B of Fig. 7 shows the result in one instance. The effect of the increased respiration is clearly marked, the new rate of output g7 being sharply defined from the normal rate ? @ preceding it. The further increase on beginning work (7s) is not so prompt in its appearance and comes on more gradually, reach- ing its maximum after a minute or more, depending on the work. In these experiments the latent period of increase on starting the work was from seventeen to twenty-two seconds. It is evident that the latent period will vary with the rapidity of the circulation and also with the rapidity of diffusion, so that a more definite figure is not to be expected. DISCUSSION OF RESULTS. These results agree in the main with those obtained by others, as far as they lie in the same field. Johansson! found a uniform output of carbon dioxide during uniform muscular work, but his determina- tions were made for periods of several minutes’ length, while our results show the output from minute to minute. Our work involved more extensive muscle area than his, but on the other hand we did not investigate the question during breathlessness and extreme fatigue, because the apparatus that we used did not supply air enough for a subject doing excessive work. We see no reason, however, why with a larger apparatus the same method would not apply in all such cases. In studying the manner in which the output of carbon dioxide 1 JOHANSSON: Loe. cit. Changes in the Excretion of Carbon Dioxide. 333 changes at the beginning and end of the working period, we believe we have entered a new field. Zuntz and Schumburg;! Tigerstedt,” Johansson, and Katzenstein® state definitely that they did not begin estimating the carbon dioxide for some time after the work was in progress, and no investigator, so far as we have noticed, has studied this stage before. When we consider the manner in which the change in formation of carbon dioxide occurs in the muscles, and the manner of its elimination from the body, our results seem entirely consistent with the conditions. Since the work begins suddenly at full speed and force, the formation of carbon dioxide in the muscles must jump instantly from the normal rate to the maximum rate during the work. Since the gas must first diffuse into the blood and then be carried to the lungs before elimination can take place, there should be a latent period of a little more than half the time required for a complete cir- cuit of the blood before the first waste product formed during work can be exhaled. The latent period of 17-22 seconds found here is in full accord with the work of Stewart previously mentioned. Since the latent period is dependent upon the rapidity of the circulation, which is different at different times, even in the same subject, a varia- tion of several seconds is to be expected. The sudden increase in carbon dioxide in the tissues produces a gradual increase in output from the lungs, because of the manner of elimination. The diffusion process tends to round off the sharpness of the change, and the transportation in the blood stream does so still more, since the portion of the gas that is absorbed in the blood that traverses the centre of the stream will reach the lungs con- siderably sooner than that following the walls of the vessels; there are also routes of different lengths from different muscles. The maximum output will evidently be reached when the blood leaving the muscles after work has begun and travelling by the slowest and longest route, finally gives off its gases in the lungs. Upon cessa- tion of work the production of carbon dioxide drops back to the normal rate as suddenly as it arose, but the reasons just given show why there is again a latent period, and then a gradually diminished output. A temporary increase in output sometimes found at the instant work ceases is readily accounted for by the fact that the respirations, which are more or less hindered by the muscular con- ZUNTz and SCHUMBURG: Loc. cit., p. 212. TIGERSTEDT: Skandinavisches Archiv fiir Physiologie, 1895, vi, p. 170. 1 2 8 KATZENSTEIN: Archiv fiir die gesammte Physiologie, 1891, xlix, p. 338. 334 G. O. Higley and W. P. Bowen. tractions involved in driving the wheel, become somewhat deeper when this hindrance ceases, since the tissues are still as fully saturated with carbon dioxide as at any time, so that there is for a few moments no diminution of the respiratory need. Fig. 6 shows a very slight diminution of the output of carbon dioxide during the latter part of the working period. This is probably accidental, as it does not occur in all cases, some experi- ments showing a slight increase. The change may be due to a change in the manner of doing the work — the subject driving the wheel with slightly less muscular expenditure at some times than at others. The great variation in internal work that may occur in the doing of a certain amount of external work has been pointed out in another paper of this series.1_ One possibility thought of in connec- tion with the carbon dioxide curve in Fig. 6 is the warming up of the absorbing beaker by continuous respiration through it, thus causing the air within it to become lighter, tending to counteract the in- creased weight due to carbon dioxide. As the experiments made gave a slight increase during work as frequently as a decrease, this was not considered likely. A group of longer experiments in which the air was drawn through the absorbing beaker only one minute of each five showed no difference in this respect, also indicating that change of temperature of the absorbing beaker was not the cause of the irreg- ularity. The change, however, is so slight as to have no effect on the general interpretation of the results, whatever may be the cause. When we compare the curves of pulse rate and carbon dioxide, we see plainly that the primary rise of pulse frequency coincides in time approximately with the rise in output of carbon dioxide, and the same can be said of the corresponding fall. The short latent period of the pulse shows, as has been stated elsewhere,” that if the production of carbon dioxide is in any way responsible, even in part, for the change in pulse rate when work begins, the influence must be brought to bear through nervous channels, rather than by an effect of the gas itself upon the heart or the cardiac centres. We see no reason why the prompt increase in heart action may not be due in part to sensory impulses arising in the muscles as a result of the waste products suddenly set free there, as advocated by Athanasiu? BowEN: Loc. cit., p. 485. BowEN: Loc. ctt., p. 475. 1 8’ ATHANASIU and CARVALLO: Archives de physiologie, 1898, pp. 554, 567. Changes in the Excretion of Carbon Dioxide. 335 The results obtained show no evidence of any relation of cause and effect between the production of carbon dioxide, and the secondary rise in pulse rate. In Fig. 6 we see that the output of carbon dioxide is constant during the entire period of the secondary rise of pulse rate, while the secondary fall of pulse rate during recovery is continued for a long time after the output of carbon dioxide has returned to the normal. The lack of correspondence in the two curves practically amounts to a demonstration that the secondary changes in pulse rate have nothing to do with the production of carbon dioxide and its elimination from the system. SUMMARY. 1. The problem of finding the changes in rate of output of carbon dioxide resulting from muscular work and other causes is practically solved by the method used in this research. 2. The latent period of increase in output of carbon dioxide from the lungs in case of beginning work is in the close vicinity of twenty seconds, and the increase reaches its maximum in about two minutes. 3. The output of carbon dioxide from the lungs is practically uniform from minute to minute during uniform muscular work, after the blood has had time to take part fully in the process of elimination. 4. Upon cessation of work the output of carbon dioxide decreases to the normal amount in about the time occupied by its increase, and after a like latent period. 5. The results obtained show no indication of any connection of cause and effect between the production and elimination of carbon dioxide and the secondary rise of pulse rate. The writers desire to acknowledge their great indebtedness to Prof. W. P. Lombard, without whose assistance this research could not have been carried on. ON THE ABSORPTION AND UTILIZATION) 3GE PROTEIDS WITHOUT INTERVENTION OF THE ALIMENTARY DIGESTIVE PROCESSES. By LAFAYETTE B. MENDEL anp ELBERT W. ROCKWOOD. [From the Sheffield Laboratory of Physiological Chemistry, Yale University.] Se experiments which have given rise to this paper were origi- nally planned to ascertain the fate of proteids introduced into the circulation without the intervention of the alimentary diges- tive processes. For a variety of reasons proteids of vegetable origin were selected for study. They are naturally foreign to the blood; their relatively easy preparation and purification offers advantages over the more commonly employed products from animal sources; and, furthermore, the recent investigations of the vegetable proteids selected indicate that they are by no means identical in chemical make-up with the serum- or tissue-proteids. Nevertheless, the vege- table proteids are common constituents of the food of animals, and are unquestionably utilized by these organisms. The problem at once arises, whether—and if so, to what extent — these foreign proteids must be subjected to digestive changes before they can be assimilated. The teaching regarding the transformations which proteids undergo in the alimentary tract preliminary to absorption has lately experienced considerable modification. The much-quoted experi- ments of Voit and Bauer, Eichhorst, Czerny and Latschenberger, and others! were interpreted to indicate a direct absorption of proteid from the intestine without previous digestive changes, Clinical experience with nutrient clysters contributed to strengthen this view. Accordingly various writers have assumed that while peptonization may facilitate the rate of absorption, the native proteids can readily be used without the intervention of proteolytic enzymes, provided they are introduced in soluble form. Indeed, Bunge has insisted that there is no @ priori ground for supposing that proteid is not absorbed 1 A review of the literature on this subject by I. MunxK will be found in the Ergebnisse der Physiologie, 1902, i, (1), p. 310. 336 On the Absorption and Utilization of Proteids. 337 unchanged. If, he says, fat droplets visible under the microscope, and even entire leucocytes, can leave the blood-capillaries and travel through the tissues, why may not a proteid molecule find its way through the capillary wall? The more recent literature on this subject presents a radical change in the views expressed. The discovery of the crystalline cleavage products of the proteids in the intestinal contents of the living animal, and especially Cohnheim’s observations! on the existence of the peculiar enzyme erefsin, capable of breaking down the more soluble proteid derivatives (proteoses and peptones) into crystalline products which no longer respond to typical proteid (biuret) tests, have brought new ideas. We are now told that the organism must synthesize the proteids peculiar to its tissues and fluids; that assimilation finds ex- pression in the reconstruction of amido-acids, diamido-acids, etc., into new molecules. In support of this the experiments of Loewi? and of Henderson and Dean,’ indicating the maintenance of nitrogenous equilibrium upon a diet free from proteids, but containing the pro- ducts of profound proteolysis, are advanced. It must be admitted that there is something attractive in the theory which assumes a complete breakdown of the food-stuffs prior to their anabolism into living tissue or circulating blood constituents. For it is less difficult to conceive how the organism can construct from these fragments the tissues peculiar to itself, and maintain its chemi- cal integrity, although the ingesta may vary widely in composition. The available experimental data are, however, by no means adequate to afford a definite answer to the problems presented. The non- proteid nitrogenous cleavage products are not found in the blood stream ;* the presence of proteoses in the blood under normal con- ditions has not been successfully demonstrated ;® and the toxic 1 COHNHEIM: Zeitschrift fiir physiologische Chemie, Igo, xxxi, p. 451; 1902, REY, ps 194s 1O02, XXxXvVi, p. 13. ; 2 Loew: Archiv fiir experimentelle Pathologie und Pharmakologie, 1902, xlviii, p. 303. 3 HENDERSON and DEAN: This journal, 1903, ix, p. 386. Since the above was written, somewhat different conclusions have been arrived at by ABDERHALDEN and Rona: Zeitschrift fiir physiologische Chemie, 1g04, xiii, p. 528. * Cf. KUTSCHER and SEEMANN: Zeitschrift fiir physiologische Chemie, 1go2, XXXiv, p. 528. 5 The experiments of ABDERHALDEN and OPPENHEIMER: Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 155, seem decisive on this point, in opposition to the positive results recorded by EMBpDEN and Knoop: Beitrage zur chemischen Physiologie, 1902, iii, p. 120, and LANGSTEIN: /did., 1903, iii, p. 373. 338 Lafayette B. Mendel and Elbert W. Rockwood. properties of the latter when introduced directly into the circulation have been clearly established in this laboratory by Underhill,’ despite the contrary views which Pick and Spiro? have advanced. Further- more, the unaltered food proteids are ordinarily not present in the general circulation in quantities detectable by chemical means; and observations like those of Abderhalden, Bergell, and Dérpinghaus ® suggest that even under the unusual conditions of protracted hunger the blood- and tissue-proteids preserve their chemical individuality. THE ABSORPTION OF PROTEIDS FROM THE SMALL INTESTINE. To the list of publications on the absorption of native proteids, as reviewed in the cited monograph by Munk on absorption, may be added those of E. W. Reid.* In dogs, he noted a disappearance of the animals’ own serum from loops of the gut. All of the previous investigators have studied the behavior of either egg-albumin, serum proteids or other proteids obtained from animal sources. We have made use of the crystallized vegetable proteid edestin (obtained from hemp-seed)® and of casein, as distinctive types of foreign native proteids. For comparison, simultaneous trials were also made in some cases with edestin-proteoses and caseoses. Methods employed.— Unfed dogs were used in these experiments. Mor- phine sulphate was given hypodermically in the dose of about one centi- gram per kilo of body-weight, the operative procedures being carried out during ether anesthesia. A loop of small intestine was withdrawn from the abdomen, carefully packed in cotton kept moistened with warm physio- logical saline solution, and thoroughly irrigated with the latter until the wash water ran perfectly clear from the lumen, the intestine being very gently massaged to facilitate the emptying of it. Ligatures were placed at two suitable places, so as to interfere as little as possible with the mesen- teric circulation. The intestine was then tied at the lower ligature, the warm proteid solution introduced slowly from a burette through a small 1 UNDERHILL: This journal, 1903, ix, p. 345. 2 Pick and Spiro: Zeitschrift fiir physiologische Chemie, 1900-1901, xxxi, Pp: 237: 3 ABDERHALDEN, BERGELL, and DORPINGHAUS: Zeitschrift fiir physiologische Chemie, 1904, xli, p. 153. 4 REID: Journal of physiology, 1895, xix, p. 240; Proceedings of the Royal Society, London, 1899, Ixv, p. 94. 6 By OSBORNE’S method. See CHITTENDEN and MENDEL: Journal of physiol- ogy, 1894, xvii, p. 50. -On the Absorption and Utihzation of Proteids. 339 incision, and the second ligature tied to retain the contents of the loop. As soon as the intestine had been ligated at both sides of the incision, the loop was replaced in the abdomen and the wound closed. After a suitable interval, the animals were bled to death, the loop of intestine excised and measured, and the adherent blood wiped away. ‘The con- tents were then carefully withdrawn. The fluid had nearly always dis- appeared and it became necessary to remove the residual precipitate of proteid mechanically (by gentle scraping and washing). ‘The contents were dissolved in salt solution or hot water, and nitrogen estimated quan- titatively in an aliquot portion by the Kjeldahl method. That this pro- cedure involved no serious error through introducing contaminations from the intestinal wall was shown in experiments in which ‘the unabsorbed edestin was recrystallized and weighed as such. In some experiments control (empty) loops, as well as loops containing proteoses, were simultaneously ligated in the same animal. The edestin solutions used were quite concentrated, sufficient sodium carbonate being added to prevent precipitation at the temperature of the body. The exact content of proteid was determined from N-estimations by Kjeldahl’s method. The edestin was not materially altered by the alkali used, since it could be recovered from the gut in typical crystalline form, and was identified in this way. The edestin-proteose solutions con- tained mixtures of proteoses, peptones, etc., prepared as follows: Re- crystallized edestin was digested with pepsin-hydrochloric acid until no precipitate formed on neutralization. ‘The concentrated solutions were , precipitated with alcohol. The aqueous non-coagulable solution of this precipitate, containing 1.28 gm. N per too c.c., was used. Acid edestin was prepared by allowing 0.4 per cent hydrochloric acid to act upon crystallized edestin for several days. It was precipitated by neutrali- zation and redissolved in warm o.g per cent sodium carbonate solution. Casein (caseinogen) was prepared by Hammarsten’s method, being re-precipitated twice, and was added in excess to a dilute sodium car- bonate solution. The filtered and very slightly alkaline casein solution was used in the trials. A solution of caseoses was prepared by the same methods as were applied to the edestin products. Admitting that the absorption figures obtained by these methods tend to be too low rather than high, owing to the technical difficulties in recovering the unabsorbed proteids, our protocols indicate that both edestin and casein disappeared only slowly from the intestine within a period of four to five hours. This is in marked contrast with the findings in the case of the proteoses from the same sources, studied in the same animals. It seems unnecessary to repeat all the 340 Lafayette B. Mendel and Elbert W. Rockwood. details of the experiments here. A few typical protocols will indicate the mode of procedure; the more important observations are briefly summarized in a table below. Typical protocols. — Experiment 7. Relative absorption of edestin and edestin-proteoses. Dog of 7.2 kilos. Loops of intestine (as indicated in the summary below) were prepared in the upper and lower parts of the small intestine, with an empty control loop of 38 cms. between them. Fifty-eight c.c. of edestin solution containing 0.587 gm. N or 3.12 gms. edestin were introduced into the upper loop (61 cms. long). The lower loop of 51 cms. received 23 c.c. of edestin-proteose solution. The animal was killed by bleeding, after 54 hours. ‘The intestinal mucosa presented a normal appearance. None of the ligated portions contained any liquid contents. From the edestin loop hard, curdy masses were removed by gentle scraping and washing, and treated with warm salt solution in which most of the material dissolved.t By diluting the filtered solution with warm water and cooling gradually, typical edestin crystals were obtained. A nitrogen estimation ona portion of the solution indicated a recovery of 0.35 gm. N, or 62 per cent; by weighing the crystals recovered from an aliquot portion, an edestin content of 69 per cent was calculated. The proteose loop was washed out with water, slightly acidified with acetic acid, and heated to precipitate any coagulable proteid. The filtrate, which gave a slight biuret reaction, contained 0.052 gm. N of the 0.295 gm. introduced, or 17 per cent. RECOVERED, Edestinpis seus os 3 « « « s Ogipermcent. WceStiN-pROLGOSES; @s0.) . & a La Ss Experiment 12. Relative absorption of sodium casein and caseoses. Dog of 11.5 kilos. Sodium casein and caseose solutions were introduced into two separated loops as indicated in the table below. When the animal was killed, after 44 hours, a middle control loop was practically empty. The washings from it gave no biuret reaction in the filtrate obtained after slightly acidifying and heating. The casein loop contained a sticky mass soluble in hot water and yielding 0.312 gm. N, or gt per cent of the amount introduced. ‘That the proteid was unaltered sodium casein was made evident by its ready precipitation by acetic acid, and its solubility in dilute alkali, and further by its content of organic phosphorus. From the caseose loop only 4 mgm. of the 121 mgm. N introduced could be recoy- ered in a similar manner. 1 The coagulation-temperature of edestin is high. Ox the Absorption and Utilization of Proterds. 341 RECOVERED. ASCII (7m Rel we ee iety Ga eo percent: CASEOSESH ss ted oy emiae ee bbe ue Ou ee SUMMARY OF RESULTS OF EXPERIMENTS ON THE ABSORPTION QF PROTEIDS FROM THE SMALL INTESTINE, ; Location Proteid equivalent fluid introduced. Number of experiment. Substance Weight of Nitrogen introduced of nitrogen.* Volume of Length of loop used. Duration of recovered. per cent 68 s9 | 30 cm..from cecum. 69 | 30cm. from py- lorus; + con- trol loop. 75 | 10 cm. from py- lorus;+ 27 cm. control loop. 75 | 50cm. from py- lorus; + con- trol loop. Edestin- i ate none | 44 cm. in upper proteose : third of gut. “ iy; 15 cm. from czcum. ef : | AC : 14 12 cm. from czecum. Acid A te 84 20 cm. from edestin czcum. Casein ! be § 69 3 cm. from cz- cum. Ss : [ ao 91 5 cm. from cz- cum + 24 cm. control loop. Caseose ; ae 35 3.3. | 30 cm. from pylorus. ° Q 8 Rea EF nw: ms Oo (9/0) = (S) zis he * Calculated from the nitrogen-content of the fluid by multiplication, using the appropriate factors. These experiments indicate no essential differences in the fate of the two typical proteids edestin and casein within the intestine under conditions which practically exclude the normal digestive processes. Cohnheim?! has found that the intestinal ferment erepsin slowly digests casein, although it is without action upon the other native proteids studied by him. In our experiments it is evident that little if any digestive change could be attributed to this enzyme in view of 1 COHNHEIM: Zeitschrift fiir physiologische Chemie, 1902, xxxv, p. 140. 342 Lafayette B. Mendel and Elbert W. Rockwood. the large quantities (70 to 90 per cent) of unaltered proteid recov- ered. Whether the proteoses underwent a cleavage to simpler com- pounds through the action of erepsin prior to their disappearance from the loops of intestine, or whether these readily diffusible diges- tion products were absorbed as such, cannot be ascertained from our observations. The protocols emphasize most strikingly, however, the importance of the preliminary digestive changes in facilitating the absorption of proteids. In Experiment IV, in which a relatively large disappearance (32 per cent) of edestin from the gut was noted, the lymph was col- lected from the thoracic duct of the dog for a considerable period before and after the introduction of proteid into the loop. The lymph samples thus obtained were compared, with reference to their con- tent of proteid, in the manner already described by one of us in a study of the paths of absorption for proteids.1. In accordance with the earlier experience, no evidence of a passage of proteid into the lymph channels was obtained. As evidence of the direct passage of unaltered proteids from the intestine into the tissue fluids, the instances of so-called ‘ alimen- tary” albuminuria are frequently cited. By the use of the precipitin test, Ascoli? believes that he has demonstrated a passage of foreign proteid into both the blood and lymph of animals, the reaction being obtained in these fluids as well as the urine when excessive quantities of appropriate proteids are introduced. These statements have not remained unchallenged, however,? nor is the “ biological” test suffi- ciently developed, as yet, to give it unquestioned significance in absorption trials of this character.4 At any rate it has not given data of any quantitative value. We have made two attempts to find edestin (or its derivatives) in the urine of rabbits after introducing solutions of this proteid (in 0.75 per cent sodium carbonate solution) into the stomach through a sound. No proteid could be detected in either case. In one trial in which 17 gms. edestin in a volume of 100 c.c. were introduced, the animal was starved for two days and muzzled before and after the feeding, in order to keep the stomach 1 MENDEL: This journal, 1899, ii, p. 137. 2 Cf. ASCOLI and VIGANO: Zeitschrift fiir physiologische Chemie, 1903, xxxix, p. 283, and earlier papers by ASCOLI in the Miinchener medicinische Wochen- schrift, 1902 and 1903. 8 See OPPENHEIMER: Beitrage zur chemischen Physiologie, 1904, iv, p. 265. 4 Cf MOLL: Beitrage zur chemischen Physiologie, 1904, iv, p. 578. On the Absorption and Utilization of Proteids. 343 empty by preventing the ingestion of the faces. It is not unlikely that the edestin, being precipitated in the stomach, is thus prevented from being absorbed before it has undergone digestive changes. But even in the event of a considerable absorption of unchanged proteid, we could not expect to detect it subsequently in the urine, in the light of our later experiments in which this proteid was introduced in considerable quantity directly into the circulation without being excreted again as such. THE ABSORPTION OF NATIVE PROTEIDS FROM THE PERITONEAL CAVITY. The studies on absorption from the peritoneal cavity have acquired additional importance in view of MacCallum’s denial of the existence of open communications between the peritoneum and the lymphatics, and likewise of the idea that the peritoneal cavity forms a part of the lymphatic system.2, Orlow’ and Hamburger * both noted an absorp- tion of blood-serum and ascites fluid to some extent from the cavity. The difficulties of obtaining anything more than approximate quan- titative data in measuring the residual proteid in the peritoneal spaces are apparent, especially when the possibility of an exudation into the latter is considered. Nevertheless, we have made a few experiments with vegetable proteids on cats and rabbits, and have seen these substances disappear in not inconsiderable quantity. Special atten- tion was directed to the possibility of an excretion of proteid in the urine in these trials. Methods. —Edestin from hemp-seed and excelsin from the Brazil-nut® were used, each being obtainable in crystallized form. The former was dis- solved in one-half to three-quarters per cent sodium carbonate solution; the excelsin in o.g per cent sodium chloride solution. The warm solutions were introduced during anzsthesia through a small opening in the abdom- inal wall which was immediately closed again. Only such animals as secreted a proteid-free urine before the experiment were used. The urine was collected after the peritoneal injection and carefully tested 1 Cf. Swirski: Archiv fiir experimentelle Pathologie und Pharmakologie, 1902, xlviii, p. 282. * MacCatvum: Johns Hopkins Hospital bulletin, 1903, xiv, p. 115. 8 OrLow: Archiv fiir die gesammte Physiologie, 1895, lix, p. 170. * HAMBURGER: Archiv fiir Physiologie, 1895, p. 281. 5 Prepared by OSBORNE’S method: American chemical journal, 1892, xiv, p. 662 344 Lafayette B. Mendel and Elbert W. Rockwood. (heat; Heller’s reaction; acetic acid and potassium ferrocyanide). When the peritoneal cavity was opened after varying intervals, the fluid had usually disappeared, while the unabsorbed foreign proteid remained deposited in curdy masses. Edestin thus recovered could be recrystal-: lized. The estimation of the extent of absorption was made on the basis of the nitrogen content of the soluble proteid material recovered. The following table gives a summary of the more interesting observations : SUMMARY OF OBSERVATIONS ON THE ABSORPTION OF VEGETABLE PROTEIDS FROM THE PERITONEAL CAVITY. Substance i = | Proteid Remarks. used. : recovered. grams c.c. Edestin, 2.10 | Rabbit VIII 50 ; 4 of glob- | No coagulable ulin N proteid ; trace of proteose. Cat I 50 4 Very little | No proteid | Injection whatever syringe used. 1.82 Dog II 44 1 | 4 of glob- Do. ulin N 1.90 Cat IT 46 Little Do. No pusor| conges- tion. Excelsin, '0.96 Rabbit I 75 ! Some un- | Proteose (15%) absorbed, and coagula- not deter- ble proteid. mined 0.73 | Rabbit II Do. Do. 1.53 | Rabbit III Do. Do. 0.80 Rabbit V Not killed | Proteid free. THe Fate oF FOoREIGN (VEGETABLE) PROTEIDS INTRODUCED DIRECTLY INTO THE CIRCULATION. The disappearance of a proteid from the intestine or any other cavity is no proof, necessarily, of its passage into the blood ; it may be retained within the absorbing membrane. The question as to the capacity of the animal body to assimilate foreign proteids without the intervention of the digestive processes has been answered in various ways. Numerous experiments by Neumeister led him to announce a distinction between assimilable and non-assimilable pro- teids, according to their behavior when injected directly into the On the Absorption and Utilization of Proterds. 345 circulation.!. The non-assimilable group included those compounds (egg-albumin, casein, and others) which speedily reappeared in the urine, and thus apparently were eliminated like other foreign sub- stances. Regarding the various blood sera, the reports have been most conflicting. But Friedenthal and Lewandowsky? showed that if toxtc sera of different species are first heated to 60° C., they can readily be assimilated after intravenous injection. More significant from our standpoint is the investigation of Munk and Lewandowsky,? demonstrating that the most diverse types of proteids can be assimi- lated in noteworthy quantities directly from the circulation, provided only that they are injected very slowly and the rate of their intro- duction is thus made comparable with that pertaining during absorption from the gut. Since our experiments were undertaken, Oppenheimer # has reported the results of intraperitoneal and intra- venous injections of egg-proteid and blood-serum in rabbits, and has introduced the term parenteral to refer to those modes of introduction which avoid the alimentary canal. Like the previous authors, he has found the serum to be very completely retained ; the egg-proteid was utilized less completely, being in part eliminated again in the urine. Repeated injections appeared to establish a certain degree of tolerance. Methods.— Our experiments involve the fate of the crystallized vegetable proteids edestin and excelsin introduced slowly into the jugular vein during ether anzesthesia. The edestin was dissolved in sodium carbonate solution of just sufficient strength (0.2-0.7 per cent) to prevent precipi- tation at body temperature. The content of edestin (containing 18.7 per cent N) varied from 4 to 12 per cent; while the excelsin, dis- solved in o.g per cent sodium chloride solution slightly alkaline with sodium carbonate, varied in amount from 0.6 to 4 per cent.. The urine of all animals used was carefully tested beforehand and found proteid-free. When coagulable proteid was present after the injections, it was estimated by analysis of the well-washed heat-coagulum by the Kjeldahl N-method. As will be noted in the protocols, after injection of excelsin, the urines frequently contained a proteose-like substance which could not be 1 NEUMEISTER: Lehrbuch der physiologischen Chemie, 1897, p. 301, contains a brief review of the literature. 2 FRIEDENTHAL and LEWANDOwSKyY: Archiv fiir Physiologie, 1899, p. 531. 8’ Munk and Lewanpowsky: Archiv fiir Physiologie, 1899, Supplementband, P: 73: + OPPENHEIMER: Beitrage zur chemischen Physiologie, 1904, iv, p. 263. 346 Lafayette B. Mendel and Elbert W. Rockwood. OF EXPERIMENTS IN WHICH VEGETABLE PROTEIDS Conditions after recovery. Depression. Not abnormal. Do. Slight depression. Lio. Good recovery. Much depression. Good. Fair; weak pulse during injection; rapid recovery. Fair. Bad; 60 cc. blood had been taken. 30 c.c. blood had been taken. Good. Great depression ; had had kittens. Good. Good. 5 days after Experiment Ia. Good. Bile experi- ment. Good. Do. Do. SUMMARY WERE INJECTED DIRECTLY INTO THE CIRCULATION. as) ° ; 2 |e |.5 Sul Se eee ey ntact Substance : 2) | eid 2.c servations on used. Animal. 2 oO 1S a oe the urine. s.6 | 8S! g Siuiieeea| pe > a S grams c.c. grams min. Edestin, 3.18 | CatIV | 78 | 0.62 | 78 | Trace of coagulable F proteid; no pro- teose iA 108 | “ Vaj 20 | 0.54 {| 15 | No proteid what- ever 4 LOS | 5s ew) B20 0:54 | 20 Do. ¥ LBD: |, Waly ess OZ al 25 Do. * 2200) ea oN LU es 1.08 | 41 Do. 4 P68) |) = VIEL) 2651 3055 23 Do. is 3:35. *o UX 53 0.88 | 35 Do. ESB Gaal © FO! 1 2:90 12705 Zot N recovered “s 3.24 | “ Xb] 40 | 0.75 | 90 | Proteid-free H 195 | “ XIb| 20 | 0.97 | 30 | Trace of coagulable proteid y 4.38 | “ XII | 36 | 0.88 | 40 | 2.8% N recovered S 3.68 | “XIIL| 63 | 0:97 | 43..| Proteid-tree 38) | - 2 NG Sb Oey Do. i Si 22 | PON 35 1.19 | 25 | Little proteid ss 3.72 | Dog Ia | 60 1.05 | 55 | Little coagulable proteid nS 230) 22 Sie! A736 0.64 | 35 Do. 535°) se |heee 0.76 | 60 Do. ¥" 118 |) SLED) TOO O93" |) 60 . | Proteid-free $ S27 RLV 42 OLAS || 3) ’ Do. Excelsin, 0.42 | Rabbit I} 45 0.23 | 42 | No coagulable pro- teid; proteoses. 49 % of N re- covered - O67) SEE OM 035°) 90 | Do.," 21 7 cake recovered 42.05.) Dog Ti) M139h 1 50:525) 55. |-Much. protease: ; Trace of coagu- lable proteid xe 1k4 See Gat clan SO |O:s6nl so) sl broteoses on, mot N recovered. No peptones On the Absorption and Utilization of Proteids. 347 precipitated by coagulation, but responded to other proteid tests. Over twenty-five injection trials were made, the conditions as regard strength of solution, dose, etc., being widely varied. The skin wounds usually healed satisfactorily, and some of the animals were used for several experiments. [Illustrative protocols follow. Experiment 11.— A cat of 2.6 kilos received an injection of 50 c.c. of excelsin solution, containing 1.45 gms. proteid, in the course of thirty-five minutes. The cat recovered speedily, and the urine collected during the following day gave only the faintest tests for proteid by heat or Heller’s test. The coagulation filtrate contained a proteose-like substance precipitable by zinc sulphate, yielding o.o125 gm. N, or 4.7 per cent of the amount injected. There was no peptone present. On the second day the urine was free from all proteids. Ten days later the cat’s weight had diminished to 2 kilos, owing to insufficient feeding. Twenty c.c. of edestin solution, containing 1.95 gms. proteid, were now injected in thirty minutes, respira- tion and pulse remaining fairly good. Urine collected two hours afterwards was proteid-free ; the urine of the following two days contained very slight amounts of coagulable proteid, and none thereafter. The body-weight increased to 2.2 kilos. These experiments demonstrate that in their bebavior after direct introduction into the circulation in various species of animals, the typical vegetable proteids present no essential difference from that noted with proteids of animal origin. Their utilization — or properly speaking, their failure to reappear as such in the urine — was unusu- ally good in practically all of the trials. The close resemblance between edestin and excelsin in chemical make-up is indicated by the differential analyses made according to Hausmann’s method by Osborne and Harris,! namely : Total | Nitrogenas| Basic Non-basic | N in MgO nitrogen. | ammonia. | nitrogen. | nitrogen. | precipitate. Edestin (hemp-seed) . . 18.64 1.88 5.91 10.78 0.12 Excelsin (Brazil-nut) . . 18.30 1.48 5.91 10.97 0.17 Yet in distinction from edestin, the excelsin injections were followed by an elimination of a proteose-like substance, not coagulable by 1 OsBORNE and HARRIS: Journal of the American Chemical Society, 1903, XXV, p. 348. 348 Lafayette B. Mendel and Lilbert W. Rockwood. heat, but precipitable by zinc sulphate and other proteid precipitants, and giving a characteristic biuret reaction. This body was found in the urine of the dog, cat, and rabbit alike, and was also discovered after intraperitoneal excelsin injections. That it does not arise in the urine itself by transformation from excreted excelsin was shown by trials in which pure excelsin was found to retain its coagulability after solution in urine from the same animals. Whether —as seems likely — the product represents a derivative of the injected excelsin, and where the transformation occurs, cannot be told at present. We recall the related experience with various proteoses which have been found to undergo a digestive change prior to their elimination through the kidneys after intravenous injection.! With reference to the immediate action of the intravenous injection, symptoms comparable to those described by Brodie? as typical for serum-proteid injections were frequently observed, particularly in the cats. They consisted in a temporary inhibition of the heart’s action and of respiration, and were occasionally of sufficient moment to cause death. The severity of the symptoms was usually increased with she dose and rapidity of injection. The sudden discharge of a few cubic centimetres of proteid solution into a vein may cause a profound effect, but if care is taken to prevent a repetition the heart slowly recovers. We usually made the injections with an intermittent flow. The appearance of more or less proteid in the urine seems to be determined less by the quantity injected than by the general condition of the animal. In agreement with Brodie’s experience, we found the symptoms far less marked in dogs than in cats. In none of the ex- periments was there any evidence (hemoglobinuria, choluria) of a destruction of erythrocytes. Obviously paths of elimination other than the kidneys suggest themselves for the foreign proteids, as they have been emphasized by one of us for inorganic compounds.’ Giirber and Hallauer* have quite recently considered the bile in this connection. In experiments on three rabbits with temporary fistulz, they found a large part of the proteid in both dle and urine after injection of 2.7 gms. casein, although the properties of the proteid in the bile no longer exactly 1 Cf. CHITTENDEN, MENDEL, and McDermott: This journal, 1898, i, p. 275 ; CHITTENDEN, MENDEL, and HENDERSON: /d7d., 1899, ii, p. 165. * BRODIE: Journal of physiology, 1901, xxvi, p. 48. 3 MENDEL and THACHER: This journal, 1904, xi, p. 5. * GURBER and HALLAveER: Zeitschrift fiir Biologie, 1904, xlv, p. 372. On the Absorption and Utsthzation of Protetds. 349 resembled those of the original casein. Remembering the very fav- orable results obtained by Munk and Lewandowsky in their careful injection experiments with casein (of which traces only were elimi- nated in the urine) one may well question whether the results ob- tained by Giirber and Hallauer were not due to unusual pathological disturbances, as in the experiments on the elimination of proteid in the bile reported by Brauer! and Pilzecker.2, We have made observa- tions on three dogs with temporary biliary fistula, to determine whether injected edestin is excreted through this channel. The results were negative in every case; an illustrative protocol of one experiment will suffice here. We have not investigated the possi- bility of an excretion of proteid into the intestine. A dog of about 9 kilos body-weight was fed at 8 a. mM. to promote the subse- quent flow of bile ; at 9.30, 60 mgms. morphine hydrochlorate were given subcutaneously, and later anaesthesia was induced with chloroform-ether mixture. At 10.30 a fistula was established and ox-bile injected into the rectum to accelerate the normal flow of bile. The normal bile contained considerable mucoid substance precipitable with acetic acid, but no other proteid precipitable by salt and heating or by alcohol. From 12.20 to 1.10, go c.c. of a solution containing 6 gms. of edestin were injected into the jugular vein. The bile collected until 4.15 behaved precisely like that collected before the injection. In another experiment, bile obtained after injection of 5 gms. of edestin into a small dog was dialyzed with the object of precipitating in crystalline form any edestin which might have been-excreted. The results were likewise negative. How is the injected proteid utilized ?—From results obtained by means of the ‘ biological” tests Sachs® has reported that after trans- fusion of ox-blood into rabbits, it-is in large part retained in the circulation for two or three days. We have attempted to ascertain whether the nitrogen equivalent of the injected proteid is eliminated in some form other than unutilized proteid. The method consisted in starving dogs until a constant daily N-elimination was attained, and then noting the effect of proteid injections on the composition of the urine. This plan was abandoned after two trials, because it was impossible to estimate the contribution of other factors such as anzesthesia,‘ etc., to the increase in the eliminated nitrogenous com- 1 BRAUER: Zeitschrift fiir physiologische Chemie, 1903, xl, p. 182. 2 PILZECKER : Zeitschrift fiir physiologische Chemie, 1904, xli, p. 157. 3 SacHs: Archiv fiir Physiologie, 1903, p. 495. * A slightly increased elimination of urinary nitrogen after ether anesthesia has been reported by HAWK: This journal, 1904, x, p. xxxvii. 350 Lafayette B. Mendel and Elbert W. Rockwood. pounds after the injection, or to determine whether, and in what degree, the mere presence of the foreign proteid stimulated proteid katabolism. In each case the introduction of the proteid was followed by an increase in the nitrogen output considerably larger than could be accounted for by the substance injected. The urines were obtained by catheterization and were proteid-free in these trials. The pro- tocols are of interest in emphasizing the uncertainty of conclusions drawn from a study of the nitrogen elimination as an index to parent- eral absorption. This criticism applies to some of the earlier investi- gators by whom the method was employed. NITROGENOUS METABOLISM AFTER INTRAVENOUS INJECTIONS OF PROTEIDS. Urine. Body-weight. Volume. N. kgm, c.c. grams ] 8.7 67 3.07 2 8.5 54 2.56 3 8.4 54 2.69 4 8.4 70 2.38 5 6 8.3 55 2.31 Dog XIII. 8.2 130 2.51 (no food) Intravenous injection of 7.78 gms. edestin (containing 1.46 gms. N) in 100 c.c. fluid during one hour. ee) 7.09 Hell 4.36 ths 3.01 7.4 2.83 4.6 2.03 4.4 2.25 Dog XIV. Intravenous injection of 3.27 gms. edestin (containing 0.615 (no food; 150 gm. N) in 42 c.c. fluid during 35 min. c.c. water daily) 4.2 42 fell Ox the Absorption and Utelization of Proterds. 351 Finally, we have attempted to follow the fate of the foreign globulins by estimating the relative proportions of the plasma pro- teids in cats at varying periods after the injections. Our method closely resembled that employed by Lewinski.! The results of a large number of experiments were so variable that we have been unable to draw any decisive inferences from them. SUMMARY AND CONCLUSIONS. Vegetable proteids (crystallized edestin from hemp-seed and ex- celsin from the Brazil-nut), slowly introduced in solution into the circulation of animals, can apparently be retained in the organism for the most part, even when the quantities introduced almost equal that of the globulins normally present in the blood. At any rate they are not eliminated unchanged in the urine (or in the bile, in the few experiments tried). When solutions of vegetable proteids are injected too rapidly or in too great concentration, toxic symptoms, including an inhibition of the cardiac and respiratory activities, may be observed, especially in cats. This corresponds with observations of Brodie after serum- proteid injections in these animals. The chemically similar proteids, edestin and excelsin, show slight differences in physiological action, a small amount of a proteose-like substance being found in the urine after intravenous or intraperi- toneal (parenteral) introduction of excelsin, but not with edestin. The observation suggests the further possibility of applying chemico- biological reactions in distinguishing related proteids. The vegetable proteids soon disappear in considerable part when introduced into the peritoneal cavity. That they reach the circulation is made probable in the case of excelsin at least, by the appearance of the typical urine proteose-body noted after direct intravenous injections. For the most part, however, the proteids do not reappear in the urine, The unaltered proteids edestin and casein are absorbed to a very small extent, if at all, from portions of the living small intestine in which the ordinary digestive processes are excluded as far as possible. On the other hand, the proteoses and peptones obtained by peptic digestion of these proteids readily disappear from the intestine under the same conditions. It is not necessary to assume that in these 1 LEWINSKI;: Archiv fiir die gesammte Physiologie, 1903, c, p. 613. 352 Lafayette B. Mendel and Elbert W. Rockwood. cases they are first completely broken down by the intestinal enzyme erepsin; for casein (upon which erepsin can act) may remain un- absorbed. Dissolved edestin could be recovered in crystalline form, z.e., unchanged after remaining in the intestine for several hours. The typical vegetable proteids show no marked differences from those of animal origin in their relation to the processes of metabolism. The attempts to learn the fate of the foreign proteids retained in the system have been rather unsuccessful. It will be of interest to ascertain something further regarding their destination, and the exact mode of utilization which they undergo. iii? PRODUCTION, OF CHOLIN FROM LECITHIN AND BRAIN-TISSUE. By ISADOR H. CORIAT. [From the Chemical Laboratory of the Worcester Insane Hospital.| INTRODUCTION. iid bee of the recent work on the autolysis of animal organs has related mainly to the detection of the hexon, purin, or pyrimidin bases, or to the estimation of the nitrogen content, either as ammonia, soluble nitrogen, in the heat coagulum, or in the zinc sulphate precipitate. Levene found that all the animal glands so far studied by him gave similar end-products on prolonged self-digestion. The work on brain-tissue is limited to the investigations of Levene! and the nitro- gen determinations that were made sufficiently proved the presence of a probable intracellular, proteolytic enzyme. Austin? had previ- ously produced oxalic acid in the digestion of brain-substance with pancreatin, but failed to detect it by heating lecithin in sealed tubes with barium hydrate. The most important of the decomposition- products of lecithin is cholin, which is found in the central nervous system, blood, and cerebrospinal fluid in those conditions where active degeneration is taking place. Studies in the production of cholin in various pathological states of the nervous system, especially in general paralysis and epilepsy, have been carried out by Mott and Halliburton,? Donath,* Wilson,® and also by myself.6 The conclusions ' P. A. LEVENE and L. B. STOOKEY: Journal of medical research, 1903, x, p. 212. 2 A. E. Austin: Boston medical and surgical journal, tgor, clxv, p. 181. 3 W. D. HALLIBURTON: The Croonian Lectures on the Chemical Side of Ner- vous Activity, 1901; W. D. HALLIBURTON and F. W. Mott: Philosophical transactions, 1899, p. 211; F. W. Morr: Archives of neurology, 1903, ii, p. 858; W. D. HALLIBURTON and F. W. Mort: British medical journal, 1899, p- 1082. * JuLius DoNATH: Zeitschrift fiir physiologische Chemie, 1903, xxxix, pp. 526-544; Deutsche Zeitschrift fiir Nervenheilkunde, 1904, xxvii, p. 71. 5 M.S. WILSON: Revue neurologique, 1904, p. 41. 6 I. H. CorraT: American journal of insanity, 1903, lix, p. 3933; 1904, Ix, p. 733; This journal, 1903, x, p. III. 353 354 Lsador H. Coriat. arrived at show that the intracerebral injection or application of cholin produces severe paralytic effects, epileptiform and tetaniform seizures. It differs but little from neurin in its convulsive action, except that it is less toxic, but it shows some variations in its effect on blood-pressure. It is found in the cerebrospinal fluid, brain, and cord only in those conditions where there is active myelin decay, leading to a decomposition of the lecithin, and Donath has recently shown! that in these states there is a coincident increase in the phosphoric acid of the cerebrospinal fluid. This is easily explained if we remember that both cholin and glycerophosphoric acid pass into the fluid on the decomposition of lecithin, while the stearic acid combines with the glycerol, forms neutral fat, and accumulates beneath the neurilemma. This latter combination explains the osmic acid reaction. In two of his cases (tabes, Jacksonian epilepsy) Donath also detected lecithin, in one of which the myelin forms were present. Gilson? failed to find cholin as the result of the action of weak sulphuric acid on lecithin, but found instead small quantities of glycero- phosphoric acid, another phosphorus-containing compound ( distearyl- glycerophosphoric acid), in still smaller quantities, and abundant amounts of free phosphoric acid. Dianconow,’ on shaking an ethereal solution of lecithin with sulphuric acid, found cholin-sulphate and distearylglycerophosphoric acid as the result of this reaction, Bokay* found that lipase splits egg lecithin into glycerophosphoric acid, stearic acid, and neurin; but as none of these products appeared in the urine or faeces, he concluded that they must be absorbed and resynthetized. According to Hasebrock,® glycerophosphoric acid, fatty acids, and cholin may be produced by the action of putrefactive bacteria on lecithin; the further action of these organisms splits the cholin into carbon dioxide, methane, ammonia, and trimethylamin, and these latter products may also be produced by cooking cholin with potassium or barium hydrate. Vincent and Cramer ® find that normal ox blood, which contains no cholin, gives an organic platinum double salt consisting almost 1 Jutius DonaTH: Zeitschrift fiir physiologische Chemie, 1904, xlii. 2 E. GILson: Zeitschrift fiir physiologische Chemie, xii, p. 585. 8 DIANCONOW: Centralblatt fiir die medicinischen Wissenschaften, 1868, P- 434- 4 A. B6KAY: Zeitschrift fiir physiologische Chemie, 1877, i, p. 157. 5 K. HASEBROCK: Zeitschrift fiir physiologische Chemie, 1888, xii, p. 142. 6 S. VINCENT and W. CRAMER: Journal of physiology, 1903, xxx, p. 143. Production of Cholin from Lecithin and Brain-Tissue. 355 entirely of potassium and ammonium platinochloride; while ina watery extract of nerve tissue, they find a platinum salt of dicholin anhydride. Gulewitsch! has also detected small quantities of cholin in the alcoholic extract of fresh ox brain, while neurin was absent ; and Halliburton has found that normal human cerebrospinal fluid contains traces of cholin. These minute amounts, however, are con- fined to an exceedingly small number of crystals, and do not influence either the qualitative detection or quantitative estimation. The objec- tions raised by Allen and French,” that these yellow octahedra of the double platinum salt may be either ammonium or potassium, will not stand critical inquiry from the standpoint of solubility, size, form, and arrangement of crystals, percentage of platinum, melting point, and the various reactions to alkaloidal reagents. Mott and Barrett,® in the examination of a degenerated cord from a case of hemiplegia, find a diminution of lecithin on the diseased side. Noll‘4 also in experimental section of the sciatic nerve, with consequent degeneration, found a decrease of phosphorus and of the alcoholic extract. In a series of experiments on rats, not yet pub- lished, I found that acute poisoning with strychnine, morphine, vera- trine, cocaine, phosphorus, and alcohol is also effective in splitting up the lecithin, as cholin was found in both the brain and cord. While these various decomposition-products, especially cholin, have been the subject of much attention, the factors concerned in the splitting up of the lecithin have remained unnoticed. The various lecithins from brain, egg, yeast, barley, and malt show a wide differ- ence, not only in the fatty acid group, but in the phosphoric acid and its relation to the methyl content. The three methyl groups attached to nitrogen which form the cholin are more stable, and this is so well marked that lecithin can be quantitated by the estimation of these groups, according to the method of Herzig and Meyer. While this splitting can be accomplished in the laboratory by heating lecithin with barium hydrate, or by the action of lipase or putrefactive bac- teria, yet in the human organism other factors must be sought. To this end, the recent work on autolysis and intracellular ferments is stimulating, and it was along these lines that the present research was conducted. 1 W. GULEWITSCH: Zeitschrift fiir physiologische Chemie, 1899, xxvii, p. 81. 2 R. W. ALLEN and H. FRENCH: Journal of physiology 1903, xxx, p. xxix. 3 F. W. Mott and W. BARRETT: Archives of neurology, 1899, i, p. 346. 4 A. NOLL: Zeitschrift fiir physiologische Chemie, 1899, xxvii, pp- 4, 5. 356 Lsador Ff. Coriat. METHODS. The lecithin used was absolutely pure. It was isolated from calf brain, yolk of egg, and human brain (senile dementia), according to the method of Koch.!’ The pepsin and trypsin were very active preparations. The human brain material was obtained from a case of senile dementia, and was free from cholin, when analyzed accord- ing to the method detailed below. The same method was used in all the experiments. After digestion had proceeded the required length of time, the solution was filtered (in case of the brain-material, by means of an exhaust filter), evapo- rated to dryness on the water bath, extracted with absolute alcohol, and again filtered. This was repeated twice, in order to insure the absence of potassium salts and any proteid. The operations were conducted at a temperature not exceeding 75° C., as lecithin decom- poses at this point. If found necessary, the solution was decolorized with animal charcoal. To the final extract there was added an excess of a 5 per cent solution of platinum chloride in absolute alcohol. This resulting yellow precipitate was washed several times with absolute alcohol in order to remove all traces of the platinum salt, the residue dissolved in warm (40° C.) 15 per cent alcohol, and allowed to crystallize in a large watch-glass over calcium chloride. Although crystallization took several days, I failed to observe any separation of oleic acid, which,as Thudichum points out, takes place under these conditions with the platinum salt of lecithin and which, according to him, is due to a decomposition of the platinum compound. Cholin was designated as present, only if the large single and twin octahe- dral crystals were found, and if these were freely soluble in water and 15 per cent alcohol. The amount of platinum in this compound is 31.64 per cent, from which the absolute weight of the pure alkaloid was calculated. My compounds yielded 31.41 per cent platinum. In previous work along this line, I have further identified cholin by its positive reactions with alkaloidal reagents. The other platinum compounds liable to be formed are kephalin (3.595 per cent platinum), lecithin (10.2 per cent platinum), and neurin (33.6 per cent platinum) ; but only the platinum salt of cholin is soluble in water and 15 per cent alcohol. The other compounds do not dissolve in water or alcohol of any strength, but are precipitated from their ethereal solu- 1 W. Kocu: Zeitschrift fiir physiologische Chemie, 1902, xxxvi, pp. 2, 3- Production of Cholin from Lecithin and Bratn-Tissue. 357 tions by absolute alcohol. Of these, neurin alone is liable to be mis- taken for the cholin salt, but the solubility, percentage of platinum, shape of the crystals, and different melting point should readily dis- tinguish it. Neurin crystallizes in single small octahedra, never in the large twin variety; it is soluble in hot water only with great difficulty, and the percentage of platinum is higher than that of cholin, and the melting point of the double salt is lower. Cholin melts at 240°—241° C., neurin at 213°-214° C. I have been unable to find neurin as a decomposition product of lecithin in any pathological cases, even when large quantities of cholin were present. A case of Huntington’s chorea was carefully worked over for neurin, but none was found, although cholin was present in abundance. The platinum compound of lecithin is insoluble in alcohol of any strength. On account of the great differences in solubility, none of the above compounds except cholin, even if precipitated from a solution by platinum chloride, will be a source of error. Of the other decomposition-products of lecithin, neither the fatty acids nor glycerophosphoric acid yields a precipi- tate with platinum chloride. In the estimations, cerebrin also can easily be omitted, as it does not contain a methyl group and does not influence lecithin estimation, while kephalin contains only one methyl group. In the protocols which follow, the amounts of cholin are expressed in terms of the absolute alkaloid, calculated from the platinum salt. DETAILS OF EXPERIMENTS. Series I.— Pepsin and trypsin digestion, and decomposition of pure lecithins. In the following experiments 0.2 gram each of pure human, calf, and egg lecithin was used and allowed to digest in the different media for seventy-two hours at 38°-40° C. The pepsin and trypsin were absolutely pure and very active preparations. For each digestion experiment, 0.2 gram of the enzyme was used ; in the case of pepsin dissolved in 100 c.c. digestive hydrochloric acid, while the trypsin was dissolved in 100 c.c. 0.5 per cent sodium-carbonate solution. To this latter a few drops of chloroform were added to prevent putrefaction. In the pepsin and trypsin experiments but little of the lecithin went into solution, whereas those in the neutral media (water) made a cloudy emulsion which remained uniform even on decomposition. An odor of putrefaction developed, and the solution became slightly acid, probably due to glycerophosphoric acid. Lsador H. Coriat. Os on 170) No. of experiment. Lecithin. Medium. Cholin. Human Pepsin Human Trypsin Calf Pepsin Calf Trypsin Egg Pepsin Egg Trypsin Human Water Egg Water \O* (60) “SGN (ON) ER os hon Calf Water Series ,II. — Decomposition and autolysis. In the second series of experi- ments, fresh human brain-tissue, obtained from a case of senile dementia, was freed from blood and membrane; finely minced and accurately weighed quantities (10 grams) were distributed in neutral, acid, and alkaline media, and also subjected to the action of pepsin and trypsin, under the variations detailed below. The amounts of fluid in each case were Ioo c.c., and the same quantities of pepsin and trypsin were used as in Series I. The temperature and time of digestion corre- sponded to the first series of experiments with one exception (Series V, Experiment 3). wlio. 0k Medium. Cholin. Remarks. experiment. gram Normal salt solution . . . . | 0.0513 | Slight odor of putrefaction. 0.2% acetic acid solution. . . | None No odor. 0.5% sodium carbonate solution | 0.0547 | Slight odor of putrefaction. Series III. — Decomposition prevented by chloroform; autolysis alone. No. of Medium. : Remarks. experiment. Normal salt solution ... . No odor. 0.2% acetic acid solution. . . No odor. 05% sodium-carbonate solution No odor. Production of Cholin from Lecithin and Brain-Tissue. 359 Series IV. — The brain substance, after being placed in the various media, was boiled and cooled before autolysis was allowed to proceed. Lo chloroform was added; decomposition alone. No. of Medium. Cholin. Remarks. experiment. gram Normal salt solution . . . . | 0.0219 | Slight odor of putrefaction. 0.2% acetic acid solution , . . | None No odor. 0.5% sodium-carbonate solution | 0.0329 | Slight odor of putrefaction. Series V.— Zhe mixture was boiled and cooled as in Series LV ; but chloroform was added to prevent decomposition, in order to show the negative action of putrefaction and autolysis. No. of é Medium. Remarks. experiment. Normal salt solution ... » None No odor. 0.2% acetic acid solution. . . | None No odor. 0.5% sodium-carbonate solution | 0.013 No odor.* * Very slightly heated, and allowed to digest for 96 hours, in order to destroy only part of the enzyme and show the long-continued action of the small amount present. Series VI. — Action of pepsin and trypsin without previous heating of the brain- substance; combined action of the proteolytic enzymes and autolysis. No. of experiment. Enzyme. Cholin. Remarks, Pepsin Absent No odor. Trypsin Absent No odor. ‘ In Experiment 2, chloroform was added to prevent putrefaction. Series VII.— Same as Series VI, except that the brain-substance was heated and cooled before adding the solution containing the enzymes ; isolated action of enzymes, without autolysis. 360 Isador HF. Coriat. No. of experiment. Enzvme. Cholin. Remarks. % 1 Pepsin Absent No odor. 2 Trypsin Absent No odor (chloroform added). THEORETICAL AND ACTUAL YIELD OF CHOLIN. In Table I, the theoretical yield of cholin required from pure lecithin, if all were split up, was calculated from the equation of the decomposition-products of lecithin. The yield from 1 gram of leci- thin, if entirely decomposed, should be 0.1405 gram cholin. By splitting a sample (1 gram) of human lecithin with barium hydrate, 0.140 gram was obtained, which is within the limit of experimental error. The amounts of lecithin in brain-tissue were taken from Koch’s figures of the analysis of the brain of an epileptic,’ the average of the white matter (corpus callosum) and gray matter (prefrontal cortex) being 4.16 per cent. TABLE III. (Showing the amounts of cholin that should be theoretically produced if all the lecithin were split up, and the amounts practically found. Calculated from the formula of the decomposition-products of lecithin and from Koch’s figures on the percentage of lecithin in brain-tissue.) Amount Series. calculated. gram 0.0281 0.0281 0.0281 0.0584 0.0584 0 0584 0.0584 0.0584 0.0584 0.0584 1 W. Kocu: This journal, 1904, xi, p. 303. Production of Cholin from Lecithin and Brain-Tissue. 361 SUMMARY AND CONCLUSIONS. 1. The putrefaction of either human, calf, or egg lecithin, in a neutral medium, yields cholin, the reaction at the same time becoming acid. (Series I, Experiments 7, 8, g.) 2. Human lecithin, by putrefaction alone, yields less cholin than either egg or calf lecithin, but the amount is about equal to the yield from brain-tissue during autolysis (Series III, Experiments 1, 3), but is less than the theoretical yield by calculation from the equation of the hydrolysis of lecithin. . 3. The putrefaction of brain-tissue alone produces cholin and in a greater quantity than autolysis alone. (Series III, 1 and 3, and Series IV, 1 and 3.) 4. Lecithin is not split on prolonged contact with acids (hydro- chloric and acetic, as in Series I, Experiments 1, 3, and 5, and Series Bee ilo LV 2: Vi Qe Vie VEL 1). 5. Lecithin can be split by heating it with barium hydrate, and in this case the entire theoretical yield of cholin is obtained. 6. Neither pepsin nor trypsin is effective in splitting off the methyl group from any of these lecithins, so that cholin may be produced. (Series I, Experiments 1-6 inclusive.) Lipase, however, is capable of splitting lecithin. 7. Pepsin and trypsin not only fail to act on the lecithin of brain- tissue, but actually seem to inhibit or even destroy autolysis. (Series VI and VII.) 8. There is an enzyme present in brain-tissue, capable of splitting © cholin from lecithin. g. The enzyme acts only in neutral or slightly alkaline media, and the yield of cholin in the latter is greater than in the former. The enzyme is inactive in slightly acid media. Levene found that the autolysis of brain-tissue, so far as the proteolytic process was con- cerned, was favored by the presence of an acid and inhibited by an alkali. The lecithin-splitting enzyme appears to have a contrary action. In the body, the action of the enzyme is favored by the normal alkaline reaction of the nerve-substance and cerebrospinal fluid. The ease with which this enzyme acts on lecithin is probably explained by the fact that the lecithin in the central nervous system, as in the stroma of red blood-corpuscles, is not in a chemical com- bination, but in an emulsiform condition, and is therefore capable of mechanical solution. According to Koch, brain lecithin, which . 362 Lsador FI, Coriat. has been a part of living tissues, gives a more perfect emulsion than egg lecithin, which is merely stored-up food material. These emul- sions have many of the physical properties of living protoplasm in that precipitation of the emulsion takes place by divalent kations, and is prevented by univalent and trivalent kations. This I have also observed. 10. The enzyme can be destroyed by heating, and then, if the suspension of the brain-tissue be kept absolutely sterile, no cholin is produced (Series V, 1); if putrefaction is allowed to supervene, cholin will be formed in a greater quantity than by autolysis alone. (Series IV, 1 and 3.) 11. Prolonged action with antiseptic precautions, and with very slight preliminary heating, so as to destroy only part of the enzyme, produces a very small amount of cholin. (Series V, Experiment 3.) 12. As with all enzymes, there is an inhibitory influence of reaction-products which cannot be removed, and this explains the low percentage of cholin obtained. 13. Efforts to isolate this enzyme have so far been unsuccessful. 14. The cholin produced in the combined action of autolysis and putrefaction is nearly equal to the sum of each, when acting sepa- rately in similar media, and nearly approaches the theoretical amount which should be yielded by the percentage of lecithin in the weight of brain-tissue used. 15. The amounts of cholin actually produced, both by the decom- position of pure lecithin and in the autolysis of brain-tissue, is less than that theoretically required for the quantity of lecithin present, if all of it were split up (according to the equation of decomposition). The only approach to the theoretical amount is when lecithin is saponified with barium hydrate, or by the combined action of putrefaction and autolysis. PURTHER EXPERIMENTS ON THE HAMOLYSINO- GENIC AND AGGLUTININOGENIC ACTION OF LAKED (\CORPUSCLES: BY GoM. SULEWART: [From the Hull Physiological Laboratory, University of Chicago.| N a recent paper! I investigated the distribution of the hamoly- sinogens and agglutininogens in the stromata and extracorpus- cular liquid of colored corpuscles laked in various ways, injecting into one series of animals the washed ghosts, and into another series the liquid of foreign corpuscles, and testing the sera of the injected animals, after a suitable interval,as regards their power of agglutinat- ing and laking corpuscles of the same kind as those injected. In the preparation of the material for injection five methods of laking were employed (heat, freezing and thawing, foreign serum, saponin, and water). Some of these methods, in accordance with the classification adopted in a previous communication 2 may be considered “ violent,” since they can be shown to affect the stroma more profoundly than the others, which we may designate by contrast as “mild.” The general result of that investigation was that although after most methods of laking the stroma was more strongly agglutininogenic than hemolysinogenic, and the liquid the reverse, no complete separa- tion from the corpuscle either of agglutininogen or of hamolysinogen could be effected by any of the methods of laking employed. In other words, the stromata of the laked corpuscles still possessed, in general, both agglutininogenic and hzemolysinogenic powers, and both of these powers had also been acquired by the extracorpuscular liquid. This result is opposed to the conclusion of Nolf,? who, working with water-laked corpuscles, convinced himself that the agglutinino- genic substance was entirely contained in the stroma, and the hemolysinogenic entirely in the liquid. Ford and Halsey * also, 1 STEWART: This journal, 1904, xi, p. 250. 2 STEWART: Journal of physiology, 1899, xxiv, p. 211. 3 NoLF: Annales de |’Institut Pasteur, 1900, xiv, p. 297. * Forp and HALseEy: Journal of medical research, 1904, xi, p. 403. 363 264%; G. N. Stewart. e after carefully investigating the action of water-laked corpuscles, and, for the sake of comparison, following in all its details the procedure recommended by Nolf, have been unable to confirm his statement. In all their experiments, agglutination and hzmolysis were both caused by any serum which caused one of them. In some cases where the serum had no hemolytic effect, but nevertheless produced ag- glutination, they proved that the absence of haemolysis was due not to the want of the specific intermediary body, but to the lack of complement. Here I desire to record the results of some further experiments which were not completed in time to be included in the pre- vious paper. They were made on four fresh rabbits and two fresh guinea-pigs. In addition, observations were carried out on two of the TABLE I. Aggluti- nation. eee Material injected. Animal. q Liquid of water-laked dog’s corpuscles, sepa- rated by centrifuge in first two et ag in the third filtered through clay . . Rabbit F Ghosts of water-laked dog’s corpuscles “3 in- jections) . Rabbit G Liquid of dog’s corpuscles laked by freezing and thawing, filtered through clay (3 injections) . Rabbit J Ghosts of dog’s corpuscles laked by freezing and thawing (2 injections) . Rabbit K Liquid of rabbit’s corpuscles laked by " dog’s serum filtered through clay (2 injections) .| Guinea-pig R Ghosts of rabbit’s ee as laked uy ee serum (2 injections) . . Guinea-pig S rabbits and one of the guinea-pigs used in the previous work, to determine for how long a period the modification in the serum per- sisted. The two fresh guinea-pigs received respectively washed rab- bit’s ghosts laked by dog’s serum, and the liquid separated from the ghosts by filtration through unglazed porcelain. Two of the rabbits received respectively the ghosts and the liquid of water-laked dog’s corpuscles, the liquid for the first two injections being separated from the ghosts by prolonged centrifugalization, but in the third by filtering through clay. The other two rabbits were injected respec- tively with the liquid and the ghosts of dog’s corpuscles laked by freezing and thawing, the liquid being filtered through clay. All these animals received at least two injections. Some of them Experiments on the Action of Laked Corpuscles. 365 received three. The results of the examination of the sera are summarized in Table I. The moderate laking caused by the sera of rabbits F and G was shown, by activating dog’s corpuscles with them at 0° C., to be due to a specific intermediary body. The attempt made to demonstrate that the same was true for the laking action of the serum of Guinea- pig S on rabbit’s corpuscles was not successful. The results for the liquid and ghosts of the water-laked blood agree substantially with those of the previous paper. While both cause the production of serum with specific hemolytic as well as agglutinating power, it is the agglutinating power which preponderates. The same preponder- ance which was noted in the former paper, in the case of guinea- pigs injected with the stromata of rabbit’s corpuscles, laked by freez- ing and thawing, is even more evident in the case of Rabbit K, which received the ghosts of dog’s corpuscles laked in the same way. Here the serum had no noticeable hzmolytic effect, but caused fair agglu- tination of dog’s corpuscles. The want of laking power was apparently not due to the absence of complement, since the addition of dog’s serum did not alter the result. For the stromata of rabbit’s cor- puscles laked by foreign (dog’s) serum, and injected into guinea-pigs, the negative result as regards agglutination supplements the table in the previous paper where observations on this point were missing. I do not propose at present to discuss the significance of this negative result. Although there was but a single experiment, one cannot help feeling some confidence in it since agglutination is in general very easily and strikingly demonstrated with immune sera produced by the injection of the stromata of corpuscles laked in other ways. Although in the two injections which this guinea-pig (S) received, a relatively large quantity of ghosts (corresponding to 18 c.c. of blood) was introduced, a larger number of injections at shorter intervals might possibly give a different result. The most interesting point shown in the Table is the absence of both specific hamolysin and agglutinin from the serum of Rabbit J, which received the liquid of dog’s corpuscles laked by freezing and thawing, and from the serum of Guinea-pig R, which received the liquid of rabbit’s corpuscles laked by dog’s serum, both liquids having been filtered through a pot of porous earthenware before injection. ‘In the last paper it was stated that the injection of the liquid of cor- puscles laked by freezing and thawing confers upon the serum of the animal into which it is injected the power of agglutinating and laking 306 G. N. Stewart. the corresponding corpuscles. Here the liquid was separated from the stromata by the centrifuge and was not filtered. Apparently, then, the porous clay prevents the agglutininogens and hzemolysino- gens from passing through. It is, of course, possible that a small amount does pass through the filter, and that a larger number of injections might give a positive result. The difference in the action of the centrifugalized and filtered liquids could be explained on the assumption that the former still contained, even after prolonged and repeated centrifugalization, a sufficient remnant of stromata to cause distinct agglutininogenic and hzemolysinogenic reactions. But this is by no means likely, as the result of centrifugalization was controlled by the microscope, and no liquid was injected which was seen to contain many ghosts. It is more probable that in laking some of the agglutininogenic and hzmolysinogenic substances are extruded from the corpuscles in colloidal solution, or contained in those gran- ules which are commonly seen in laked blood, and that in either case the pores of the clay refuse them passage. The agglutinin in typhoid serum, it is said, is removed by filtration through porcelain. The pot used in these experiments was shown (see protocols) to remove the greater part of the haemolytic power of normal dog’s serum for rabbit’s corpuscles. This was apparently due in part to the removal of complement, but not entirely, since the haemolytic power could not be fully restored by theaddition of rabbit’s serum. The agglutinating power of the unfiltered serum was relatively feeble, but it was dis- tinctly diminished by filtration. The fact that the agglutininogens and hzemolysinogens in the liquid of laked corpuscles are apparently also incapable of passing through such a filter, strengthens the view that the substances in blood which when injected into an animal give rise to the formation of agglutinins and haemolysins are nearly akin to the bodies whose production they cause. Since the hzmoglobin passed through the pot‘used in these experiments, the negative result confirms the statement of Ford and Halsey? that solutions of pure hzemoglobin do not lead to the formation of hemolytic or agglutinating substances. In Table II are given the results of the further examination of the sera of Rabbit A and Guinea-pigs F and J’ already reported on in the former paper. These animals were kept alive with the view of testing how long the specific alterations in the serum persist. 1 ForpD and HALSEY: Loc. cit. Experiments on the Action of Laked Corpuscles. 367 The serum of Guinea-pig J’, which, 21 days after the last injection of the stromata of heat-laked rabbit’s corpuscles, gave good agglutina- tion of rabbit’s corpuscles, caused, 107 days after the last injection, only very moderate agglutination, although still very fair laking in comparison with that produced by normal guinea-pig’s serum. The serum of Guinea-pig F, which, 41 days after a single injection of nor- mal rabbit’s’ corpuscles, caused good agglutination and laking of rabbit’s corpuscles, still after 180 days produced fair agglutination and laking, in comparison with the serum of the practically normal Guinea-pig R (Table I), although distinctly less than before. Evi- dence was obtained of the presence of specific intermediary body in the serum of F after 180 days. The serum of Rabbit A, which, 18 days after the last injection of dog’s formaldehyde-fixed corpuscles, WABER, IT: Days after Material injected. last injection. Agglutin- ee Laking. Stromata of heat-laked rabbit’s | { Guinea-pig J’ 107 Some Very fair. corpuscles (two injections) . ‘s 21 Good Good. Normal rabbit’s corpuscles | § Guinea-pig F 180 Fair Fair. (one injection) fe iake se 41 Good Strong. 18 Good Fair. Dog’s formaldehyde-fixed cor- } Rabbit A 148 Slight Very slight. puscles (two injections) . gave good agglutination and fair laking of dog’s corpuscles, caused, 148 days after the last injection, only a little agglutination and very little laking. The agglutination, however, was better marked than the laking, just as when the first test was made. So far as one can say from a determination of two points on the curve, the decline in agelutinating, ran parallel to the decline in lytic power, the latter being practically extinguished by the time of the second test. There was no evidence that the feebleness of the hemolytic action was due to deficiency of complement since the addition of normal dog’s serum did not increase it. SUMMARY, The results of the previous paper in regard to the agglutininogenic and hzmolysinogenic power of the stromata and liquid of water- laked corpuscles receive additional support. Both cause the pro- 368 G. N. Stewart. duction of sera with specific haemolytic and agglutinating power, the latter being most marked, as is also the case with the serum obtained after the injection of stromata laked by freezing and thaw- ing. Filtration through porous earthenware appears to remove the agglutininogens and hemolysinogens from the liquid of corpuscles laked by freezing and thawing and by foreign serum. Small Rabbit F. Jay 6, May 16, and Fune 15.— Injected liquid from water-laked dog’s corpuscles corresponding, respectively, to 15 c.c., 18 c.c., and 15 c.c. of blood. For the first two injections, the liquid was centrif- ugalized, after the addition of sodium chloride, till practically no ghosts could be seen with the microscope. For the third injection it was filtered through clay. Fuly 9. — Got serum (F). Made the following experiments : (1) S*+0.2 cc. F. 40m (40°),* complete L.* (2) S+0.1cc. F. 40m (40°), good A.* 55 m (40°), L not yet complete; 14 h (r),* nearly complete. (3) S+ 0.05 c.c. F. 40 m (40°), little L. 115 m (40°) and 14h (r), some L, but not much. 22m (40°), marked A. Good A in few minutes at 40°. Repeated with S’.. 15 m(40°), no L in any, but good A inall. 43m (40°), slight L in (1), none in (2) and (3); 14h (r), no L in (3), very little in (2), some L in (1), but not half laked. In all L is much less than in the corresponding experiments with S. Repeated (1), (2), and (3) with S and F serum which had been heated to 58° for 20 minutes; 60 m (40°),no L; 14h (r), slight L; 8 m (40°), excellent A in (1), and good A in (2) and (3). Repeated (1), (2), and (3) with S and heated F serum diluted with 9 times its volume of NaCl solution; 4 h (40°), no A or L in any. Repeated (1), (2), and (3) with S and normal rabbit’s serum, and with S and normal rabbit’s serum previously heated to 58°; 24 h (40°), and 18 h (0°), no A or L in any. * In the protocols, S means 0.5 c.c. of a 5 per cent suspension of washed dog’s corpuscles where rabbit’s serum was being tested, and of washed rabbit’s cor- puscles where guinea-pig’s serum was being tested. S’ means 0.5 c.c. of a Io per cent dilution of the entire blood of dog or rabbit, respectively. Where smaller quantities of serum than o.1 c.c. were to be added to o.5 c.c. of suspension, the requisite dilution was obtained by adding more than 0.5 c.c. of suspension to 0.1 c.c. of serum; but, for simplicity in the protocols, the amount of serum is stated as if it were always added to 0.5 c.c. 40 m (40°) means 40 minutes at 4o° C. 14 h (r) means 14 hours at room-temperature. 14h (0°) would mean 14 hours in the ice-chest. A stands for agglutination, L, for laking. (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) Experiments on the Action of Laked Corpuscles. 369 Mixed 1 c.c. of a sediment of washed dog’s corpuscles with 1 c.c. F serum, both previously cooled to 0° C.; kept 2} hours at o°, then took off the serum (a), washed the sediment twice in the centrifuge, and made a 5 per cent suspension of it (8). Performed the following experiments : S$ + 0.2 c.c.a+ 0.1 c.c. NaCl solution. S$ +0.1 c.c. a+ 0.1 c.c. NaCl solution. S+ 0.2 c.c.a+ 0.1 c.c. normal dog’s serum. $+0.1 cc. a+ 0.1 c.c. normal dog’s serum. S$+0.1cc.a+0.1 c.c. heated F serum. S + 0.1 c.c. normal rabbit’s serum + 0.1 c.c. heated F serum. After 50 m (40°) and 18 h (0°), fair L in (8), better than in any of the Experiments (4) to (14), except (12). Some A in (4), (6), and (8). Good A in(g). Fair Lin (g). Slight L in (4) and (5). No L in (6) and (7). 0.5 c.c. 8 + 0.1 c.c. normal dog’s serum. 0.5 c.c. 8 + 0.2 c.c. normal dog’s serum. 0.5 c.c. 8 + 0.2 c.c. normal rabbit’s serum. 05c.c.8+ 0.2 c.c. serum a. (14) 0.5 c.c. 8 + 0.2 c.c. NaCl solution. After 50 m (40°), 18 h (0°), fair A in (12) and (13); some A in the others. Good L in (12), much the best of all the Experiments (4) to (14). Slight L in (10), and some in (13). Little or no Lin (11). No L in (14) Small Rabbit G. May 7, May 16, and Fune 17.— Injected the stromata (2) (3) (5) (6) (7) (8) of dog’s water-laked corpuscles corresponding, respectively, to 20 c.c., 20 c.c., and 15 c.c. of blood. Fune 25. — Got serum (G). Heated some to 57° for 20 minutes. S$+0.2c.c.G. 20 m (36°), distinct L. 60 m (36°), complete L, preceded by A. $+0.1c.c. G. 60m (36°), good A, comparatively little L. 2h (36°), very fair L. S+0.lc.c. NaCl solution. 90m (36°), no A or L. S + 0.2 c.c. G (heated). 10m (36°), marked A. 90m (36°), no L. Mixed a sediment of dog’s corpuscles (at 0°) with G serum (at 0°), and left 1} hours at o°. The corpuscles were well agglutinated. Now separated the serum (a), washed the sediment, and made a 5 per cent suspension of it (). 0.5 c.c. 8 + 0.2 c.c. dog’s serum. Pretty good A. 13 h (36°), and 1 h (r), fair L, but not complete. $+ 0.1 c.c.G+ 0.1 c.c. dog’s serum. S+0.1 c.c. a+0.1 c.c. dog’s serum. S + 0.2 c.c. NaCl solution. 370 G. NV. Stewart. (9) S+0.1 c.c. heated G serum + 0.1 c.c. dog’s serum. (10) S+ 0.1 cc. heated G serum + 0.1 c.c. 2+ 0.1 c.c. dog’s serum. 50 m (36°), 1h (r). Fair L. (11) S+0.1 cc. heated G serum + 0.1 c.c. a. 50m (36°), 1h (r). Fair L. Agglutination is absent or inconspicuous in Experiments (5) to (11). Little if any L in Experiments (6) to (g) in 13 h (36°) and 1 h (2). Small Rabbit J. May 29, May 80, and Fune 25.— Injected liquid from dog’s corpuscles laked by freezing and thawing, corresponding, respec- tively, to 15 c.c., 10 c.c., and 12 c.c. of blood. The liquid was filtered through clay. _ Fuly 16. — Got serum (J). Heated some of it to 58° for 20 minutes. (1!) *S=FO Mere: (8) S + 0.1 cc. heated J + 0.1 cc. (2) eS) S005 cies dog’s serum. (3) S+0.1 c.c. heated J. (9) S + 0.05. c.c. heated J+ 0.1 cic. (4) S+ 0.05 c.c. heated J. dog’s serum, (5) S+ 0.1 c.c. NaCl solution. (10) S+0.2 cc. J. (6) S+01 cc. J +01 c.c. dog’s serum. (11) S+0.2c.c. J + 0.2 c.c. dog’s serum. (7) S + 0.05 cc. J + 0.1 cc. dog’s (12) 0.5c.c.ofa5 %suspension of J’s washed serum. ; corpuscles + 0.1 c.c. dog’s serum. In 14 h (40°) no A or L in any of Experiments (1) to (11). Even after 15 h (40°), and 24 h (0°) very little L. In (22) complete L in 20 minutes, and no doubt sooner. Rabbit K. May 30 and Fuly 22.— Injected all the washed stromata which could be separated from the corpuscles of 20 c.c. and 22 c.c., respectively, of dog’s blood laked by freezing and thawing. Fuly 24. —Got serum (K). Heated some to 57° for 20 minutes. (1) S+0.1 cc. K. (5) S+ 0.1 cc. heated K + 01 cc. (2) 3S Ol0sicic: Ke dog’s serum. (3) S+0.1c¢.c. NaCl solution. (6) S+ 0. cc. heated K + 02 cc. dog’s (4) S+0.1c.c. heated K. serum. (7) S+ 0.025 K. (8) S+ 0.025 K + 0.4 c.c. dog’s serum. No L after 2 hours and 40 minutes at 40°, and 12 hours at o°. Fair but not strong A in all except (3) ; seems strongest in (4). Guinea-pig F.— Injected washed rabbit’s corpuscles on Jan. 20. July 19. — Got serum (F). Heated some to 58° for 15 minutes. (1) S+01l cc. F. 25 m (40°), almost complete L, preceded by A. L began even at room-temperature immediately after addition of serum. Half this amount of F serum also caused fair laking (see protocol of Guinea-pig R). (2) S+0.1 cc. heated F. (4) S+0.1 c.c. heated F + 0.1 c.c. rabbit’s (3) S+ 0.05 c.c. heated F. serum. (5) S+ 0.05 c.c. heated F + 0.1 c.c. rabbit’s serum. Experiments on the Action of Laked Corpuscles. 371 After 4 h (40°) very little L in (2) to (5), but fair A in all. Added F serum at o° to washed rabbit’s corpuscles at 0°, and kept at o° for 3 hours and 20 minutes. A little L, fair A. Now separated the serum.and diluted it with 4 volumes NaCl solution. Call the dilute serum a. Washed the sediment and made a 5 per cent suspension ({). (6) S+0.5 cc. a. (11) S + 0.6 c.c. NaCl solution. (7) S+0.25 c.c. q (12) S + 0.25 c.c. a + 0.05 c.c. heated F. (8) S+0.5 c.c. NaCl solution. (13) S + 0.3 c.c. NaCl solution. (9) S + 0.25 c.c. NaCl solution. (14) 0.5 c.c. of 8 + 0.1 c.c. NaCl solution. (10) S+0.5c.c.a+0.1 c.c. heated F. (15) 0.5 c.c. of B + 0.1 c.c. rabbit’s serum. After 45 m (40°), fair L in (10) and (12), none in the others. After go m (40°), nearly complete L in (10) ; good L, but not complete, in (12); none in the others. Fair A in (14) and (15). Same after 3 h (0°). Guinea-pig J’, 850 grams. Yan. 26 and April 8.—Injected the heat- laked stromata of rabbit’s corpuscles corresponding, respectively, to 7 c.c. and 3.5 c.c. of blood. Fuly 24.— Got serum (J’). Heated some to 57° for 15 minutes, and 58° for 5 minutes. (1) S+Olcc. J’. 15 m (40°), good L, not quite complete. 85 m (40°), complete L. No doubt complete earlier. (2) S+0.05 cc. J’. 15 m (40°), very fair L,and some A. 30 m (40°), L is nearly as great as in (5) at same time. 85 m (40°), good L, though less than in (5). Some A. 3h (40°), L still incomplete. (3) S+0.1 cc. heated J’. 85 m (40°), no L; some A, but not marked. (4) S+0.1. cc. J’+ 0.1 cc. rabbit’s serum. 10m (40°), complete L. (5) S+ 0.05 c.c. J’ + 0.05 c.c. rabbit’s serum. 15 m (40°),some L; 80m (40°), almost complete L; 3 h (40°), complete L. Some A, but not strong. (6) S+0.025 c.c. J’. 1h (40°), some L; 23 h (40°), fair L. Some A. (7) S + 0.025 c.c. J’ + 0.05 c.c. rabbit’s serum. 1h (40°),some L, but less than in (6). 24 h (40°), L about same as in (6), and also A. (8) S+ 0.025 c.c. heated J’. 24h (40°), A as in (6). No L. (9) S+0.05 c.c. NaCl solution. No Lor A. (10) S+0.1 c.c. heated J’+ 0.1 c.c. rabbit’s serum. 40 m (40°),some L. 55 m (40°), good L; in one hour more, nearly complete L. (11) S+0.1c.c. heated J’ + 0.2 c.c. rabbit’s serum. 75 m (40°), complete L. (12) S + 0.025 c.c. J’ + 0.1 c.c.rabbit’s serum. 2 h (40°) and 12 h (0°), no L; fair A. A is not conspicuous in any of Experiments (1) to (12). Guinea-pig R, 750 grams. une 4 and Fune 25. — Injected the liquid from rabbit’s corpuscles laked by dog’s serum corresponding, respectively, to 9 c.c. and 6 c.c. of blood. Fuly 18. — Got serum (R). Heated some to 57°. Made the follow- ing experiments to test the laking and agglutinating power of the serum and to compare it with the serum of Guinea-pig F. 372 @ (2) (5) (6) (8) (9), (10) G. N. Stewart. $+ 0.le¢c. R. 40 m (40°),no Lor A. After 11 h (0°), some L,but much less than in (3). S+ 0.05 c.c. R. 40 m (40°) and 11 h (0°), only the very slightest L, and little, if any, A. $+0.l cc. F. 25 m (40°), almost complete L, preceded by A. S + 0.05 cc. F. 25 m (40°), fair L and A. 40 m (40°) and 11 h (09), fair L, better than in (1), and good A. Made a 10 per cent dilution of normal rabbit’s blood @’). S’+ 0.1 cc. R. 25 m (40°), no L; very little after 11 h (0°), distinctly less than in (1). LittleornoA. , S’+ 0.05 c.c.R. NoL or A. (7) S’+0.025cc.R. NoLorA. S*+ 0.l cc. F. 14 m (40°), complete L. S’+ 0.05 cc. F. 14m (40°), some L. 25 m (40°) and 11h (0°), fair L, but less than in (4). Good A, S + 0.025 c.c. F. Fair L, but less than in (9) under same conditions. Good A, much better than in (2). Guinea-pig S$, 800 grams. $une 3 and Fune 26. Injected the stromata (1) (2) (6) (7) (8) (9) (10) of rabbit’s corpuscles laked by dog’s serum corresponding, respectively, to 12 c.c. and 6 c.c. of blood. Sune 28. — Got serum (Ser). Heated some to 57° for 20 minutes. S + 0.1 c.c. Ser. (3) S+0.1 c.c. NaCl solution. S$+0lcc. Ser + 0.1 c.c. rabbit’s (4) S+0.05 c.c. Ser. serum. (5) S+0.1 c.c. heated Ser. In (1) some L in 12 m (36°), but by no means complete even after 45m (36°) and 12h (r). In (2) good L in 20 m (36°), not quite com- plete. Nearly complete in 80 m (36°), and distinctly better than in (1). In (4) very slight L in 80 m (36°), although more than in control (3). After 12 h (0°) slight but distinct L in (4). No Lin (5). A is absent or inconspicuous in all. Added to a sediment of washed rabbit’s corpuscles at o° serum of Guinea-pig S at 0°, and left at o° for 14 hours. Separated the serum (a) and made a 5 per cent suspension of the washed corpuscles (f). 0.5 c.c. 8 + 0.2 c.c. rabbit’s serum. (11) S+0.lc.c. heated Ser + 0.1 cc. 0.5 c.c. 8 + 0.2 c.c. NaCl solution. rabbit’s serum. S + 0.l cc. Ser + 0.1 c.c. rabbit’s (12) S +0.1 c.c.a+ 0.1 c.c. heated Ser + serum. 0.1 c.c. rabbit’s serum. $+ 01 cc. a+4+0.1 c.c. rabbit’s serum. (13) S+0.1c.c. a+ 0.05 heated Ser. S + 0.2 c.c. NaCl solution. (14) 05 cc. 8+ 0.2 c.c. a. (15) 0.5 c.c. 8 + 0.2 c.c. rabbit’s serum. (16) 0.5 c.c. 8 + 0.2 c.c. NaCl solution. After 34 h (36°), no L in any of Experiments (6) to (16), except in (8), where L is very fair, though not complete, and in (12), where there is a trace of L. A is absent or inconspicuous in all, except in (6) and (7). Experiments -on the Action of Laked Corpuscles. 373 Filtered serum. — Filtered dog’s serum through the clay pot, throwing away (1) (2) (3) (4) the first part of filtrate. Call unfiltered serum D and filtered D’. Heated some of D and D! to 57° for 25 minutes. During the heating nearly all the blood-pigment went out of solution in D’/, while no change occurred in D. The spectrum of the heated D’ showed feeble oxyhzemoglobin bands. Madea 5 per cent suspension of washed rabbit’s corpuscles (S) and a ro per cent dilution of rabbit’s blood (S’). S’+ 0.05 c.c. D. 40 m (40°), partial L; 21 h (40°), very fair L. S’+ 0.1. cc. D. 40m (40°), complete L. S’+ 0.05 c.c. D’%. 5h (40°), no L. S’4+ 0.l cc. D’. 2¢h (409), no L. 5h (409), slight but distinct L. Repeated (1) to (4) with S instead of S’. Practically the same result. Repeated (1) to (4) with S instead of S’,and heated D and D’ instead of unheated. No Linany. Fair A in (2), none in the rest. $+ 0.1 cc. heated D+ 0.2 c.c.D’. 70 m (40°), some L. 90 m (40°), 17 h (09), fair A, very fair L (quite half laked). S+0.1¢.c heated D + 0.2 cc. rabbit’s serum. | 14 h (40°), 17 h (09). ; S$+0.1 cc. D’ + 0.2 cc. rabbit’s serum. No A or Lin (6) or (8). Slight 5 ioe L but no A in (7). Fair Lin S+ 0.1 c.c. heated D’ + 0.2 c.c. rabbit’s serum. (9), but distinctly less than in S + 0.1 c.c. NaCl solution + 0.2 c.c. D’. in (5). Fair A in (9). D, whether heated or not, gives, in general, much better A than D’. FURTHER PROOF OF ION ACTION IN PHYSIOrsGre PROCESSES. By C. HUGH NEILSON Ann ORVILLE H. BROWGR: [From the Department of Physiology in the St. Louis University.] HE purpose of this paper is to present additional evidence in support of the statement made by us! in a previous paper, that the results obtained from the action of electrolytes upon the decom- position of hydrogen dioxide by platinum black, and by a watery extract of pancreas, could be explained, in general, by the assumption that the cations or positive ions have a depressing or retarding action and the anions or negative ions have a stimulating action. Kastle and Loevenhart 2 in a recent article in Science took exception to the use of the term ‘ion action” by physiologists and pharmacologists in general, and by the writers of this article in particular. Our statement, that the negative ion stimulates, and the positive ion in- hibits catalytic action does not imply that the negative or positive charge is necessarily the active factor. We merely gave the facts and offered no explanation as to how the stimulation or depression was produced. The influence of ions may be due solely to the charge, and to the physical properties of the ion, as given by Dr. A. P. Mathews® in a recent paper. Again, it may be explained, according to Kastle and Loevenhart, by the formation of insoluble films. It seems to be just for us to remark that Kastle and Loeven- hart apparently see no explanation for the stimulating action which some electrolytes exert upon the decomposition of hydrogen dioxide by finely divided metals. Again, will this theory of an insoluble film hold when soluble ferments are substituted. for the finely divided metals ? The theory of the dissociation of a compound into ions has been so thoroughly substantiated by such numerous and brilliant examples 1 NEILSON and Brown: This journal, 1904, x, p. 335. 2 KASTLE and LOEVENHART: Science, 1904, xix, p. 630. 8 A. P. MATHEWS: This journal, 1904, xi, p. 455. 374 Further Proof of Lon Action in Phystologic Processes. 375 that it need no longer be considered atheory. Jones! says, “ Since the theory was proposed, it has been tested both theoretically and experimentally from many sides; with the result that, when all the evidence available is taken into account, the theory of electrolytic dissociation seems to be as well established as many of our so-called laws of nature.” Dry metallic sodium and dry sulphuric acid do not react with each other. Dry hydrochloric acid and dry fumes of ammonia may be mixed without chemical change. Hydrochloric acid in toluol pos- sesses none of its acid properties, 2z. ¢., it is not dissociated. Osmotic pressure depends on the number of particles in solution. Ionization explains why a solution of sodium chloride has a higher osmotic pressure than a sugar solution of the same concentration. This also explains the too high molecular weight of salts when determined by the freezing point methods. It also places the question of chemical equilibrium and the facts of precipitation on a definite mathematical basis. Substances, as urea, sugar, etc., known as non- electrolytes are relatively inactive chemically, while the electrolytes, or acids, bases, and salts, are active chemically. In the application,of the Arrhenius theory of electrolytic dissocia- tion to the explanation of physiological processes, the names of Loeb in animal physiology, and Kahlenberg and True in plant physiology, stand out as pioneers. Loeb? found that the gastrocnemius muscle of the frog contracts rhythmically only in solutions of electrolytes, the non-electrolytes being ineffective. Kahlenberg and True® found that solutions of acids, bases, and salts, completely dissociated, owe their toxicity on the bean Lupinus albus L. to the specific ions in solution, Lingle’s* work with electrolytes and non-electrolytes on turtle’s heart further substantiates the view that the phenomenon of a contracting heart strip is dependent on specific ions. A. P. Mathews ® shows that the physiological effect of electrolytes is due to ions, and that this action is due to the electrical and physical properties of these ions. Howell’s work on the heart, Ringer’s work, and the work of a host of others can be satisfactorily explained by the action of ions in so far as their action on protoplasm is known. 1 Jones: The Modern Theory of Solutions, p. 9. 2 Logs: Archiv fiir die gesammte Physiologie, 1899, lxix, p. 99. 8 KAHLENBERG and TRUE: Botanical gazette, 1896, xxii, p. 81. * LINGLE: This journal, 1900, ii, p. 205. 5 MATHEWS: Loc. cit. 376 C. Hugh Nelson and Orville H. Brown. The inversion of sugars by acids may, perhaps, be considered as truly a physiological process as when it is brought about by animal or vegetable diastases. It has long been known that the power of an acid to invert a sugar is in direct ratio to the number of hydrogen ions present. The strongly dissociated hydrochloric acid splits ina given time an amount which may be designated as 100 per cent, while the weakly dissociated acetic acid splits in the same length of time only a fraction of 1 per cent. The importance of hydrochloric acid in the stomach in a concentration where dissociation is very great is an example of ion-function well known to all. Neilson and Brown,! Kastle and Loevenhart,? Cole,? McGuigan,! and others have demonstrated clearly that enzymic activity in the presence of a salt solution is dependent on the ions in the solution. The experimental data of this paper will be presented in three divisions, as follows: Part I, the influence of the non-electrolytes upon the decomposition of hydrogen dioxide by platinum black and a watery extract of kidney; Part II, the influence of salts with less degree of ionization than normal, upon the decomposition of hydrogen dioxide by platinum black; Part III, the effects of a compound with one ion suppressed, upon the decomposition of peroxide of hydrogen by platinum black, and a watery extract of kidney. METHODS. The platinum black was mixed with distilled water and measured out while the mixture was being constantly stirred. The constancy of the controls showed that equal amounts of the metal were measured out each time. A watery extract of kidney was used instead of the watery extract of pancreas which was used in our former experiments. The kidney extract is more stable than the pancreas extract, in which autolytic processes are rapid. The kidney extract is also more active than the pancreas extract, and at the strength necessary to produce the desired rate of evolution of gas it has much less albuminous material than the pancreas extract producing the same evolution of oxygen, and therefore it is much more accurately and easily measured by a pipette. Fresh beef kidney was used, and the extract was made 1 NEILSON and Brown: Loc. cit. 2 KASTLE and LOEVENHART: American chemical journal, 1903, xxix, p. 563. 3 COLE: Journal of physiology, 1903, xxx, p. 202. * McGuIGAN: This journal, 1904, x, p. 444. further Proof of Lon Action in Physiologic Processes. 377 by mincing and grinding the kidney in a mortar, adding water, and filtering through filter paper. Two large-mouthed bottles of 200 c.c. capacity were fitted with double-holed rubber stoppers. From one hole of each stopper, by means of rubber and glass tubing, a connec- tion was made with a eudiometer tube of 75 c.c. capacity. After the solution to be tested and the kidney extract had been put into the bottles, the stoppers were adjusted, and through the second holes in the stoppers, which were about a quarter of an inch in diameter, the hydrogen dioxide was introduced simultaneously by means of two pipettes. Immediately and at the same time the two holes were closed by glass rods. By this method of procedure and the use of the AARC Ry Ee PLATINUM BLACK. CONCENTRATIONS. SUBSTANCES. Cubic centimetres of oxygen given off in . ° | . 5 . = 5 . 1 min. | 2min.! 1 min.| 2 min.) 1 min.| 2 min.}| 1 min.| 2 min. SUCAm mya Kose colts: Te A, 11 20 27 27 13 27 Wivcaeicea ate tls ne dacs 2 16 Zi, 13 27 Glvcenilemee sti 4. Ge 27 28 12 27 micohols(ethyl) |. . . . 26 27 12 27 W ater watery extract of kidney or the platinum black, prepared as given above, the control experiments did not vary more than I c.c. of oxygen without some apparent cause. The amount of oxygen given off was recorded at the end of one and two minutes. To insure the free liberation of the gas, the bottles were shaken. In order to have a uniform shaking, the bottles were fitted in holes in a block of wood which was shaken by hand. The shaking might have been done by a mechanical device, but the results were so constant that variations in shaking can be ignored. Unless otherwise stated, the amount of water or the solution to be tested was 25 c.c.; the amount of kidney extract, 5 c.c.; the amount of hydrogen dioxide, 10 c.c. The kidney 378 C. Hugh Neilson and Orville H. Brown. extract and the peroxide were kept in ice-water, and the solution was kept at room-temperature. The dioxide of hydrogen used was that made by the Mallinckrodt company of St. Louis. I. THe ActTIon oF NoN-ELECTROLYTES. Solutions of non-electrolytes, as urea, glycerin, sugar, etc., in con- centrations isotonic with the blood, do not stimulate nerves. If the solution has a concentration giving an osmotic pressure of fourteen atmospheres or more, the nerve is stimulated if placed in it. This is due to a purely physical process, — 2. ¢., the extraction of water from the nerve. By the use of non-electrolytes we have a way of deter- TABLE II. KIpNEY EXTRACT. CONCENTRATIONS. SUBSTANCES, Cubic centimetres of oxygen evolved in 1 fae a Al 29 2 min. min.} min.| min.| min. Sugar Urea. Glycerin . Alcohol (ethy]) Water. mining whether molecules as such have any action upon the decom- position of the peroxide of hydrogen by platinum black and a watery extract of kidney. Weiss! and Arnheim? found that a solution of gum-arabic accelerated pepsin digestion; but Mugdan? showed pepsin digestion to be delayed by gum-arabic, Meisenheimer * found zymase action delayed by glycerin solutions. Braeuning,® who tried 1 WEISS: Zeitschrift fiir physiologische Chemie, 1897, xl, p. 488. 2 ARNHEIM: Zeitschrift fiir physiologische Chemie, 1897, xl, p. 238. 8 MuGDAN: Berliner klinischen Wochenschrift, 1891, p. 788. * MEISENHEIMER: Zeitschrift fiir physiologische Chemie, 1897, xxx, p. 520. 6 BRAEUNING: Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 80. Further Proof of lon Action in Physiologic Processes. 379 many non-electrolytes on different kinds of ferment action, sums up his work by saying, the stronger the solution of the non-electrolyte, the more the ferment action will be delayed. Our results are similar, as is shown in Tables I and II. When the concentration is 5 mol or more, there is considerable inhibition; but in the 1 mol and weaker there is little or no effect. Urea has a greater inhibitory action than the other substances in the same concentration. This may be explained by the fact that ionization does occur to a limited extent in urea. This was shown by the ability of the solution of urea which we used to conduct the current better than could the distilled water from which it was made. TABLE III. Methylalcohol 4, water }. uy. Methyl-alcohol. - 59.82 62.46 65.36 67.11 69.26 70.53 II. SALTS WITH LESSENED IONIZATION. The power of dissociating the molecules of a salt into its ions differs in various solvents. Water is the strongest ionizer known, with the possible exception of hydrogen dioxide. The alcohols are considerably weaker in their ionizing power. Wakeman! showed that the addition of water to a solution of an organic acid in ethyl- alcohol increased proportionately the conductivity of the solution. Zelinsky and Krapiwin ? found, however, that substances dissolved in methyl-alcohol presented an exception. It was shown by these workers that certain salts had a higher conductivity in pure methyl- 1 WAKEMAN: Zeitschrift fiir physikalische Chemie, 1893, xi, p. 49. 2 ZELINSKY and KRAPIWIN: Zeitschrift fiir physikalische Chemie, 1896, xxi, p- 189. 380 C. Hugh Nelson and Orville H. Brown. alcohol than in a 50 per cent solution. This is shown by Table III taken from the work of Zelinsky and Krapiwin, in which v equals the © number of litres containing a gram molecule, and wy equals the molec- ular conductivity. It occurred to us that, if the stimulating or depressing power of a salt upon the hydrolysis of hydrogen dioxide by platinum black, depended at all upon the ionic condition of the salt,-we should be able to obtain a difference in the amount of oxygen given off when the salt was dissolved in water alone, and when it was dissolved in equal parts of water and methyl-alcohol. Of course the retarding influence of the alcohol was necessarily taken into consideration. As TABLE IV. PLATINUM BLACK. Cubic centimetres of oxygen liberated in SUBSTANCES. 1 min. 2 min. Water SEAS cua Nea Rod 60 88 50% micthyl-alcohols. Gs jah 6 eof. 24 35 2 calciumiehlonide mwater 6 © . » « 24 40 * calcium chloride in 50% methyl-alcohol . . 11 19 SWALCH.o) psu ee ee ateeees. seuee) al ses 60 50% methyl-alcohol zo Mercuric chloride in water . zip Mercuric chloride in 50% methyl-alcohol . the methyl-alcohol has a precipitating action upon proteids, the experiment was not performed with the kidney extract. The results will be seen in Table IV. It will be observed from Table IV that the amount of oxygen given off in one minute in the methyl-alcohol is 36 c.c. less than that in pure water. This represents the inhibitory effects of the alcohol. In the calcium chloride there is also 36 c.c. less than in the water. The sum of these, 72 c.c., then represents the inhibitory effects of the mix- ture of calcium chloride and methyl-alcohol, if each one exerts, when mixed, the full power that it possesses when separate. But these further Proof of Lon Action in Phystologic Processes. 381 two substances mixed together allowed 11 c.c. to come off. There- fore their inhibitory power when mixed is the remainder of sixty minus eleven, which is forty-nine. Then the lessened inhibitory power is the remainder of seventy-two minus forty-nine, which is twenty-three. The only explanation that we can see for this is the decreased dissociation of the calcium chloride when dissolved in equal parts of methyl-alcohol and water. By making a calculation of the amounts at the end of the second minute, it will be found that the lessened inhibition corresponding to the lessened dissociation is 32 c¢.c. of oxygen. When mercuric chloride is used instead of the calcium chloride, the result is still more striking. By calculating as before, it will be found that the lessened inhibition of the mercuric chloride for one and two minutes is respectively 35 c.c. and 50 c.c. of oxygen. This was also tried upon the stimulating salts and was found to hold. Ill. Tue Errect or LESSENING THE CONCENTRATION OF EITHER THE POSITIVE OR NEGATIVE ION. In qualitative and quantitative chemistry, advantage is taken of the fact that the addition of a common ion to a substance in solution lessens the dissociation of that substance. Based upon the equi- librium of electrolytes, the fact is well known that a weak base, as ammonia hydroxide, is made still weaker by the addition of an am- monium salt or another hydroxide. Likewise a weak acid is made still weaker by the addition of one of its salts. For example, acetic acid by the addition of sodium acetate has a much weaker acid effect. The addition of the common acetate ion depresses the dissociation of the acetic acid, and thus lessens the number of hydrogen ions. This is shown by the following formula: Bia CH..COO Ch COOr acetic acid dissociation = 754,55. During the first interval of time after adding sodium acetate the formula becomes H x CH,COO + ~« acetate ions CH,COOH equilibrium is established we have Equilibrium equation = = , or the constant of > for acetic acid. When constant (H—y) (CH,COO + +—y) _ , CH,COOH + y That is, acetate ions plus hydrogen ions forms molecular acetic acid, which decreases the product of ion concentration in the nu- merator, and increases the denominator, thereby decreasing the dis- 382 C. Hugh Neilson and Orville H. Brown. sociation and the number of H ions. In the application of this fact - to physiological problems not much experimental work has been done. Paul and Kronig! found that the addition of sodium chloride to mercuric chloride caused the toxicity or the germicidal action of the latter to be lessened. They found that anthrax spores lived longer in a mixture of the sodium and mercuric chlorides than in the TABLE V. Cubic centimetres of oxygen liberated in SUBSTANCES. 1 min. 2 min. ZOO? WALEI srt 7, CeO ae kee he baits os 46.0 72.0 20 c.c. water + 5 ¢.c. 3%5 mercuric chloride. . . . 15.0 26.0 20 c.c. | calcium chloride +5c.c. water... . . 29.0 46.0 20 c.c. '% calcium chloride + 5 c.c. 335 mercuric chloride 5g ob NG 5.0 25 c.c. water. 23 c.c. water + 2 c.c. =%, mercuric chloride . 200 23 c.c. ¢ calcium chloride + 2 c.c. water . at 23 c.c. § calcium chloride + 2 c.c. 335 mercuric chloride ; Seer =) As 25 c.c. water . 24 c.c. water + 1 c.c. 35 mercuric chloride . 24 c.c. 2 calcium chloride + 1 c.c. water . 8 24 c.c. § calcium chloride -+ 1 cc. 335 mercuric chloride 5 oe pe OO Oe eee mercuric chloride alone. Their explanation for this was, that the common chlorine ion decreases the dissociation of the mercuric chloride, and therefore the number of mercuric ions is lessened. These workers found also that the germicidal action of mercuric chloride, bromide, and cyanide is dependent in the main on the con- centration of the mercuric ions, z. ¢., upon the degree of dissociation 1 PauL and Kronic: Zeitschrift fiir Hygiene und Infektions-Krankheiten, 1897, xxv; cited from COHEN’S Physical Chemistry. Further Proof of lon Action in Phystologic Processes. 383 ‘of the salts. Similar results were also obtained with gold, silver, and copper salts. In testing the concentration of the positive ions upon the rate of catalysis of hydrogen dioxide, we used mercuric chloride as the solution to be tested, and to this was added sodium chloride or cal- cium chloride. The addition of the common chlorine ion decreases the dissociation of the mercuric chloride by increasing the amount of TABLE VI- KIDNEY EXTRACT. Cubic centimetres of oxygen a : liberated in SUBSTANCES. 1 min. 2 min. LSVIC (Ens WELISIO” Eg on Camas. sof "idee c= Op oe Gun mo 15.0 30.0 20 c.c. water + 5 c.c. 335 calcium acetate . .. . 26.0 45.0 20 c.c. | calcium chloride + 5c¢.c. water. . .. . 4.0 7.0 20 c.c. ¢ calcium chloride + 5 c.c. 55 calcium acetate 10.5 19.0 POMOC EWOtGEE fe 6% bos) ieee en sep ee Ae 30.0 23 c.c. water + 2 c.c. sjy calcium acetate... . 36.5 23 €.¢, = calcium chloride +- 2 c.c. water... =. « 6.0 23 c.c. % calcium chloride + 2 c.c. =} calcium acetate 25 c.c. water 24 c.c. water + 1] c.c. 545 calcium acetate 24 c.c. | calcium chloride + 1 c.c. water . 24 c.c. 3 calcium chloride + 1 c.c. <}> calcium acetate molecular mercuric chloride. Ifnow the mercury ion has a depress- ing action on the rate of the decomposition of hydrogen peroxide by platinum black or a watery extract of pancreas, the addition of calcium chloride ought to lessen this depressing action. Such was found to be the case. By consulting the first part of Table V where 5 c.c. of 5%, mercuric chloride were used, it will be noticed that 72 c.c. of oxygen were liberated at the end of the second minute in water. Where the mercuric chloride was used, 26 c.c. were given off. 384 C. Hugh Nelson and Oroville H. Brown. The depression of the mercury then was 46 c.c. of oxygen. Where the calcium was used alone, 46 c.c. of oxygen were freed. The depressing effect of the calcium was 26 c.c. of oxygen. When these two salts are mixed, if each exerts its full inhibitory power, there should be an inhibition of the normal catalysis of 72 c.c. of oxygen, or there should be nothing given off. This, however, is found not to be the case. 9 c.c. of oxygen were liberated at the end of the second TABLE VII. Cubic centimetres of oxygen liberated in SUBSTANCES. 1 min. 2 min. SCH Water tie en eae te Distr Seinen) 15.0 30.0 . water + 5 c.c. #5 sodium acetate .... 26.0 47.0 . = sodium chloride-+5cc. water... . - 5.0 9.0 . 4 sodium chloride + 5 c.c. 35 sodium acetate . 125 225m 2OLWaten . ek enna tes 5 ys Sar 2 15.0 30.0 . water + 2c.c. sjj sodium acetate. . .. . 19.0 37.0 . sodium chloride ++ 2c.c. water. ... . 3.0 6.0 . % sodium chloride + 2 c.c. sG> sodium acetate . 6.0 - water . water + 1 c.c. =45 sodium acetate * sodium chloride + 1 c.c. water . * sodium chloride + 1 c.c. =$5 sodium acetate . minute in the mixture. The same results will be noticed at the end of the first minute, and also where the smaller amounts of mercuric chloride are used. An objection to this might be made, — namely, that the chlorine ions of the mercuric chloride lessened the dissocia- tion of the calcium chloride. This, however, is not thecase. Calcium chloride is a strongly dissociated salt, while the mercuric chloride isa weakly dissociated salt. In case the calcium chloride is lessened in degree of dissociation a small amount, the results would not be Further Proof of lon Action in Physiologie Processes. 385 vitiated, as this would lessen the depressing influence of the calcium ion, and would also furnish proof of the point in hand. To show the effect of lessening the number of the anions or nega- tive ions, a similar method was used. For the solutions to be tested we used sodium and calcium acetates, and added in the one case sodium chloride and in the other calcium chloride. The addition of the common ion, sodium, in the one case and calcium in the other, will lessen the dissociation of the sodium or calcium acetate, and thereby will decrease the number of acetate ions, correspondingly the stimulation should be less. The results are seen in Tables VI and VII. By subtracting the amount of oxygen given off in water at the end of the second minute, from that given off in the water and cal- cium acetate, the stimulating action of the acetate will be found. This is 16 c.c. of oxygen. By subtracting the amount of oxygen given off at the end of the second minute, in the water and calcium chloride, from that given off in the calcium chloride and calcium acetate, the stimulating action of the acetate in the presence of calcium chloride will be found. This is 12 c.c. of oxygen. The stimulation of the calcium acetate is seen to be 4 c.c. of oxygen less, when in presence of the calcium chloride, than where the calcium acetate was used in water alone. Similar results are obtained in the other concentrations, but it is especially striking where 1 c.c. of the calcium acetate was used. In this amount the calcium acetate in presence of the calcium chloride has no stimulating effect, while in presence of water it stimulates 3 c.c. of oxygen. Sodium acetate is affected by sodium chloride the same as the calcium acetate is by the calcium chloride. The explanation for this is the same as before given, — namely, the strongly dissociated sodium chloride or calcium chloride, when added to the less strongly dissociated sodium acetate or calcium acetate decreases the dissociation of the latter and thereby decreases the number of stimulating negative ions. The same results were obtained with platinum black as were obtained with kidney extract. CONCLUSIONS. 1. The non-electrolytes have no effect except in solutions of 1 mol concentration, or more, upon the decomposition of hydrogen dioxide by platinum black, or by a watery extract of kidney; in concentra- tions stronger than I mol, there is an inhibitory effect which increases with the increasing concentration. 386 C. Hugh Newlson and Orville H. Brown. 2. A salt in dilute concentration exerts either a depressing or stimulating effect upon the decomposition of hydrogen peroxide by platinum black, or a watery extract of kidney, by virtue of its ionic condition. 3. The stimulating effect of a salt upon the splitting of peroxide of hydrogen by platinum black, or a watery extract of kidney, depends upon the negative ion, while the retarding effect depends upon the positive ion. Our thanks are due Professor Lyon for valuable criticisms and suggestions. THE PASSAGE OF DIFFERENT FOOD-STUFFS FROM THE STOMACH AND THROUGH THE SMALL INTESTINE. By W. B. CANNON. [From the Laboratory of Physiology in the Harvard Medical School.| CONTENTS. @hemethod .-. . . An RC lomRcA ICC: ola The treatment of the ifort foal pais i ae stomneh : The rate of gastric peristalsis : The rate of discharge from the stomach Rate of discharge of fats : Rate of discharge of carbohydrates . Rate of discharge of proteids . . . . i 0 : Comparison of the rapidity with which cavbohydrates and naoreids are dis. charged from the stomach The treatment of combinations of food-stuffs by the Seomaehy, Effect of feeding carbohydrate first, proteid second Effect of feeding proteid first, carbohydrate second Effect of mixing in various combinations equal amounts of the elites feodistutis Effect of mixing carbohydrate and proteid Effect of mixing fat and proteid . Effect of mixing fat and carbohydrate . Effect of increasing the amount of food ; The passage of the different food-stuffs through the small i intestine Rhythmic segmentation with the different food-stutfs . Rate of segmentation with the different food-stuffs . ° Rate of passage of the different food-stuffs through the small intestine Summary 407 407 409 410 411 413 413 416 : 1901, during the course of observations on the movements of the intestines, I noted that salmon began to leave the stomach later than bread and milk, and that it was slower in reaching the large in- testine ; in the report of the research I called attention to this inter- esting difference.? It seemed that a careful study of the manner in 1 A partial report of this research was presented at the meeting of the American Physiological Society in December, 1903, and was published in the Proceedings, This journal, 1904, x, p. xvii. 2 Cannon: This journal, igo2, vi, p. 263. 387 388 W. B. Cannon. which the different food-stuffs are mechanically treated by the ali- mentary canal might suggest the mechanisms by which the move- ments are controlled. The purpose of the following investigation, therefore, was primarily to study the mechanical treatment of the various food-stuffs in the stomach and small intestine. This study was to include the rate of gastric peristalsis and of rhythmic segmentation in the small intestine, and such important matters as the time-interval between eating and the first passage of food from the stomach, the rate at which the stomach empties, and the time required for traversing the small intestine. It was also desirable to know how these processes might be affected by altering the conditions, as, for example, by mixing different kinds of food-stuffs, and by changing their amount. In order to serve for a study of this nature the method should be as simple and exact as possible, and it should not interfere with the process of digestion or with the course of the food through the digestive canal. THE METHOD. Among the first essentials for simplicity of method in a study of the agencies controlling the mechanical activities of the stomach and intestines is the employment of food-stuffs as purely proteid, fat, or carbohydrate as possible. Such foods were used in this investiga- tion. Boiled beef free from fat, boiled haddock, and the white meat of fowl are examples of proteids which were fed; beef suet, mutton and pork fat are representatives of the fats; starch paste, boiled rice, and boiled potatoes, of the carbohydrates. The foods were invariably given in uniform amount, 25 c.c. They were always finely shredded or pressed in a mortar, and were moistened with sufficient water to produce, as nearly as could be judged by the eye and by manipulation, the uniform consistency of thick mush. Before the food was fed to the animals, it was mixed with 5 grams of subnitrate of bismuth. A comparison of subnitrate of bismuth with other insoluble heavy salts has shown that it has no peculiar effects on the movements of the alimentary canal; and also clinical studies by Schiile' have proved that the addition of subnitrate of bismuth to food does not interfere with normal gastric motility. 1 SCHULE: Zeitschrift fiir klinische Medicin, 1896, xxix, p. 67. The Passage of Food-Stuffs from the Stomach. 389 In all cases full-grown cats, deprived of food for the twenty-four or thirty hours previous to the experiment, served for the observations. The animals were either permitted to eat voluntarily from a dish, or were placed on the holder and were fed from a spoon, usually with little or no difficulty. The animals thus fed with food mixed with bismuth subnitrate were exposed to the X-rays and, without disturb- ing the processes of digestion, the’ movements of the food in the stomach and small intestine were observed by means of the shadows cast on a fluorescent screen. That the results secured were not due to individual peculiarities of the cats was proved by using the same cat repeatedly with different food-stuffs, and finding the results char- acteristic of the food, and not peculiar to the cat. Animals once used were not used again within three days. A B co FIGuRE 1. — Tracings of the shadows of the contents of the stomach and intestine, made two hours after feeding boiled lean beef (4) and boiled rice (2). The aggregate length of the shadows of the intestinal contents in A was 20 cms.,in B,43 cms. The small divisions of the food in some of the loops represent the process of rhythmic segmentation. One-third original size. The observations were recorded at regular intervals, — one-half hour, one hour, and then every hour for seven hours after the feeding. The records consisted of outlines of the shadows traced on transparent tissue paper laid over the fluorescent surface. If in any case there was doubt that all the shadows had been recorded, an electric light was flashed momentarily on the tracing before it was removed from the screen, and thus the outlines drawn on the paper were compared with the shadows cast by the intestinal contents, and the outlines verified. Inasmuch as the diameter of the small intestine varies only slightly (see Fig. 1), the area of cross-section of the contents may be dis- regarded, and the aggregate /ength of the shadows of the contents 390 W. B. Cannon. of the small intestine may be taken to indicate the amount of food present. By comparing the aggregate lengths of these shadows it is possible to determine the relative amounts of food in the intestine at different times after feeding, as well as the relative amounts of different foods in the intestine in different cases at the same time after feeding. For example, in the original tracings, represented in Fig. 1, the amount of proteid in the intestine two hours after feeding, indicated by the aggregate length of the masses, was 20 centimetres ; and the amount of carbohydrate, similarly indicated, was 43 centi- metres. By this method the observer, without interrupting or inter- fering with the course of digestion, can know when food first leaves the stomach, the rate at which different food-stuffs are discharged into the intestine, the interval till food first appears in the large intes- tine, and the mechanical treatment which the food receives as it passes through the canal. The method has obvious defects. The loops of intestine are not always parallel to the screen, and the loops not parallel may not always make the same angles with the screen ; the shadows cast by contents of the loops would therefore be variously foreshortened. In extenuation of this defect, it may be said that the animals were stretched on their backs, and that the ventral abdominal wall was, flattened against the back, both by the stretching and by the pressure of the fluorescent screen; the loops must therefore have been nearly parallel to the screen, except at possible short dorso-ventral turns from one loop to another. That the foreshortening of the shadows in the loops and turns was not a serious source of error was repeatedly proved by tracings made before and after a rearrangement of the loops by massage of the abdomen; the tracings showed only slight variations in the aggregate length of the shadows. Another possible defect of the method arises from the chance of such overlapping of the loops that two masses of food or parts of two masses may cast a single shadow. Care was invariably taken to obviate this error by pressing apart with the fingers loops lying close together. A further — criticism of the method suggests itself, — that the bismuth subnitrate and the food may separate, and that the shadows may then be mis- leading. Several animals were fed the three different kinds of food, and were killed at intervals from two to six hours after the feeding. The intestinal mucosa was remarkably free from any perceptible sepa- rate deposits of the heavy powder, and the well-limited masses of material scattered at intervals along the gut were mixtures of sub- The Passage of Food-Stuffs from the Stomach. 391 nitrate of bismuth and the food. Naturally, with the digestion of a part of the food and with the constant interchange of fluids between the intestinal mucosa and the food-remnant in its movement onward, the relation of the bismuth subnitrate to the remnant must vary; but examination proved that the remnant does not become fluid to a degree which prevents it from being a vehicle of transmission for the bismuth salt, nor, on the other hand, does the percentage of bismuth fall until it no longer indicates the presence of alimentary material. It is clear that the changes in the relation of the bismuth subnitrate to the food, on account of absorption of the food or secretion of digestive fluids, are much less in the early stages of intestinal digestion, when no absorption and but little digestive alteration have taken place, than they are later. The application of the method to the determina- tion of the rate of discharge through the pylorus is therefore justified only in the first two or three hours of digestion before much absorp- tion has occurred. Other minor faults of the method are to be found in the variations in the thickness of the food-masses at different times, and in the individual rates of absorption for the different foods. These defects, however, must be regarded, especially in the early stages of intestinal digestion, as relatively slight, compared with the great and characteristic differences in the amounts of food present in the intestine when carbohydrates, fats, and proteids are separately fed. In securing data in this research about 1200 observations were made on 150 cases. A few of these cases have not been reported because they were evidently pathological, — the animals were afflicted with coryza and conjunctivitis, and almost no food left the stomach for at least seven hours; other cases have not been reported, because they were preliminary and did not afford data for the regular times selected for the observations. Furthermore, the desirability of a uni- form number of cases, for comparing different conditions, required, at the time the results were collated, either the exclusion of the cases in excess of the smallest number for any condition, or many more — observations to bring the smaller up to the larger numbers. The degree of uniformity of the results justified the former course, and many cases were thus excluded. The 728 observations on 91 cases reported in the following sections are thoroughly representative. 392 W. B. Cannon. THE TREATMENT OF THE DIFFERENT Foop-STUFFS BY THE STOMACH. The rate of gastric peristalsis. — In observations made on the stomach in 18971!I noted that the peristalsis of the stomach was to be seen whenever an animal was examined during gastric digestion, and I inferred that peristaltic waves are running continuously throughout the entire digestive period. All observations made since have sup- ported that inference, — so long as food remains in the stomach the waves, ever-recurring, sweep slowly, one after another, from the middle of the stomach to the pylorus. The rate of peristalsis noted when bread and milk were fed was stated as 6 waves per minute; a new contraction would thus appear every ten seconds. A consider- ably slower rate of peristalsis (4 waves per minute), noticed after feeding a fat, suggested that there might be characteristic rates for the different food-stuffs. Observations at different intervals after feeding were therefore made on various animals which had been fed various kinds of foods, and the following results secured: Average rate per| Rate most fre- | Extreme varia- minute. quently observed.| tions in rate. Waves waves Waves Fats (23 observations). . .. 5.0 ee 4.0-6.0 Proteids (16 observations) . . 5.2 5.0-5.4 4.8-5.8 Carbohydrates (13 observations) 55 5.8 5.0-6.0 The average rate of peristalsis with the separate food-stuffs differs concomitantly with the rate most frequently observed, but the differ- ence is relatively so slight, and the variation with any given food so great, as to make it improbable that each food-stuff has a character- istic rate. The variation during a period of digestion suggested in some cases a slowing of peristalsis as time passed. The following examples, however, showing the number of waves per minute at different times after feeding different foods, prove that no general statement can safely be ventured. 1 CANNON: This journal, 1898, 1, p. 367. The Passage of Food-Stuffs from the Stomach. && Ne) Go Hours after feeding . Bacon fat Boiled haddock Boiled rice . It will be shown later that, when given in fairly pure forms, carbohydrates leave the stomach first, proteids next, and fats slowest of all. The figures presented in this section show that the average rates of peristalsis follow this same order, —carbohydrates most rapid, proteids next, and fats slowest. The differences, however, are so slight that the slower rate does not compensate for the greater duration of stay in the stomach. When equal amounts of carbo- hydrate, proteid, or fat food are given, a much larger number of peri- staltic waves, therefore a much greater expenditure of energy in muscular contraction, is required by the proteids and fats than by the carbohydrates, before the stomach is emptied. The rate of discharge from the stomach.-— Tables of the “ digesti- bility”: of foods, as indicated by the time during which the various foods remain in the stomach, have long been published. Beaumont! made such a table from his observations on Alexis St. Martin, and more recently Leube,? and Penzoldt? and his pupils have studied, by means of the stomach tube, the duration of the gastric digestion of various foods, and tabulated their findings. The use of these figures to judge the rate at which the stomach is emptied is open to criticism. First, the observations were made either on a pathological subject or on one whose digestion was interrupted by the introduction of a stomach tube. The results, furthermore, express merely the time when the stomach was found empty; they give no hint as to the moment when food first passed the pylorus, or as to the amounts,. large or small, which entered the intestine at any stage during digestion. Also, if comparisons are to be made, it is obviously nec- essary to note the amount of food given. Beaumont’s records indi- cate frequent inattention to this factor, and Leube’s observations 1 BEAUMONT: The Physiology of Digestion, second edition, Burlington, 1847, p- 292. : 2 LEUBE: Zeitschrift fir klinische Medicin, 1883, vi, p. 189. 8 PENZOLDT: Deutsches Archiv fiir klinische Medicin, 1893, li, p. 545. 394 W. B. Cannon. _have the same defect. Although Penzoldt recorded the amounts of food given, he did not give systematically the same amounts, so that the stomach was not always dealing with the same mechanical problem. And these investigators did not consider in their com- parisons the consistency of the food, a factor which Moritz? has rightly noted as important. Moreover, the endeavor of these investi- gators was to learn the manner in which ordinary articles of diet were treated in the stomach, — the simplification of the conditions by the use of fairly pure food-stuffs they did not much regard. The research here reported was an attempt to learn how different food-stuffs, other factors being as nearly as possible the same, would be acted on mechanically by the stomach and small intestine. It was the intention also to discover, if possible, the agents determin- ing the difference in treatment. Fairly pure food-stuffs, therefore, were administered; they were given in uniform amount; they were, as nearly as could be judged by the eye and by manipulation, uniform cm. 10 a eral | Pa pee aie Velehs|. |. bel ea Hours; 1 2 3 4 5 6 7 s FicurE 2, — Curve showing average aggregate length in centimetres of the masses of fat food present in the small intestine at regular intervals, for seven hours after feeding various fat foods. Sixteen cases. in consistency. It is to be remembered that the results here stated are significant for the emptying of the stomach chiefly in the early stages of intestinal digestion, before much absorption has taken. place. 1. Rate of discharge of fats. — In selecting the fats for this investi- gation, it was necessary to pay particular attention to the differences in the consistency of the fat at different temperatures. A fat of proper consistency at room-temperature might be much too fluid at body-temperature. Care was taken to choose fats or fatty tissues which, when mixed with subnitrate of bismuth, presented at body- temperature about the same degree of viscosity as the carbohydrate and proteid preparations. The rate at which fats leave the stomach may be seen in the following average figures, representing in centimetres the aggregate 1 Moritz: Zeitschrift fiir Biologie, 1901, xlii, p. 565. The Passage of Food-Stuffs from the Stomach. 395 length of the food-masses in the small intestine, at regular times after the various fats were fed : Hours after feeding . Porkefat, 4 cases... « << 6. ; : i 16357 S27 Mutton tallow, 4 cases. . ! : : ES al osk 9.4 9.0 Beef suet,4 cases ... 4 4 d : 10.7 6.6 5.4 Mutton fat,4 cases . . . 3 : : , 22.2 | 19:0 | 16.0 Average for fats, 16 cases. ; : 3 : LA el2-2) LIE The amount of variation observed in different animals when the same food (mutton tallow) was given, is illustrated in the following figures : Hours after feeding . Aggregate length (in centi- metres) of food- ie aa in small intestine . | l Averages These tables and the accompanying curve of the fat-content of the small intestine (Fig. 2), plotted from the average figures of the six- teen cases of fat-feeding above reported, show that the emergence of fat from the stomach begins rather slowly, —in eight of the sixteen cases, indeed, it did not occur at all during the first half-hour of di- gestion, — and continues at such a slow rate that there is never any great accumulation of fat in the small intestine. Fats almost inva- riably are present in the stomach during the seven hours of observa- tion; in one case an animal was killed six hours after receiving 25 c.c. of mutton fat, and about 11 c.c. of the food had not yet entered the intestine. The long, low curve, therefore, is the char- 1 These fats were not cooked. 396 W. B. Cannon. acteristic curve for fats. It indicates a discharge from the stomach at nearly the same rate at which the fat leaves the small intestine by absorption and by passage into the large intestine. Zawilski,! in 1876, while studying the duration of the fat-stream through the thoracic duct, was impressed by the length of time necessary to complete the absorption of fat. His examination of three animals killed at different intervals after feeding fat mixed with other food, showed that about 100 of the 150 grams of fat were in the stomach after five hours of digestion, and even after twenty-one hours about Io grams were still there. In the small intestine, on the other hand, the variation in amount was only slight: about 10 grams were found at five hours, and about 6 grams at twenty-one hours. Frank,? also, in investigating the absorption of fatty acids, noted in six animals that the fat stayed long in the stomach, and that in the small intestine a fairly uniform amount of fat was present at various times. Matthes and Marquadsen? have confirmed Zawil- ski and Frank, but their observations, like the previous observations, were incidental to another investigation. The evidence that fats are retained long in the stomach is meagre, but the testimony of different observers is thus far harmonious. Recently, however, Strauss* has denied that fats remain exceptionally long in the human stomach. But Strauss’s methods are hardly comparable with the methods used by previous investigators. One hundred grams of butter and 1 litre of milk containing 9 per cent fat sufficed for the fat of the diet, but this was given with almost 600 grams of other food and 1700 c.c. of other drink. The observations were few, and on only one patient. It is clear that the results tabulated above agree with and amplify the less complete evidence offered by Zawilski, Frank, and Matthes and Marquadsen. The long delay of fat in its passage through the alimentary canal is in the stomach. Fat passes from the stomach about as rapidly as the small intestine disposes of it; asa rule, therefore, the amount of fat in the small intestine is fairly con- stant in quantity and relatively slight in amount. 1 ZAWILSKI: Arbeiten aus dem physiologischen Anstalt zu Leipzig, 1876, p. 156. 2 FRANK: Archiv fiir Physiologie, 1892, p. sor. 3 MATTHES and MARQUADSEN: Verhandlungen des Congresses fiir innere Med- icin, 1898, xvi, p. 364. 4 STRAusS: Zeitschrift fiir diatetische und physikalische Therapie, 1899, iii, Pp: 279. ‘ The Passage of Food-Stuffs from the Stomach. 397 2. Rate of discharge of carbohydrates. — The following average figures, representing in centimetres the aggregate length of the food- masses in the small intestine at regular intervals after feeding, will indicate, particularly in the early stages of digestion, the rate at which various carbohydrate foods pass from the stomach.1 Hours after feeding. . . .| § Mashed potato,4 cases . .| 9.4 | 30.9 | 43.0 | 25.2 | 21.2 | 13.7 92 Crackers and water, 4 cases .| 11.0 | 22.0 | 35.4 | 39.5 | 40.0 | 27.5 | 22.1 Boiled rice,4 cases . . . .| 166 | 29.0 | 36.2 | 286 | 24.4 | 17.0 | 14.0 Starch pudding, 4 cases . .| 12.6 | 24.7 | 36.9 | 37.4 | 25.2 | 17.6 | 12.9 Av. for carbohydrates, l6cases| 12.4 | 26.6 | 37.7 | B20 1) Ald) 19:0.) 140 The following figures illustrate the degree of variation observed when the same food (mashed potato”) was fed to different animals: Hours after feeding . Aggregate length (in centi- | metres) of food-masses in small intestine . Average . In the accompanying curve, Fig. 3, plotted from the average figures for carbohydrates, the content of the small intestine at regular intervals after feeding is represented graphically. In my first obser- vations on the movements of the stomach, bread was used to feed the animals, and I noted that this food appeared in the duodenum 1 Glucose also was tried, but proved by itself to be such a violent stimulus to the motor activities of the alimentary canal that its use was not continued. 2 MARBAIX (La Cellule, 1898, xiv, p. 299) notes that potatoes leave the human stomach rapidly, and that the gastric juice cannot attack them to any extent, and he suggests that an important question lies here. 398 W. B. Cannon. within ten or fifteen minutes after feeding.! The tables and the curve for carbohydrates show that this early emergence of the starchy food from the stomach is followed by an abundant discharge. Ina half-hour the amount present has almost equalled the maximum for fats, and at the end of an hour that amount has more than doubled. The abrupt, high rise of the curve to a maximum at the end of two hours indicates the rapidity of the rate of discharge. And as the stomach is usually empty about three hours after feeding carbo- cm. ae oe | 1h | eae alata tage |) BE ea Ei ee te) NI) Sa i alee elle | Se es Aa loa | sa 30 20 Hours } 1 FIGURE 3.— Curve showing average aggregate length in centimetres of the masses of carbohydrate food present in the small intestine at regular intervals for seven hours after feeding various carbohydrate foods. Sixteen cases. hydrates, the slow fall in the curve during the last four hours records the gradual departure of the food from the small intestine through the absorbing wall and into the colon. Carbohydrates, therefore, begin to be discharged early from the stomach (ten or fifteen minutes after the food is swallowed), and the discharge is so rapid and abundant that the stomach is emptied in about three hours. The small intestine consequently receives a large amount in a relatively short time, and must be the resting- place for much of this food for a considerable period, for the passage of carbohydrates from the small intestine is usually slow and gradual. 3. Rate of discharge of proteids. — The rate at which various pro- teid foods pass from the stomach is indicated in the following average figures, which represent in centimetres the aggregate length of the food-masses in the small intestine at regular times after feeding: 1 CANNON: This journal, 1898, i, p. 369. The Passage of Food-Stuffs from the Stomach. 399 Hours after feeding . Hiawle4.cases. . . 3. E : : 20.6 18.0 Lean beef, boiled, 4 cases . : , 24.5 16.0 iibrins 41cases*®, a... i : 25.2 17.8 Haddock, boiled, 4 cases . i 4 129 DS Be ded Vive Av. for proteids, 16 cases . : : : 20:6), 1930 Teas | 122 A series of observations was made also with egg-albumin. The unchanged albumin, 25 c.c., mixed with 5 grams of subnitrate of bismuth and 3 grams of lycopodium spores (to offer a vehicle for the bismuth if the albumin should be absorbed) was first fed. The food passed from the stomach at the carbohydrate, not at the proteid rate. It seemed possible that the great rapidity of exit from the stomach might be due to the fluid state of the food, but when the same mix- ture of albumin, spores, and bismuth subnitrate was heated toa thick, jelly-like consistency, the rate of discharge was not diminished.t In the following table the first two cases were unchanged egg- albumin ; the second two, egg-albumin coagulated by heat: Hours after feeding . ( Aggregate length (in centi- | metres) of food-masses 4 in small intestine Average . Comparison of these average figures with the averages of other proteids and of the carbohydrates, proves at once that egg-albumin is 1 There is a marked discrepancy between these results and the statement made by Roux and BALTHAZARD (Archives de physiologie, 1898, p. 91) that in the dog coagulated egg-albumin did not begin to leave the stomach for about three hours after feeding. Never have I seen so long a delay except in animals that were manifestly not in good health. 400 W. B. Cannon. treated mechanically by the stomach much more like a carbohydrate than like other proteids. Gluten paste was also tried, and the averages of three observations were as follows: Hours after feeding... = = 4 1 2 2 4 5 6 v 27.2 49.2 47.0 39.7 35.2 21.0) (22530 Here again the figures are like those from carbohydrates, not like those from proteids; but in this case the “gluten” was found to consist very largely of starch. No such explanation is at hand for the difference between the usual proteid averages and the averages for albumins, and any explanation of the differences between the mechanical treatment of proteids and of carbohydrates must take this exceptional treatment of albumin into consideration. As an indication of the variation observed when the same food (boiled lean beef) is fed to different animals, the following figures are presented: Hours after feeding . 9.5 | 14.0 | 18.0 Aggregate length (in centi- : : 14.0 | 20.0 | 23.5 metres) of food-masses 4 in small intestine. .. h ‘ 11.0 | 19.0 | 19.0 18.5 | 22:0 | 22.0 PAW ETARE 5 1c 5\)-. 1 ee SNe : i 1331 | LE eZue Fig. 4 is a curve plotted from the average figures for the content of the small intestine after feeding the four representative proteids in the sixteen cases reported in the above table. Comparison of these average figures with the figures for the different proteids reveals a remarkable similarity during the first two hours of digestion. The striking feature of this part of the proteid curve (Fig. 4) is its very slow rise. In nine of the sixteen cases no food had left the stomach at the end of the first half-hour, and in eight cases the small intestine had not received at the end of an hour more than 4 centimetres of food. Some of the food usually remains in the stomach for about six hours; the fall in the curve, therefore, represents, as in the case of the fats, the resultant of the outflow from the stomach and of the departure ° The Passage of Food-Stuffs from the Stomach. 401 of the food from the small intestine, either by absorption or by moving onward into the colon. The difference between the initial discharge of carbohydrates and of proteids makes a further investigation of these relations desirable. Comparison of the rapidity with which carbohydrates and proteids are discharged from the stomach.— The main portion of adiet is more likely to be composed of carbohydrates or proteids or of the two com- bined than of fats alone. To digest a diet consisting chiefly or even largely of fat is an unusual task for the digestive apparatus. The mechanical treatment of carbohydrates and proteids is, therefore, of more importance practically than the treatment of the fats; and the fact that the stomach is more habituated to the presence of carbo- Aa ea eae Dee he ae ae Se a 7 Oe a PSA see ee BA Hee ee ee Hours 4 1 cm. 20 10 FIGURE 4.— Curve showing average aggregate 8 in centimetres of the masses of proteid food present in the small intestine at regular intervals for seven hours after feeding various proteid foods. Sixteen cases. hydrates and proteids in large amounts makes a consideration of the differences in the manner in which these foods are treated of greater significance than a comparison involving the fats. The curves indicating the carbohydrate and proteid rapidity of discharge from the stomach are strikingly different. At the end ofa half-hour eight times as much carbohydrate as proteid has left the stomach; at the end of an hour more than five times as much, and even at the end of two hours, after much carbohydrate food has in all probability been absorbed, considerably more than twice as much carbohydrate as proteid is present in the small intestine (see Fig. 1).1 This remarkable difference between the carbohydrate and the proteid rapidity of departure from the stomach assumes special significance when the action of the gastric juice on these two food- stuffs is considered. That the carbohydrates, which are not digested 1 It should be remembered that these comparisons are comparisons of average figures; commonly at the end of a half-hour, and sometimes at the end of an hour the stomach has not released any of its proteid content. Under normal conditions carbohydrates are never thus retained. 402 W. B. Cannon. by the gastric juice, should begin to leave the stomach soon after being swallowed, and should pass out rapidly into a region where they are digested, while the proteids, which are digested by the gastric juice, should be retained in the stomach for a half-hour, and frequently for an hour, without being discharged in any considerable amount, indicates the presence of an important digestive mechanism. With the purpose of securing further evidence as to the action, of this probable mechanism, various combinations of food-stuffs were fed, and the rate of passage from the stomach studied by the method already described. THE TREATMENT OF COMBINATIONS OF FOooD-STUFFS BY THE STOMACH. ; Effect of feeding carbohydrate first, proteid second. — When different kinds of foods are fed one after another, the first food swallowed fills the antrum pylori and lies along the greater curvature of the stomach, and the later food lies along the lesser curvature and fills the cardiac end. Thus if carbohydrates are fed first and proteids second, the carbohydrates are in contact with the pylorus and predominate in the pyloric end of the stomach, while the proteids are found chiefly in the fundus. Does the presence of proteids in the fundus of the stomach retard the exit of carbohydrates lying near the pylorus? To answer this question, about 12.5 c.c. of crackers and water, mixed with 2.5 grams of subnitrate of bismuth, were fed, and then about 12.5 c.c. of boiled lean beef with 2.5 grams of subnitrate of bismuth. The following results were obtained in four such cases: Hours after feeding . | Aggregate length (in centi- metres) of food-masses 4 in small intestine. . . | | Average . 1 CANNON: This journal, 1898, i, p. 378; also GRUTZNER: Archives italiennes de biologie, 1901, xxxvi, p. 29. The Passage of Food-Stuffs from the Stomach. 403 Comparison of these figures with the average figures for crackers and water (see p. 397), and for boiled lean beef (see p. 399), proves that the presence of proteids in the cardiac end of the stomach and along the lesser curvature does not materially check the departure of the carbohydrate food lying at the pylorus. Effect of feeding proteid first, carbohydrate second, — The converse of the relations described in the foregoing paragraph is observed whea proteid is fed first, and carbohydrate second. The proteid is then at the pylorus, and the carbohydrate rests chiefly in the fundus of the stomach. Under these circumstances does the presence of the proteid near the pylorus retard the natural early exit of the carbo- hydrate? To answer this question about 12.5 c.c. of boiled and shredded lean beef with 2.5 grams of bismuth subnitrate were fed, and this was followed by the same amount of moistened crackers with 2.5 grams of bismuth subnitrate. The following results were obtained in four such cases: Hours after feeding . ( Aggregate length (in centi- | metres) of food-masses { in small intestine. Average . The nearness of these figures to the average figures for boiled lean beef during the first four hours should be noted. And the rate of discharge when carbohydrates are fed first should be compared with the rate when proteids are fed first (Fig. 5). When the crackers are first at the pylorus, the discharge for two hours is almost as rapid as when crackers alone are given. At the end of two hours, how- ever, the curve ceases to follow the normal curve for crackers, — there is a checking of the outgo from the stomach, which is most reasonably explained by assuming that the beef has by that time come to the pylorus in considerable amount, and is as usual passing out slowly. On the other hand, when the beef is first at the pylorus, the curve is in close approximation to the normal curve for beef dur- ing the first four hours, and after that time, as the crackers come to Cannon. 404 Bs the pylorus in greater amount, the curve continues to rise, while the curve for beef falls. At no time during the first three hours is there half as much food in the small intestine as when crackers alone are fed. It is evident that the presence of proteid near the pylorus dis- tinctly retards the onward passage of carbohydrate food ane in the cardiac end of the stomach. It is noteworthy that when proteid was given first the stomach still contained considerable food even six hours after feeding, — a doubling of the time during which carbohydrates alone remain in the stomach. On the other hand, when carbohydrates were fed first, the food was last observed in the stomach at the end of four hours. When the pro- teid was fed second, therefore, the food was retained in the stomach only about an hour longer than the carbohydrates alone. When one considers the ease with which carbohydrate fermentation may proceed in the cardiac end of the stomach, unchecked for a long time by any acid Hours} 1 2 3 Es FIGURE 5.— Curve showing average aggregate length of the food- masses in the small intestine at regular intervals for four hours after feeding moistened crackers first, and lean beef second, four cases (heavy continuous line); and after feeding lean beef first, and crackers second, four cases (heavy interrupted line). The light continuous line is the curve for moistened crackers alone; the light interrupted line, the curve for lean beef alone. Four cases are represented in each - curve. fluid,! the desirability of avoiding such a stasis of the carbohydrate food in the fundus as occurs when proteids lie near the pylorus, becomes at once apparent. Two rational courses seem to be open: either the carbohydrate food should be eaten first, and the proteid second, if the two are taken separately at a meal; or, in case the proteid is eaten first, the carbohydrate should be very carefully chewed. The chewing mixes saliva with the starchy food, and the food is thereby changed in the cardiac end of the stomach either into sugar, which as a fluid flows downward toward the pylorus, or into dextrin, which undergoes fermentation with difficulty or not at all.? Effects of mixing in various combinations equal amounts of the three food-stuffs. — Inasmuch as food is generally given as a mixture of the 1 CANNON: This journal, 1898, i, p. 379. 2 CANNON and Day: This journal, 1903, ix, pp. 407-413. The Passage of Food-Stuffs from the Stomach. 405 various food-stuffs, it is important to learn what effect the combining of the various food-stuffs from which characteristic curves have been secured may have upon those curves. For this purpose carbohydrates, fats, and proteids were mixed in equal parts to make 25 c.c. of food, and this mixture, with 5 grams of subnitrate of bismuth, was fed, and the results recorded in the manner already described. LEAN BEEF AND MOISTENED CRACKERS, IN EQUAL PARTS. Hours after feeding. . . 5 9.0 7.0 Aggregate length (in centi- | 18°55) 16:5 metres) of food-masses 4 in small intestine. . . | . 16.0 9.5 l 11.5 | 10.0 iAveragels cs 1 ‘ ; : : W372) LOL BorLED HADDOCK AND MASHED POTATO, IN EQUAL PARTS. 1170) 920.0), 36:0) |) 38:0) | 34-0) 17-0 Aggregate length (in centi- 14:0) | 26.0) |) 29:0) || 29:0: | 17.0 8.0 metres) of food-masses in small intestine. . . 16:07)|)927-05 935.5.) 37.0) || 26:0 LMOMP2 5s 28:00) 29:0) | 12:0 13.0) 2LAS | S21) 33:22) 22.2 1 That there is more of mixed haddock and potato than of potato alone in the small intestine at the end of three hours is undoubtedly due to the fact that the mixed food passed into the colon one hour later than the potato. That the amount of mixed crackers and lean beef in the small intestine is less than the lean beef alone after three hours is explained by the appearance of the mixed food in the large in- testine two hours earlier than the beef appears there when fed by itself. 1. Effect of mixing carbohydrate and protetd. — In order to test the effect of mixing carbohydrate and proteid on the rate of discharge from the stomach, equal parts of lean beef and crackers in one series, and equal parts of boiled haddock and mashed potato in another 406 W. B. Cannon. series, were fed in 25 c.c. amounts. The table on the preceding page presents the results. In Fig. 6 will be found a comparison between the rate of discharge of the mixed foods, and of the same foods fed separately: in (A) are the curves for crackers (heavy line), and lean beef (light line), and for mixed crackers and beef (dotted line); in (4) are the curves for mashed potato (heavy line), boiled haddock (light line), and mixed potato and haddock (dotted line). Only the changes during the first three hours are taken for consideration, since they are most cin. . Hours} 1 2 3 4 1 2 3 FiIGuRE 6.—In (A) the average aggregate length of the food-masses in the small intes- tine during the first three hours after feeding equal parts of moistened crackers and boiled lean beef (dotted line), compared with crackers alone (heavy line), and beef alone (light line). In (&) the average aggregate length of the food-masses in the small intestine during the first three hours, when equal parts of boiled haddock and mashed potato are fed (dotted line), compared with mashed potato alone (heavy line), and haddock alone (light line). Four cases are represented in each curve. significant in judging the rapidity of discharge from the stomach. In both cases here presented, the amount of the mixed food in the small intestine at the end of a half-hour is nearer the carbohydrate than the proteid figure; indeed, in the case of mixed haddock and potato, there had left the stomach during the first half-hour a little more of the mixed food than of the potato alone. But the above tables and the accompanying curves (Fig. 6) clearly show that in general when carbohydrates and proteids are mixed in equal parts, the rate of discharge through. the pylorus is intermediate; the mixed food does not leave the stomach so slowly as the proteids, nor so rapidly as the carbohydrates. 2. Effect of mixing fat and protetd. — Boiled lean beef and beef suet served for the observations on the effect of mixing fat and proteid. These substances were mixed in equal amounts, and fed in the manner already described, with the following results: The Passage of Food-Stuffs from the Stomach. 407 Hours after feeding . f Aggregate length (in centi- | metres) of food-masses { in small intestine. ‘| ( Comparison of these results with the average figures for beef suet and for lean beef (Fig. 7) reveals at once that when the two are mixed in equal parts, the rate of discharge from the stomach is diminished to a degree below that of either the lean beef or the suet fed by itself. In other words, the presence of the fat causes proteid to leave the stomach even more slowly than the proteid by itself would leave. This con- cm +r T T T ‘ isan clusion was corroborated by feeding had- 8 | = dock and mutton fat in equal amounts, in Peps which case at the end of two hours there 1? nee: ik A Ae was in the small intestine only two-thirds as much of the mixed food as of the haddock when it was given alone. ———— Hours4 1 ez 3 In all Frcure 7.—Curves showing the probability the long delay of an hour or an hour and a half before the initial pas- sage of food from the stomach, noted in an earlier research when salmon was fed, was due to the presence in salmon of more than half as much fat as proteid. 3. Effect of mixing fat and carbo- hydrate. — Mashed potato and mutton average aggregate length of the masses of lean beef (heavy line), of beef suet (light line), and of beef and beef suet mixed in equal parts (dotted line), in the small intestine during the first four hours after feeding. Four cases are represented in each curve. fat, and moistened crackers and beef suet, mixed in each case in equal parts, were used in studying the effect of combining carbo- hydrates and fats. results of these observations. The table on the following page presents the In both series of observations the passage of the mixed food from the stomach is more rapid at first than the normal rate for the carbo- hydrate used (see Fig. 8). Very soon, however, the fats have a 408 W. B. Cannon. retarding effect on the outgo of the carbohydrate from the stomach, so that the curve for the mixed food-stuffs, after the first hour, ceases to rise, and never even approximates the height of the carbohydrate curve. The failure of the curve for mixed crackers and beef suet to rise might in part be ascribed to the remarkably early discharge of the contents of the small intestine into the colon; but the curve for MASHED POTATO AND MUTTON FAT. Hours after feeding . Aggregate length (in centi- | metres) of food-masses 4 in small intestine | | MOISTENED CRACKERS AND BEEF SUET. 18.0 | 28.5 | 15.0 9.0 65 Aggregate length (in centi- 22,0))), 30:0) || 32:05 | TSO So metres) of*food-masses in small intestine. . . 2+.0 | 27.0 | 21.0 15.0 10.5 | 18.0 10.5 ARSTAGE) .1 A Se ee eee 186 4 2425 5: 1187 mixed potato and mutton fat remains low for more than three hours without that factor becoming operative in more than one case. It is reasonable, therefore, to conclude that the mixture of fat in large amount (50 per cent) with carbohydrate has the same effect, though not to so great a degree, as the mixture of fat with proteid, — namely, the fat causes the outgo of the carbohydrate and proteid from the stomach to be slower than the outgo of either of the two foods fed separately. The Passage of Food-Stuffs from the Stomach. 409 EFFECT OF INCREASING THE AMOUNT OF FOOD. All the observations detailed thus far have been made in each instance on 25 c.c. of food. It is desirable, especially in the case of carbohydrates and of proteids, to know if the feeding of larger amounts of food will result in curves differing from those secured when smaller amounts are fed. Boiled rice, representing carbo- hydrates, and lean beef, representing proteids, were fed in twice the amount previously given. The results are seen in the following table : BoILED Rice. 50 c.c. Hours after feeding . | 19.0 | 31.0 0 | 36.0 | 28.0 | 21.0 | 13.0 | Ageregate lengths (in cen-|| 12.0 | 25.0 : 0 | 38.0 | 36.0 | 24.0 | 13.0 | timetres) of food-masses 4 in smallintestine. . .|| 32.0 | 48.0 : : 80} 65] 4.5 22.0 | 44.0 | 28.0 | 27.0 | 17.0 eee. i DE | 37.0 ; 5 |i a22 LEAN BEEF, BOILED. (| 00 | 00 | 1210 | Aggregate Jengths (in cen- || 0.0 0.0 4.0 | timetres) of food-masses 4 insmallintestine .. . | 0.0 0.0 6.0 | { 0.0 10.5 | ISSSEERS let oh) 2 | 0.0 In Fig. 9 are the curves of each of these food-stuffs in the 25 c.c. and in the 50 c.c. amounts. The rapid rate of carbohydrate dis- charge, and the slow rate of proteid discharge, have both been in- creased. At the end of an hour over 25 per cent more carbohydrate has left the stomach when the amount fed is 50 c.c. than when the amount fed is 25 c.c. And with the proteid food, when the amount was 50 c.c., nothing left the stomach for an hour, and at the end of AIO W. B. Cannon. two hours only half as much had left as when the amount was 25 c.c. Increasing the amount of food, therefore, increases the rate of discharge of carbohydrates from the stomach, and retards the out- going of the proteids. THE PASSAGE OF THE DIFFERENT FOOD-STUFFS THROUGH THE SMALL INTESTINE. The chyme emerges from the stomach in small amounts. Usually there is an accumulation of these small discharges into a slender mass in the duodenum before the food is further advanced in the intes- tine. But as time goes on, these masses tend to gather into longer Hours} 1 2 3 42: 1 2 3 4 Ficure 8.— (A) Curves showing the average aggregate length of the masses of mashed potato (heavy line), of mutton fat (light line), and of mixed potato and mutton fat (dotted line), in the small intestine during the first four hours after feeding. (4) Similar curves for crackers (heavy line), beef suet (light line), and mixed crackers and beef suet (dotted line). Four cases are represented in each curve. masses, so that at the end of three or four hours the food in the coils is often largely collected in a few unbroken strings. This condition occurs earlier with carbohydrate than with proteid food (see Fig. 1, veand 2). In the small intestine the food is mixed with the intestinal di- gestive secretions, exposed to the absorbing intestinal wall, and from time to time moved forward into new areas of intestinal activity. The mixing of the food with the secretions, and the exposing of the food for absorption, are functions performed in the rhythmic segmen- tation of the food by circular constrictions of the gut; the forward movement is the result of ordinary peristalsis; but at any moment most of the food in the intestine lies undisturbed! In studying the 1 CANNON: This journal, 1902, vi, p. 263. The Passage of Food-Stuffs from the Stomach. 411 passage of the different food-stuffs through the small intestine, there- fore, attention must be paid to the process of segmentation and to the rate of passage from the stomach to the large intestine. ; Rhythmic segmentation with the different food-stuffs. — The segmenta- tion of the food occurs with all three kinds of food-stuffs. The fre- quency with which it is seen at the regular times of observation adhered to throughout this research has varied, as shown in the following table. The figures represent the number of cases in which segmentation was observed at the hours indicated. Hours after feeding. . . 3 1 2 Carbohydrates, 16 cases. 7 al 11 iadtsyIGicases'. 3 4. 3 6 5) Proteids, 16 cases .. - 0 6 9 Comparison of these figures with the average figures for the intes- tinal content of the different foods at the regular times of observation (see pp. 395, 397, and 399) shows that the frequency with which segmentation occurs corresponds roughly to the amount of food present in the intestine. Thus, when carbohydrates are fed, seg- mentation is observed frequently, but more frequently during the first hours than later. Similar comparisons of the fats and proteids, show that, as a rule, when much food of any kind is present in the intestine, segmentation is more likely to be seen’ than when a small amount is present. The amount of work the musculature of the intestine performs with the different foods is indicated better by the total length of the masses segmented, than by the frequency of the occurrence of that process. The length of all the segmented masses in the small in- testine in the sixteen cases of each food-stuff is as follows: Hours after feeding . Carbohydrates _. .| 40.3 | 89.4 | 132.1] 76.6 | 90.3 Pforerds. J. so 0.0 | 34.1 54.0] 51.6 | 46.4 ai 5 gg oo] AOHO) || HSNO) | ZO eixoye all ra) Ane W. B. Cannon. These figures do not, of course, record the total segmentation taking place during a complete digestive period, but they indicate the amount of activity which may be expected at given intervals during the first seven hours after feeding. These figures also indi- cate that the muscular energy expended on carbohydrates in the seven hours far exceeds that expended either on proteids or fats. The longer presence of proteids and fats in the small intestine, because of the slower discharge of these foods from the stomach, would later compensate in part for the relatively small amount of segmentation during the first seven hours; but at the end of seven cn. TEER | cEZ a 80 [A if iN O a mie ie | (2 _ Le Ry Hours# 1 2 3 4 5 6 7 FiGurE 9.— (A) Curves showing the average aggregate length of the food-masses in the smail intestine when 50 cc. boiled rice are fed (heavy line), and when 25 e.c. are fed (light line). (4) Similar curves for 50 cc. boiled lean beef (heavy line), and for 25 c.c. lean beef (light line). Four cases are represented in each curve. hours the proteid content of the small intestine is almost the same as the carbohydrate content (compare Figs. 3 and 4), and so slight is the segmenting activity in the presence of fats that, should their rate continue, —and that seems hardly probable when the dwindling of the segmentation during the last hours is considered, —they must remain in the small intestine for almost nineteen hours to produce an activity equal to that produced by carbohydrates. It appears very likely, therefore, that in the small intestine a much greater amount of muscular activity in the form of rhythmic segmentation is provoked The Passage of Food-Stuffs from the Stomach. 413 by carbohydrate food than by either of the other two food-stuffs. It is noteworthy that the carbohydrates, which cause the greatest amount of muscular activity in the small intestine, cause the least amount in the stomach.! Rate of segmentation with the different food-stuffs. — In previous ob- servations on segmentation in the small intestine, the rate was found to vary from eighteen to thirty movements per minute when salmon was fed. Observations with other foods reveal no greater variation than that seen with salmon, and there was nothing characteristic in these variations. Rate of passage of the different food-stuffs through the small intestine. —In studying the passage of food through the small intestine of a woman with a fistula at the ileo-colic junction, MacFadyen, Nencki, and Sieber? noted a considerable variation in the time between the ingestion of the food and its appearance at the fistula. For example, peas first arrived at the colon on one occasion two and a quarter hours, and on another occasion five and a quarter hours, after being eaten. Demarquay,® who studied a case similar, but apparently less normal, reports also a wide variation in the time of the first appearance of food at the fistula. The method used in this research does not permit a statement of the moment when food first entered the colon; all that can be re- ported is the first observation when food was seen in the colon. Inasmuch as the observations were an hour apart, the results, except 1 Curiously the proteid (egg-albumin), which leaves the stomach with carbo- hydrate rapidity, more nearly resembles the carbohydrates than the proteids in undergoing much segmentation by the intestines. The following figures, giving the total length of the segmenting masses in five cases in which egg-albumin was fed, illustrate the abundant segmentation to which this food is subjected : Hours after feeding ..... $ 1 2 3 4 ‘) 6 Om Ios: 25.2 44 8.7 GO s7 00 It was the frequency with which segmentation takes place when egg-albumin is fed, compared with the frequency of segmentation observed in a few cases in which starch paste, and others in which olive oil was administered, that led to an erro- neous statement that proteids were more provocative of segmentation than fats or carbohydrates. See Proceedings of the American Physiological Society, This journal, 1902, viii, p. xxii. 2 MacFapyENn, NENCKI, and SIEBER: Journal of anatomy and physiology, 1891, XXV, Pp. 393. 3 DEMARQUAY: L’union médicale, 1874, xviii, p. 906. AIA W. B. Cannon. in their negative aspect, are not as exact as could be desired. The following figures, therefore, represent the number of cases for each food, in which, at the hours stated, the food was first observed in the colon: NUMBER OF CASES IN WHICH, AT VARIOUS HOURS AFTER FEEDING, FOOD WAS FIRST OBSERVED IN THE LARGE INTESTINE. Hours after feeding . Mashed potato. Moistened crackers . Rice, boiled . Carbohydrates | Starch pudding Total Fowl Haddock, boiled . Lean beef, boiled . | | | Fibrin . Total { Tallow a j Beef suet . Los) F4 | Pork fat . | Mutton fat . Total 1 In four cases no food had appeared in the colon at the end of seven hours. In this table, these cases are regarded as coming in an eight-hour class. The Passage of Food-Stuffs from the Stomach. 415 NUMBER OF CASES IN WHICH, AT VARIOUS HOURS AFTER FEEDING, FOOD WAS FIRST OBSERVED IN THE LARGE INTESTINE.—continued. Hours after feeding . { Mutton fat and potato . Beef suet, and crackers . ; | Lean beef and crackers. S Haddock and potato. Haddock and mutton fat . = e oO = =] = 5 no) = S = o ) | Lean beef and beef suet Cc Crackers, first; beef, second | Beef, first ; crackers, second Rice, 50 c.c. . Beef, 50 c.c. . The variation reported in cases of low intestinal fistula in human beings appears also in these tabulated results. Although the mean time after eating at which the food reaches the colon is about four hours for carbohydrates, about six hours for proteids, and about five hours for fats, the divergence from the mean in each of the three classes is considerable. It is interesting to note that the divergence among the carbohydrates is chiefly due to one food, the moistened crackers. The long stay of this food in the small intestine accounts for the prolonged high curve for crackers (Fig. 6, 4), so different from the quick rise and rapid fall of the curve usual with carbo- hydrates (Fig. 6, 8B). Among proteids also the divergence from the mean is most marked in the case of one food, — the boiled had- dock. This food arrives at the colon about two hours earlier than most of the other proteids, and its relatively rapid passage through the small intestine explains the exceptional early decline of the curve of the intestinal content in the case of this proteid (see Fig. 6, A and B, light lines). The two combinations, crackers and lean beef, and crackers and beef suet, may be compared with crackers alone, as to the time re- quired to traverse the small intestine; both combinations pass through the small intestine more rapidly than this carbohydrate food 416 W. B. Cannon. by itself. Even feeding crackers before feeding lean beef seems to make the crackers reach the colon earlier. On the other hand the combination of potato with haddock and with mutton fat apparently causes a checking of the rate of passage of this carbohydrate from the stomach to the colon. Hence no conclusion can be drawn from © these cases. The combination of proteids and fats, in most instances, has the effect of delaying the appearance of the mixed foods in the colon to a time later than that recorded when the fats alone were fed. The average figures for carbohydrates, proteids, and fats in the above table demonstrate that, as a rule, the carbohydrates reach the large intestine about one hour before the fats and about two hours before the proteids. As there is no considerable passage of proteid food from the stomach into the small intestine until about an hour after the carbohydrate food has begun to leave the stomach in large amount, the difference of an hour still remains between the carbo- hydrate and proteid stay in the small intestine. It is probable that, in general, the proteids pass from the stomach to the large intestine more slowly than do the carbohydrates, while the fats have a rate intermediate between these two. SUMMARY. Fat, carbohydrate, and proteid foods, uniform in amount (25 c.c.) and consistency, were mixed with a small amount of subnitrate of bismuth and fed to cats deprived of food for at least twenty-four hours. The rate of gastric peristalsis observed by means of the Rontgen rays was usually slower for fats (5.2 waves per minute) than for carbohydrates (5.8 waves per minute), but the variation was so great as to make a more definite statement unsafe. At regular intervals for seven hours after feeding, the shadows of the intestinal contents were traced on transparent paper by means of a fluorescent screen and the Roéntgen rays. Since the intestinal contents vary only slightly in diameter, the aggregate length of the shadows can be taken to indicate the relative amount of food present in the small intestine at various intervals, and in various animals at the same interval after feeding. In the early stages of intestinal digestion, before much absorption has occurred, the aggregate length of the shadows at different intervals indicates the rate of discharge from the stomach. The Passage of Food-Stuffs from the Stomach. 417 Fats remain long in the stomach. The discharge of fats begins slowly and continues at nearly the same rate at which the fat leaves the small intestine by absorption and by passage into the large in- testine. Consequently there is never any great accumulation of fat in the small intestine. Carbohydrate foods. begin to leave the stomach soon after their ingestion. They pass out rapidly, and at the end of two hours reach a maximum amount in the small intestine almost twice the maximum for proteids, and two and a half times the maximum for fats, both of which maxima are reached only at the end of four hours. The carbo- hydrates remain in the stomach only about half as long as the proteids. Proteids frequently do not leave the stomach at all during the first half-hour. After two hours they accumulate in the small intestine to a degree only slightly greater than that reached by carbohydrates an hour and a half earlier. The departure of proteids from the stomach is therefore slower at first than that of either fats or carbohydrates. An exception to this general statement was found in egg-albumin, which, both in its natural state and in coagulated form, was discharged from the stomach at about the carbohydrate speed. When carbohydrates are fed first and proteids second, the presence of proteids in the cardiac end of the stomach does not materially check the departure of the carbohydrate food lying at the pylorus; but the presence of proteids near the pylorus, when proteids are fed first and carbohydrates second, markedly retards the onward passage of the carbohydrates which under these circumstances predominate in the cardiac end of the stomach. When carbohydrates and proteids are mixed in equal parts, the mixed food does not leave the stomach so slowly as the proteids, nor so rapidly as the carbohydrates, —the discharge is intermediate in rapidity. In a mixture of fats and proteids in equal parts, the presence of the fat causes the proteid to leave the stomach even more slowly than the proteid by itself. Fat mixed with carbohydrate in equal amounts also causes the carbohydrates to pass the pylorus at a rate slower than their normal. Doubling the amount of carbohydrate food (50 c.c. instead of 25 c.c.) increases the rapidity of the carbohydrate outgo from the stomach during the first two hours ; whereas doubling the amount of proteid food strikingly delays the initial discharge of proteid from the stomach. 418 W. B. Cannon. The process of rhythmic segmentation is seen with all three kinds of food-stuffs, and the frequency of its occurrence corresponds roughly to the amount of food present in the intestine; a measurement of the length of the segmenting masses in a given number of cases shows that at the regular times of observation, during the first seven hours after feeding, the amount of segmenting activity in the presence of carbohydrates was much greater than in the presence of either fats or proteids. Egg-albumin is excepted in this general statement. The interval between the feeding and the appearance of food in the large intestine is variable, but the mean for carbohydrates is about four hours, for proteids about six hours, and for fats about five hours. After time is allowed for the later start of proteids from the stomach, there still remains a probability that the proteids pass through the small intestine more slowly than do the carbohydrates. The discussion of the observations here presented, and the relation of these observations to the work of other investigators, is deferred. toa later paper. This paper will report experiments undertaken to explain the characteristic differences of treatment of the food-stuffs, which have been described in the foregoing pages. fib TOXIC. AND. ANTI-ITOXIC ACTION OF SALTS. by AL PF MATHEWS. [From the Marine Biological Laboratory, Wood's Hole, and the Hull Physiological Laboratories, Chicago.] INCE Ringer’s! observations on the action of potassium and calcium salts on the heart, it has been known that the physi- ological action of certain salts could be modified by other salts. Ringer? discovered also that this antagonism was not confined to electrolytes, but that the action of veratrin, one of the most power- ful of protoplasmic poisons, could be neutralized very completely by potassium and calcium salts.2, This antagonism of common electro- lytes to so powerful a drug is of importance theoretically, for the light it may throw on the means of action of drugs on protoplasm, and practically, since it suggests the possibility of alterations in the toxic value of poisons by salts. To discover if other drugs could be neutralized by calcium salts, Mr. Rothrock, a medical student, examined under my direction, two years ago, the drug digitalis. He found at least a partial antag- onism between digitalis and calcium chloride, but as his results were not entirely convincing they were not published. I obtained a better result with physostigmine and these results were extended and pub- lished by S. A. Matthews and O. S. Brown.? There is, therefore, little doubt that the whole group of drugs which act like barium chloride, namely, suprarenal extract, physostigmine, veratrin (?), and digitalis may be largely or completely antagonized by calcium chlo- ride. Recently McCallum‘ has found an antagonism between cas- cara and calcium for the intestinal movements, and Hektoen has shown that the activity of some bacterial hamolysins is greatly altered by inorganic salts present in the solutions. 1 RINGER: Journal of physiology, 1884, v, p. 247. 2 RINGER: Jdid., p. 352. 3 MATTHEWS and Brown: This journal, 1904, xii, p. 173. 4 MacCaLium: Journal of experimental zodlogy, 1904, i, p. 179. 419 420 A. P. Mathews. : These facts show that the same kind of an antagonism exists’ between salts and drugs not ordinarily classed as electrolytes as between two electrolytes. This fact, in my opinion, supports the conclusion of a former paper! that both electrolytes and non-electro- lytes act physiologically by their dissociation products. The anti- toxic action of the toxins may therefore be most easily approached by a study of the anti-toxic relations of electrolytes. I have studied the antagonistic action of salts on the developing eggs of the fish Fundulus heteroclitus. While I have not yet been able to determine fully the nature of the anti-toxic action, I have decided to publish the results, for the reason that they necessitate certain changes in the conclusions drawn by Loeb from a similar study on the same form. Loeb,? extending Ringer’s observations on the action of solutions on developing ova, found that pure solutions of sodium, potassium, lithium, and ammonium chlorides were fatal to Fundulus embryos; but that the addition of small amounts of calcium, magnesium, barium, strontium, manganese, cobalt, and other salts would offset the poisonous action and permit development to take place. This fact may be easily confirmed by any one, and my own observations are in accord with it. My results, however, are not in accord with Loeb’s several explanations of the fact. In his first paper Loeb states that the anti-toxic power depends upon the valence of the cation. A solution containing only mono- valent cations was poisonous, but the addition of any salt contain- ing a bivalent cation, or the addition of a small quantity of a salt with a trivalent cation, neutralized this action. There was hence a toxic and anti-toxic action between monovalent and polyvalent cations.2 This action concerned only the cations, since the same re- sults were obtained with chlorides, nitrates, and acetates. That the toxic and anti-toxic action depended upon the valence of the cation and not upon its chemical composition was shown by the fact that most bivalent metals had the anti-toxic action which was about equally powerful in all. It will be perceived that the correctness of the con- clusions depends upon the truth of the last statement, since if it hap- pens that the anti-toxic power of bivalent metal salts is not the same, 1 MATHEWS: This journal, 1904, x, p. 290. 2 Loes: Archiv fur die gesammte Physiologie, tgot, Ixxxviili, p. 68. 3 LOEB: So 3S m 5000 aS So a oO m 3333 ~~ j=) m 2500 om ~v — m 2000 \o Oo m I3s3s mt T000 mm Z00 (oy) te) fey ey ey (Sy) te (=) should, according to Loeb’s hypothesis, be equally well neutralized. This is, however, not the case. While there is no marked difference in the ease of neutralizing potassium and sodium chlorides, lithium chloride is much harder to neutralize, and ammonium chloride re- quires at least ten times as much calcium chloride for neutralization as does sodium chloride. This may be seen in Table IV. The figures for lithium and sodium are given in Table I, and further experiments are cited for ammonium chloride later. Asam solution of ammonium chloride is not a fatal dose for some eggs, the difficulty of neutralizing by calcium is the more significant. The mono- chlorides arrange themselves in ease of neutralization in the order of their electrolytic decomposition tensions, this being highest in potassium chloride and lowest in ammonium chloride. The quantity of calcium chloride necessary to neutralize ammonium chloride is so much greater than that required to neutralize sodium chloride that little significance can be attached to the small quantity required in the former case. If the anti-toxic action is a function of valence the polyvalent ions ought to be far more efficient than the bivalent. Loeb thought that aluminium was, as a matter of fact, more efficient than other ions; on 428 A. P. Mathews. CAB Gn VE Anti- toxic salt. Anti- Toxic salt. toxic salt. Toxic salt. Concentration. Concentration. Per cent of em- bryos developed. Concentration. Concentration. Per cent of em- bryos developed. Sth the other hand, ferric chloride had no anti-toxic action! My own results with aluminium chloride do not confirm the great activity of this salt, as may be seen in Table V. A solution of 3 m# NaNO, ae 7 : 1 1 1 Wt containing FeCl, in the concentrations .. 7059) ro000 iaoa00! oun WL W1 W1 W1 71 V1 V1 E0000? 30000’ ZOH0D’ 15000" 10000’ F500" BOO Produced in no casea single embryo. The minimum anti-toxic dose of aluminium chloride is therefore about an er as contrasted with an oh of calcium chloride. It will be seen that this is a smaller difference than that between mag- nesium chloride and calcium chloride. In other words, there is a smaller difference in the efficiency of these. bivalent and trivalent 1 Cf. LILLIE: This journal, 1904, x, p. 419. The Toxic and Antt-Toxic Action of Salts. 429 TABLE V. Anti- toxic salt. Anti- toxic salt. Concentration. Per cent of em- bryos developed. Concentration. Concentration. Per cent of em- bryos developed. = S) Ss 3 col ~ = oO 1S) = ° O wilco ocleo > SSS aero cation salts than between the two bivalent salts. This certainly can hardly be interpreted as favorable to the valence hypothesis, although it may be that the relationships really existing are masked by the failure of some ions to penetrate the egg easily. But one other combination remains to be tested, and that is the toxic and anti-toxic action of monovalent for monovalent, and bi- valent for bivalent salts. That it is possible to neutralize salts by other salts of the same valence may be seen from Ringer’s observa- tions with sodium and potassium chlorides on the heart, and from my observations on motor nerves. Loeb himself states that the poisonous action of sodium chloride can be in part neutralized by potassium chloride, although heavy doses are required. Table VI gives the results. TABLE VI. Per cent of embryos developed. Concen- Anti-toxic Concen- tration. salt. tration. us) 28 10 In other words, to neutralize sodium nitrate a concentration of potassium chloride less than #% is necessary. Contrast this with the amount of magnesium chloride required to give a similar number 430 A. P. Mathews. of embryos. An ,”, magnesium chloride solution gave only 8 per cent of embryos. To get 17 per centan 7% solution must be used. In other words, it takes about twice as much potassium chloride to neutral- ize sodium chloride or nitrate as is required of magnesium chloride. This difference is immaterial when it is remembered that it takes twenty times as much magnesium chloride as calcium chloride. We find, as we did with aluminium, that there is no sharp difference between bivalent and monovalent cations in their neutralizing power, but that far greater differences exist between salts having ions of the same valence than between salts of ions of different valence. With lithium chloride I obtained no neutralization of sodium chloride ; but with hydrochloric acid the following result was obtained : AB ICE VL Anti- toxic salt. Anti- toxic salt. Concentration. Concentration. Per cent of em bryos developed. | | Concentration. Concentration. Per cent of em- bryos developed. Control alco oles wlco oko > HCl T0S000 “ Ct oe ie Ge) “ There is here an unmistakable anti-toxic effect although it is slight except in the last concentration. This is nearly the minimum fatal dose for this acid, and the embryos barely reached the form of a line on the egg before dying. With potassium hydrate, on the, other hand, no anti-toxic action was obtained. The anti-toxic action of bivalent metals upon each other was also tested with positive results (Table VIII). The experiments in Table VIII show very clearly that calcium chloride can neutralize the toxic action of magnesium chloride and of manganese. The anti-toxic power of the calcium salt for magnesium chloride is indeed greater than the anti-toxic action of strontium chloride upon sodium salts. Thus, magnesium chloride gave embryos with calcium chloride present in an ,%, concentration, while sodium chloride required an ,“, solution of strontium chloride. The Toxic and Anti-Toxic Action of Salts. 431 TABLE VIII. Anti- toxic salt. Anti- toxic salt. Concentration. Concentration. Per cent of em- bryos developed. Concentration. Concentration. Per cent of em- bryos developed. - Oo Oo o To neutralize manganese chloride, more calcium is necessary; but it will be seen that an ,” solution gives 30 per cent of embryos in an # manganese chloride solution. A quantitative difference, there- fore, between the amounts of bivalent salts necessary to neutralize univalent, and the amounts necessary to neutralize bivalent does not exist. How can this anti-toxic action upon each other of salts of which the metals have the same valence be explained upon the valence hypothesis ? TABLE IX. Concen- Embryos Concen- Embryos tration. | P88s- developed. tration. | E88S- developed. m 1000 2 19 m 500 m 33s m 2350 432 A. P. Mathews. I found it even possible to neutralize cobalt and nickel salts with calcium, and as these experiments are interesting for two reasons, I will give them in full (Tables IX and X). TABLE X. Concentration. 2 c.c. CaCl, 3m ++ + + + + + + 4+ Embryos developed. 12 10 10 None of the embryos in the first six mixtures of the calcium and cobalt were coagulated, whereas without the calcium the cobalt soon coagulates. There is no mistaking, I think, the protective action of the calcium chloride. . The following experiment with nickel chloride will serve to show the peculiar action of this metal (Table XI). Concen- tration. Eggs. TABLE XI. Embryos Concen- developed. tration. Embryos developed. This experiment shows that while nickel chloride in great dilution will kill nearly the whole of the eggs, a few eggs will resist enor- The Toxic and Antt-Toxic Action of Salts. 433 mously greater doses. The figures show no increase in poisonous action in passing from an,’ to an % solution. Beyond this all eggs are killed. This result is of interest for the reason that nickel and cobalt were marked exceptions to the rule of the relation of poison action and decomposition tension. According to the formula given in my paper, nickel chloride ought to be fatal in concentrations far below those found. This experiment indicates, I think, that the minimum fatal dose for this salt ought to be very close to its cal- culated value and far below that I assigned to it. For some reason, nickel chloride enters the egg with great difficulty. A feweggs hence escape its action, and these few resist enormous quantities and make the minimum fatal dose appear higher than it ought. The action of calcium chloride on the poisonous power of nickel chloride is shown in the following experiment (Table XII): TAMILS, QO Concen- Embryos tration. developed. 100 c.c. 100 c.c. +- 2 c.c. CaCl, mz 3 100 c.c. 100 c.c. + 2 c.c. CaCl, m 3 100 c.c. 100 c.c. + 2 c.c. CaCl, m 3 100 c.c. 100 c.c. + 2 c.c. CaCly me 2 100 c.c. 100 c.c. + 2 c.c. CaCl, m 3 100 c.c. 100 c.c. + 2 c.c. CaCle m 2 The anti-toxic action of calcium for nickel chloride is thus shown to be very great. A fatal dose of nickel chloride is no longer poisonous in the presence of #4 calcium chloride in which nearly go per cent of the eggs develop into embryos, and many of them hatch. 434 A. P. Mathews. The next experiment shows how small were the quantities of cal- cium chloride required (Table XIII). TABLE XIII. Embryos Concentration. Eggs. developed 29 “ + 0.2 c.c. CaCl, 3m 42 roe «| 60 “ +06 48 « +08 22 “ +10 4] ee 15 28 A plain effect can be seen with the calcium chloride present in 77, concentration, and its action is therefore comparable to that of stron- tium upon sodium chlorides. Is there a definite numerical relationship between the toxic action of a given number of molecules of one salt and one molecule of the anti-toxic salt? It will be remembered that Hardy found a numerical relationship between the precipitating powers of mono- and bivalent cations. A bivalent cation was many times as powerful as a monovalent cation. Loeb! concluded that a somewhat similar relationship existed for the toxic and anti-toxic action. To neutralize a single molecule of potas- sium chloride a very small fraction of a molecule of calcium chloride was necessary, whereas to neutralize the poisonous action of a biva- lent cation salt greater quantities were required. This numerical re- lationship, however, was determined for a single concentration, z. ¢., 3” or $x of the toxic univalent salt. It was desirable to see whether any such relationship actually existed. If one molecule of calcium chloride would neutralize 1000 of potas- sium chloride in 2 solution, it ought to be possible by greater con- centrations of calcium chloride to neutralize the poisonous power of more concentrated solutions of potassium chloride. I tried experi- 1 Logs: Archiv fiir die gesammte Physiologie, 1901, Ixxxvili, p. 72; zb2d., p: 76, 4G The Toxic and Anti-Toxic Action of Salts. 435 ments to see how concentrated solutions of the different monovalent salts it was possible to neutralize with calcium chloride. The results showed a great variation in different salts (Table XIV). TABLE XIV. Molecular relationship. Per cent of embryos developed. Concen- | Anti-toxic Concen- tration. salt. tration. Toxic salt. Toxin. Anti-toxin, 1200 The foregoing experiment illustrates the fact that the molecular relationships of the toxin to the anti-toxin are not fixed. When potassium chloride is present in a § z solution, it takes but one mole- cule of calcium chloride to 1200 of the potassium to yield 20 per cent of embryos. If, however, an ,8; solution is used, one molecule of calcium chloride to 640 of the potassium yields only 25 per cent of embryos. Ina 75 solution, the concentration of the calcium must be increased so that only one hundred and nineteen molecules of potassium chloride are present to one of calcium chloride in order to give any embryos. Above a ;% normal solution I have not obtained any embryos, although greater concentrations of calcium chloride than I have tried might possibly yield a few. With sodium acetate the following results were obtained, Table mV 436 A. P. Mathews. TABLE XV. Molecular relationship. | per cent of embryos developed. Concen- 3 Anti- ‘ : en- Toxic salt. tration. Conc toxic salt. tration. Toxin. Anti-toxin. NaC,H;0, Bin CaCl, 5000.0 2500.0 1665.0 1250.0 1000.0 665.0 119.0 117.0 180.0 239.0 38.7 It may be seen in this table that while sodium acetate in 3 x solu- tion is easily neutralized by very small quantities of calcium chloride, z.é., one molecule to sixteen hundred of the acetate, in the slightly greater concentration of 8, normal, a concentration of calcium ten times as great is wholly unable to yield a single embryo. With ammonium chloride still more striking results were obtained. In a 3 # solution of this salt some eggs developed. Even ina § m solution an occasional egg may form an embryo. Yet it is pos- sible only with large quantities of calcium to neutralize a 8 m solu- tion, as may be seen in Table IV. In a $2 ammonium chloride solution, still larger concentrations of calcium are required. No em- bryos were obtained with calcium chloride present in 7%’, and 74 con- centrations. In an % solution I obtained 13 per cent of embryos. This is a molecular relationship of 1 mol. calcium chloride to 16 of ammonium chloride. With lithium chloride the result was obtained that while this salt may be easily neutralized in concentrations less than 2 molecular, above a ¢ molecular concentration no amount of calcium chloride would give any embryos. This is shown in the following table Clablenx wi: The Toxic and Anti-Toxic Action of Salts. 437 TABLE XVI. Molecular relationship. | Per cent of embryos developed. Concen- Anti-toxic Concen- tration. salt. tration. Toxin. Anti-toxin, Ca(NOg)o 5 240.0 120.0 48.0 320.0 106.0 32.0 13.0 200.0 100.0 40.0 17.0 36.0 18.0 157 While, therefore, in a 3 a lithium chloride solution, a molecule of calcium chloride to each 120 of lithium chloride will yield 100 per cent of embryos, in a $ z solution, one molecule of calcium chloride to 106 of lithium chloride yields nothing, and 1 to 32 lithium chloride gives only II per cent. These figures show that a definite molecular relationship between the toxin and anti-toxin does not exist,and the conclusion is not jus- tified that one molecule of a bivalent cation salt is anti-toxic for a very large number of molecules of the monovalent toxic salt. This relationship is true only for certain concentrations of sodium and potassium salts. Lithium and ammonium salts require quite differ- ent molecular amounts of calcium from these other salts. The conclusions from these experiments may be summarized as follows: 438 A. P. Mathews. 1. The toxic action of any salt shows no definite relationship to the valence of either ion, but is related to the decomposition tension of the salt. 2. The toxic action is due to both ions equally. The anti-toxic action must hence involve both ions. 3. The anti-toxic action of the salts of the bivalent metals for the salts of the univalent metals involves all ions both of the anti-toxin and the toxin, and is not due to an anti-toxic action between dif- ferent cations of different valence. 4. The salts of different bivalent metals show the widest varia- tions in anti-toxic action, ranging from no anti-toxic action at all, to very strong anti-toxic action. Magnesium chloride is only 54 as strong as calcium chloride. 5. It is possible to neutralize the toxic action of monovalent by monovalent salts and bivalent by bivalent salts. | 6. The amount of calcium chloride or other bivalent salt required to neutralize different monovalent salts is different for every salt. Ammonium and lithium chlorides are neutralized with the greatest difficulty. : 7. The number of molecules of calcium chloride required to neu- tralize one molecule of any monovalent salt is not a fixed quantity, but varies with the concentration of the monovalent salt. In some cases a very slight increase in the concentration of the monovalent salt makes it impossible to neutralize its toxic action except by enormous doses of calcium chloride. No characteristic quantitative relationship between monovalent and bivalent salts in their anti- toxic relationships is to be discerned. 8. From these facts it ts clear that valence, as such, either of the anion or the cation ts of secondary or no importance in determining either the toxic or anti-toxic action of the salts. So far the conclusions reached have been negative. What is then the true explanation of anti-toxic action ? Iam not able to answer this question as a whole, but the following considerations explain some of the phenomena. In the first place, is there any relation between the decomposition tension of the salt and its anti-toxic power?! The facts show that only those salts of sodium, potassium, and lithium of a high decom- position tension can be neutralized. As soon as an anion of low 1 LILLIE (This journal, 1904, x, p- 443) finds such a relationship for the anti- toxic action of ions on ciliary movement. The Toxic and Anti-Toxic Action of Salts. 439 solution tension is added, it becomes much more difficult or al- together impossible to neutralize the compound. In the second place, the anti-toxic metals all fall practically in the list of solution tensions above cobalt. They are metals for the most part of high decomposition tension. Whether, however, this relationship can be extended farther than this general statement, and an explanation of the facts be founded upon it, cannot at present be said. In the second place, a part of the anti-toxic action is undoubtedly due, I think, to an alteration in permeability of the egg membranes «by the anti-toxic salt. The blood and presumably the tissue of the teleost fishes have an osmotic pressure of about one-half that of the sea-water. The monovalent salts which can be counteracted by cal- cium and other salts are only poisonous in concentrations approxi- mately equal to, or in some cases in excess of the osmotic pressure of the sea-water. When a Fundulus egg is brought into a sodium chlo- ride solution of 8 7 concentration a difference of about twelve atmos- pheres of pressure must exist between the inside and the outside of the egg. It is clear that were the egg membranes not readily perme- able to salts, this great pressure would at once plasmolyse the egg. A very slight change in permeability could, therefore, profoundly ef- fect the ease of plasmolysis. There is evidence that the poisonous action of sodium, ammonium, lithium, and potassium chlorides is due in part to the fact that they cannot enter the egg with entire freedom. When put into such solutions, the protoplasmic envelope of the egg may be seen to shrink away from the outer membrane, and in case the solution is strong enough, complete plasmolysis takes place and the membrane is driven into a compact mass at the centre of the egg. A part of the poisonous action is, therefore, undoubtedly due, I think, to the plasmolysing action of the salts. Itis this plasmolysing action particularly which calcium salts and other bivalent metals are able to offset. If the eggs are placed in $-7 7 potassium chloride solutions containing calcium or any other anti-toxic salt, this plasmolysis does not occur. The egg does not shrink, but remains uncoagulated and of its normal size. These facts point to the conclusion that a part, at least, of the anti-toxic action is due to an alteration in permeability of the cell- membrane by the calcium or other anti-toxic salt, as suggested by ' Logs refers to Fundulus eggs as being less permeable to salts than Arbacia. The contrary is the case, as otherwise Fundulus would be plasmolysed more easily than Arbacia. 440 A. P. Mathews. Stewart.!. This alteration is of such a sort that there is an increased permeability for lithium, sodium, potassium, and ammonium com- pounds. By this means, that part of the poisonous action of the salt which is due to plasmolysis is neutralized. Whether this explains the whole of the anti-toxic action is at present uncertain. It may be that calcium may also neutralize the specific poisonous action of the salts, but this appears to me doubtful. This explanation clears up also the fact that while potassium chloride may be neutralized up to a 0.9 z solution, lithium chloride can hardly be neutralized in concen- trations beyond its minimum fatal dose. The reason for this would be that the specific poisonous power of lithium chloride, as computed from its decomposition tension, is almost that of its minimum fatal dose. That is, only a small part of the poisonous action of a $2. lithium chloride solution is due to plasmolysis; it is, therefore, im- possible to neutralize lithium in much greater concentrations. With potassium, however, much of its poisonous action is due to plasmoly- sis. Its true poisonous dose should be 0.97. It is possible to neu- tralize potassium chloride in concentrations weaker than this, but not above it. This explanation can obviously not apply to such salts as manganese chloride, cobalt chloride, and nickel chloride, which can also be neu- tralized by calcium chloride, but which are poisonous in concentrations far lower than that necessary to plasmolyse. The explanation of this antagonism, in my opinion, is that calcium reduces the permeability of the egg membranes to cobalt and nickel and manganese salts. The evidence for this view is as follows: Normally, cobalt and nickel find their way into the egg with great difficulty. This is shown by the fact that the action of these salts is extremely irregular. Whereas the great majority of the eggs exposed to their action are killed by concentrations less than ;745, a certain number of the eggs withstand concentrations as high as #6. Further- more, when these salts enter the egg, they produce a quite different change from that produced by poisonous doses of sodium chloride. In the former case there is no plasmolysis, but instead the embryo turns white or pink and is coagulated zz sz¢éu. The action of the salts is also very slow, as if they entered with difficulty. While many embryos may live in an ,”%, solution for twenty-four hours, most 1 STEWART: American year book of medicine and surgery (medicine), 1904, p. 527; Cf This journal, 1903, ix, p. 96; Comptes rendus, xiv, Congres interna- tional de médecine, 1903, p. 84, Section de physiologie. The Toxic and Anti-Toxic Action of Salts. 441 of these are coagulated before forty-eight hours. The same appear- ances are true for manganese, though the embryos are not so easily coagulated by this salt. The appearance of the embryo when developing in the presence of a mixture of calcium and cobalt or nickel salts is quite different from that in cobalt or nickel without the calcium. In the latter case most eggs are coagulated and white; in the former they do not coagulate. We may conclude, therefore, that when calcium is present, the cobalt,. nickel, and manganese do not enter the egg, and accordingly the embryo is protected from them. How the permeability is altered by calcium is not clear. It may be, however, that the membrane is slightly coagulated, so that its texture is coarser, and potassium, ammonium, and sodium salts can pass through more readily. As regards cobalt and nickel, these ions may have a bulk too great to pass through even the coarser mem- brane. They may find their way into the egg ordinarily by first combining with the membrane, and in this way getting in.’ If cal- cium is already in combination they can no longer combine with the membrane and they are too large to enter the free spaces in the membrane. Whatever the exact mechanism by which a change in the permeability is produced may be, I feel confident that a portion at least of the anti-toxic action is to be explained by this alteration in permeability of the egg membranes. Further evidence of this alteration in permeability was secured in the following way: Sea-urchin eggs will not divide after fertilization if the concentration of the sea-water is raised. The careful exam- ination of the segmentation problem by Spaulding? led him to con- clude that this process involved a balance between intrinsic and extrinsic forces. Osmotic pressure is one of the forces which tends to check cell-division, since it tends to drive the egg into a spherical form and to prevent any enlargement of its surface. Now if a sea- urchin egg is placed in pure sodium chloride of an osmotic pressure equal to that of the sea-water, it shrinks, becomes spherical and will not divide, or divides very slowly. This shows that in sodium chlo- ride, although the actual osmotic pressure is the same as that of sea- water, its efficiency is increased. This increased efficiency indicates that the egg membranes are more permeable to sodium chloride in 1 It is possible that the bulk of the free ion of cobalt is very much greater than the atom in a non-ionized form. 2 SPAULDING: Biological bulletin, 1904, vi, p. 98. 442 A. P. Mathews. sea-water than to sodium chloride alone. In other words, the resist- ance of the egg membrane to the passage of sodium chloride is greater in pure solutions of this salt than when other salts are pres- ent. If this is the case, the addition of calcium chloride may be expected to reduce the effect of the sodium chloride and to permit the egg to segment. Experiment confirms this hypothesis, at least in part, as is shown by Loeb’s observations on the effect of calcium on eggs in sodium chloride solutions. The following experiment was tried to test this hypothesis further (Table XVIJ). TABLE XVII. Arbacia eggs fertilized at 10.45 A.M., and transferred to the solutions noted. Examined at 2 P.M. 100 c.c. NaCl. 100 c.c. NaCl + 0.7 c.c. 3m CaCly. } Concentration. Result. Concentration. Result. Most are dead in . 2 2+ CaCl, Most alive; 2-32 cells. meee ce there is . 332+ CaCl, Twelve eggs, 16-20 cells; one 5 cells; nine ten, 8 cells. Alive. fered nites Six- . ¢n + CaCl, Eggs in 16-64 cells. ve ere . $32 -+ CaCl, All 6f cells. Alive. ieee . 32 + CaCl, Alive. 2-32 cells. Eggs black and ; + CaCl, Undivided for the most shrunken part, but normal looking. There are, however, certain differences in the reactions of these eges from those of Fundulus. Thus calcium chloride would net neutralize the action of potassium chloride nearly as well as it did sodium. In fact I could get no neutralization of the poisonous action of potassium for these eggs, a fact which indicates that the anti-toxic action is not due in all likelihood to the same causes in different cases. Barium chloride also showed itself almost inert as an anti- toxic agent in all concentrations tried, but was itself exceedingly toxic. Magnesium chloride showed itself the best antagonist of sodium chloride. The facts indicate an increase in permeability of the cell wall due to magnesium, calcium, and barium ions, but show clearly that there are certain other harmful actions of potassium and sodium compounds which these ions cannot inhibit. The Toxic and Antt-Toxic Action of Salts. 443 The general conclusion to be drawn from these experiments is that the interpretation of the anti-toxic action of one salt upon an- other given by Loeb is incorrect, at least in part. Valence either of the cation or anion appears to have no share either in toxic or anti- toxic action. While this part of Loeb’s explanation is wrong, it may still be that the action involves the electrical charges upon the ions, but in the way stated in my papers on solution tension and nerve stimulation.! It appears probable that several different causes may operate in determining anti-toxic action. In some cases the anti- toxic salt increases the permeability of the cell-membranes, thus relieving the cell from plasmolysis injuries; in other cases the cell- membrane is rendered less permeable to a given ion, thus protecting the protoplasm within from its influence. A part of the action may, however, be due to the necessity of preserving a certain viscosity of the protoplasm, in order that delicately adjusted protoplasmic pro- cesses may go on, and for this viscosity ions of definite potential must be present in proper proportions. It is probable, I think, that the facts showing the advantage of three or more metals for ciliary and muscular movement, and possibly nuclear division, are to be thus explained. In so complicated a process as cell-division, in which surface tension, protoplasmic viscosity, and respiration play a part, it is probable that all these factors come into play, and that while the anti-toxic action of calcium upon sodium chloride may be pronounced in its effects, for example, on surface tension, the salts may not antagonize each other’s harmful action in their relations to other processes, such as karyokinesis.? If the valence of an ion is of little or no importance in determin- ing its anti-toxic action, attention must be directed to its other prop- erties. Among these the ionic potential is indicated, both by Lillie’s results and mine, to be of great importance, but it is not impossible that the ionic weight, velocity, and volume will all be found to play a part. 1 MATHEWS: This journal, 1904, xi, p. 455. ? This general conclusion is practically the same as that reached by STEWART: American year book of medicine and surgery (medicine), 1904, p. 527. THE QUANTITATIVE ESTIMATION OF CARBAMATES. By J. J. R. MACLEOD anp H. D. HASKIne= [From the Physiological Laboratory, Western Reserve University, Cleveland, Ohio.| HE physiological significance of carbamates was first pointed out by Drechsel.2, This worker found carbamates to be formed when amido acids and other nitrogenous organic compounds were ox- idized in the presence of alkalies, and he was also able to demonstrate traces of them normally present in the blood serum of dogs. Hahn and Nencki® further found carbamates present in minute traces in the normal acid urines of the dog and man. In the alkaline urine of horses very considerable amounts of carbamates have been detected by Drechsel and Abel,* and, by feeding dogs with lime until the urine became alkaline towards litmus, Abel and Muirhead ® were able to separate, in a comparatively pure state, considerable quantities of this substance. In a series of dogs in which an anastomosis of the portal vein with the inferior vena cava had been successfully established by Pawlow and Massen, Hahn and Nencki demonstrated considerable quantities of carbamate in the urine, and these workers concluded that the ner- vous symptoms which frequently make their appearance after this operation are due to the presence of excess of carbamates in the blood. The amount of carbamates was found to be still greater when, in the Eck’s fistula dogs, the liver was also removed from the circu- lation by ligature of the hepatic arteries. Lieblein’ has also noticed 1 H. M. Hanna Research Fellow, Western Reserve University. 2 DRECHSEL: Vide No.iF: Zeitschrift fiir physiologische Chemie, 1897, xxiii, De 155s 8 HAHN and Nenckt: Archiv fiir experimentelle Pathologie und Pharma- kologie, 1893, xxxii, p. 185. 4 DRECHSEL and ABEL: Archiv fiir Physiologie, 1891, p. 231. 5 ABEL and MUIRHEAD: Archiv fiir experimentelle Pathologie und Pharma- kologie, 1892, xxxi, p. 15. 6 PaAWLow, MASSEN, HAWN, und NENCKI: Archiv fiir experimentelle Path- ologie und Pharmakologie, 1893, xxxii, p. 161. 7 LIEBLEIN: Archiv fiir experimentelle Pathologie und Pharmakologie, 1894, XXxill, p.'327- 444 The Quantitative Estimation of Carbamates. 445 the presence of an excess of carbamate in the urine of dogs after partial destruction of the hepatic cells by injecting dilute sulphuric acid into the biliary ducts. The chemical method mainly employed in the above investigations for the detection of carbamates was that recommended by Drechsel.! This method consists in shaking the urine for ten minutes with an excess of milk of lime;. the solution is then filtered or centrifugalized, and the clear fluid, thus obtained, again shaken with a few drops of a solution of calcium chloride and with a few crystals of calcium carbo- nate. By this second precipitation, all] traces of soluble carbonate are converted into the calcium salt which, by the shaking with crystalline carbonate, is rendered crystalline, and therefore easily separable by filtration. By the above process any carbamates which the solution may contain are converted into the calcium salt, which is soluble in water. After standing on ice for some time, the solution is filtered into an excess of alcohol at a temperature of o° C., in which — after again standing for some time — calcium carbamate separates out asa precipitate, which, however, is highly impure. To purify it, the pre- cipitate is dried in a desiccator, and then dissolved in a small amount of 10 per cent ammonia, the resulting solution being then subjected to fractional precipitation, the third precipitate obtained being com- paratively pure carbamate. This final precipitate, when dissolved in ice-cold water and the resulting solution placed in a closed tube, gives a solution which is at first perfectly clear, but which, on raising the temperature, becomes cloudy from the separation of calcium carbonate. The preparation thus obtained is, however, far from pure;? for, when it is dissolved in water and the resulting solution heated, a very variable amount of ammonia and carbon dioxide gas is obtained.® Theoretically, for each molecule of carbon dioxide which such a solu- tion yields, there ought to be two molecules of ammonia, as is evident from the following equation: (H,NCO:O).Ca oh H,O = CO; + 2 NH; + Cio... The above method is entirely unsuited for quantitative work.® 1 DRECHSEL and ABEL: Loc. cét., p. 238. ? Aromatic sulphates and volatile fatty acids were found by HAHN and NENCKI (oc. ctt.) in the preparation. ® NEUBAUER and VOGEL: Analyse des Harns; analytischer Theil, 1898, p. 270. 446 J. J. R. Macleod and H. D. Haskins. It has been further pointed out, especially by Nolf! and Hofmeis- ter,? that during the manipulations of Drechsel’s method a consider- able amount of carbamate may be formed. These authors point out that any watery solution of ammonium carbonate must contain at least some carbamate, and that when such solution is shaken with milk of lime, the amount of carbamate will increase, on account of the liberation of free ammonia. Not only ammonium carbonate, but any carbonate, will form carbamate when shaken with alkalies in the presence of substances which readily yield ammonia (Hofmeister), or where free carbon dioxide gas exists along with ammonia salts. In all these cases the formation of carbamates can be explained by the chemical interaction between carbonates and ammonia, so that much doubt is thrown on the existence of preformed carbamates in animal fluids. The importance of a reliable test for the presence of carbamates, and the possibility of securing a method for estimating the amount of them present in animal fluids, especially in conditions where the circulation through the liver is interfered with, as after an Eck’s fis- tula has been instituted, prompted us to see whether, by an estima- tion of the carbon dioxide gas in fluids, before and after precipitation of the carbonates by suitable reagents, a more reliable method than that of Drechsel could not be obtained. In this account we will con- fine ourselves to a description of the method as used for various fluids, and to proofs of its accuracy; and we will reserve for a future paper a fuller account of the results of bio-chemical interest which have been obtained by it. The principle of the method is as follows: the carbon dioxide gas in I c.c. of the fluid under examination is estimated in the apparatus. of Barcroft and Haldane,? the temperature of the water bath being noted on a delicate thermometer. Another cubic centimetre of the fluid is vigorously shaken in a stoppered weighing bottle with an excess of a saturated watery solu- tion of barium hydroxide * containing ammonia. This causes com- plete precipitation of the carbonates, but not the carbamates, as the barium salt of these is soluble in water (Drechsel). After about one- LINGLE siacicer * HOFMEISTER: Archiv fiir die gesammte Physiologie, 1872, xii, p. 337- 3 BARCROFT and HALDANE: The journal of physiology, 1902, xxviii, p. 232. * For this shaking we have recently used a mechanical shaking apparatus driven by a water motor. The Quantitative Estimation of Carbamates. 447 half hour the fluid is transferred to a tube, which is tightly corked and centrifuged, and the carbon dioxide gas determined (at'the same temperature as that used for total carbon dioxide), either in the su- pernatant fluid or in the precipitate; in the supernatant fluid when blood, or other fluids rich in proteids, is under examination; in the precipitate in other cases. The gas found in the supernatant fluid corresponds to that of carbamates. The difference between the amount of gas found in one cubic centimetre of the total fluid, and that found in the precipitate, also corresponds to the gas of carbamates. We will first consider the method employed for simple solutions of carbamates, 1. e., where the precipitate is examined. The reliability of this method must depend mainly on the entire precipitation of carbonates by barium hydroxide and ammonia. That such is the case is proved by the following results: I. 0.5 €.c. 7 Sodium carbonate (approx.) gave, when mixed with tartaric acid in the Barcroft and Haldane apparatus, 0.584 c.c. carbon dioxide. 0.5 c.c. of the same solution was mixed with baryta water and ammo- nia, shaken and centrifuged. The precipitate washed into the gas bottle and mixed with tartaric acid gave: 2 SOSSIMNGE: 65 JOISS2icie: 6. (0:5 ,718c:c; 2. In another series of tests with 0.5 ¢.c. sodium carbonate solution which had stood for some time, 0.567 c.c. and 0.559 c.c. carbon dioxide were found, and the barium precipitate obtained as above described gave 0.557 c.c. and 0.566 c.c. The presence of other neutral salts besides carbonates does not interfere with this precipitation. r. 1 c.c. of an artificial plasma solution (made by dissolving the various salts of plasma in proper proportions in distilled water) gave 0.593 c.c. of car- bon dioxide. The barium precipitate obtained by the above method gave 0.586 c.c. 2. 1 c.c. freshly passed human urine containing added sodium carbonate gave 0.566 c.c. and 0.552 c.c. carbon dioxide. The barium precipitates treated as above gave 0.566 c.c. and 0.578 c.c. In the above cases, viz., where phosphates, etc., exist in the fluid, a heavy precipitate follows the addition of barium hydroxide, and, to entirely displace all the carbon dioxide gas from this precipitate, an 448 J. J. R. Macleod and H. D. Haskins. excess of tartaric acid must be added. This we accomplish by using, instead of the glass spoon supplied with the apparatus, a small flat bottomed tube, closed at one end, and holding about 1 c.c. of acid; this tube is placed in the gas bottle, and used in the same manner as that employed in decomposing urine with hypobromite. The presence of proteids retards considerably the precipitation of the carbonates, but, with certain precautions, the entire precipitation of these can be effected. Without adopting these precautions, results like the following are obtained : | 1. 1 c.c. of sodium carbonate solution containing egg proteid gave 0.680 and 0.699 c.c. gas. The barium precipitates from two samples of 1 c.c. each of this solution gave 0.561 c.c. and 0.593 ¢.c. 2. 1 c.c. of artificial plasma containing egg proteid gave 0.648 c.c. and 0.655 c.c. carbon dioxide. 1 c.c. of the same solution, precipitated as above with barium, and the precipitate thoroughly washed several times with weak ammonia water,— the supernatant fluid being removed by decanta- tion,— gave 0.526 c.c. gas. With blood plasma even worse results are obtained. If, however, the proteid solution be considerably diluted with weak ammonia, and be treated with excess of barium solution and frequently shaken for one and a half hours, the precipitation of carbonates is much more complete. I. 1¢.c. 25 per cent egg white in artificial plasma solution gave 0.566 c.c. and 0.566 c.c. carbon dioxide. 1 c.c. of the same solution with 8 c.c. barium hydroxide solution and 10 c.c. ammonia water (0.5 per cent), after fre- quent shaking during 14 hours, was centrifuged until the supernatant fluid was clear. The precipitate was then washed twice by decantation with 0.8 per cent ammonia, and yielded 0.753 c.c. carbon dioxide. In this case it is obvious that carbonates had been formed, during the prolonged standing, from the carbon dioxide in air. To correct for this error, a blank of 8 c.c. barium hydroxide solution and ro c.c. 0.5 per cent ammonia was carried out as above, and the precipitate gave 0.185 c.c. gas. Deducting this latter figure from that for the first precipitate, 0.568 c.c. carbon dioxide is obtained. Exactly similar determinations were made with blood serum; the following table (No. I) gives the results : The Quantitative Estimation of Carbamates. 449 TABLE I. Carbon dioxide of carbonates precipitated. Carbon Carbon dioxide in dioxide in precipitate. blank. Total carbon Fluid examined. dioxide. c.c. c.c. c.c. c.c. loodrserum, <- .. . » 0.380 0.544 0.158 0.386 icod serum: . «|... . 0.377 0.535 0.158 0.377 OOGCSELUM s+ 6! sce som 0.382 0.333 0.074 9.259 POOdISerulm « - ¢ - 0, /« 0.370 0.301 From this table it is seen that, although by frequent shaking it is possible to precipitate all the carbonates, the method is an uncertain one and difficult. This uncertainty is due to the difficulty in shaking down the carbon- ate precipitate. We will describe later a modification of the method which we use for blood serum. To summarize, all the carbonate in simple solutions, or in solutions such as artificial plasma or urine which contain no proteid, can be precipitated by shaking with a saturated solution of barium hydrox- ide, containing 0.5 per centammonia; where, however, any consider- able amount of proteid is present in such solutions the precipitation of carbonates is uncertain. The second factor on which the reliability of the method depends, is that saturated baryta solution does not have any action on carba- mates, that it does not accelerate the decomposition of carbamates into carbonates. This can be easily shown by dividing a solution of calcium carba- mate in ice-cold water into two equal parts, and adding barium hydroxide to one of these. If kept cool in tightly corked tubes, both solutions remain clear for an equal period of time. The table on the following page (No. II) depicts some of the results which we have obtained by the above method in simple solu- tions of carbamates. We give these results for the purpose of demonstrating the utility of the method, and we do not intend, in this article, to discuss in any detail the facts relating to the behavior of carbamate solutions which they demonstrate. It may be stated, however, that the results are entirely confirmatory of what is already known regarding this sub- ject, viz., that the stability of such solutions depends on temperature 450 J. J. R. Macleod and H. D. Haskins. (1 and 2, 3 and 4), tension of ammonia present (9, 10, 11, and 12), and exposure to air (2 and 3). By exposure to air the ammonia formed by decomposition of the carbamate diffuses out of the solu- tion, so that all the carbamates soon become decomposed; whereas, TABLE II. = S a = IV. Condition under which solution was before baryta, etc., was added. Nature of No. solvent used. monium added. solution stood before baryta, etc., was added. Carbon dioxide gas in precipitate of 1 c.c. of fluid. Difference be- tween IIT and IV; i. é., gas of car- Amount of car- bamate of am- Carbon dioxide gas in 1 c.c. of fluid Length of time ° a ° ° Distilled In open bottle at} 3 hour room-temperature In open bottle on} 1 hour ice Not In closed flask on | 23 hours weighed ice ce In closed flask at} 20 min. body-temperature In closed tube on| 14 hours ice In closed flask at} 1h. 20m. room-temperature In closed flask at} 20 min. body-temperature In loosely corked| 2 days flaskat room-temp. In closed tube on} 14 hours ammonia ice 0.33% In closed flask at] 1h.10m. ammonia room-temperature 05% In closed flask at | 20 min. body-temperature In loosely stop-| 2 days ammonia pered flask at room-temperature S eH) y= = ON 1 In none of these was any correction made for carbon dioxide absorbed from the air during the manipulations, nor was any great precaution taken to have the temperature of the water bath always the same. 2 Excess of gas in VI due to absorption of CO, from air. 8 Excess of gas in VI due to absorption of CO, from air. when this diffusion is prevented, a solution of carbamate may remain undecomposed for days, even at room-temperature. The results further show that the calcium salt is more stable than the ammonia salts. (Compare 3 and 4 with 5, 6, and 7.) We have also made several estimations by the above method of the The Quantitative Estimation of Carbamates. 451 amount of carbamate in the urine of a dog fed on lime.!_ The follow- ing table gives the results, as well as results obtained by the same method on normal dog’s urine: ” TABLE TL: II. III. IV. Nic VI. VII. VIII. Amount Carbon Gaho Carbon Carbon Carbon of urine | Carbon | dioxide in pateted dioxide in dioxide in | dioxide of No. | used for cee barium barium, ae eas Sake diel exami- in II. | precipitate Poe fe) Re - fo) mates nation. of II. (=1:1—V),|(=III—VI).| per c.c. dioxide in c.c. c.c. c.c. ea cE c.0. c.c. 0.189 0.1580 0.048 0.110 0.079 0.026 0.134 0.1211 0.033 0.088 0.046 0.015 1.022 0.9560 0.019 0.937 0.085 0.085 0.635 0.5460 0.020 0.526 0.109 0.109 1.934 1.7300 0.023 * 1.707 0.224 0.224 0.0335 0.0480 0.0737 0.0988 0.020 0.0788 Nos. 1 to 5 are from a lime-fed dog; Nos. 6 and 7 from a normal dog. The estimations confirm Abel and Muirhead’s discovery? that carbamates appear in the urine of dogs to which lime is administered until the urine reacts alkaline to litmus. Our results show further that the amount of carbamate steadily increases from day to day until the investigation has to be terminated on account of gastric dis- turbance. We found it essential for satisfactory results that only catheter specimens be examined. If the urine be collected in the ordinary metabolism cage, fermentation of urea is apt to set in, and ammonium carbonate to be formed, a certain amount of which becomes converted into carbamates.* In one specimen of urine (No. 4), the carbamates were isolated by Abel and Drechsel’s method. In all the estimations a control of the reagents in the same quantities 1 Vide ABEL and MUIRHEAD: Loc. cit. 2 Catheter specimens were used throughout, and were examined as soon after collection as possible. 8 ABEL and MUIRHEAD: Loc, cit. * NoiF: Loc. cit. 452 J. J. R. Macleod and H. D. Haskins. as used for precipitation was carried through, and the carbon dioxide gas found in the resulting barium carbonate precipitate — derived from air — deducted from that of the barium carbonate precipitate of the urine. On account of the uncertainty of precipitating all carbonates by the above process zz solutions containing proteids, we have found it necessary to adopt an entirely different method for estimating carba- mates in such fluids. The principle of this latter method is similar to that used for non-proteid solutions, only that, instead of the precipitate, an aliquot part of the supernatant fluid is taken for estimation of carbon dioxide gas. (See page 447.) Since in transferring this solution from the centrifuge tube to the blood-gas bottle of the Barcroft-Haldane apparatus, carbon dioxide is absorbed from the air by the barium hydroxide which the solution contains, it is necessary, in al] estimations, to carry out controls, and to deduct from the amount of gas found in the solution under exami- nation, the gas which is produced in the control. On account of the readiness with which watery solutions of carba- mates undergo decomposition into carbonates at room-temperature, we have found it necessary to insure the presence in the solution of a considerable tension of ammonia. This precaution is especially neces- sary when the percentage of carbamate is low. On the other hand, an excess of ammonia—in which, of course, decomposition of even larger amounts of carbamates would be entirely prevented — cannot be used, and for two reasons: firstly, because the ammonia would form carbamates with the carbonates present in the serum,! and secondly, because, with an excess of ammonia, the estimation of the gas in the Barcroft-Haldane apparatus is rendered inaccurate. Before proceeding further with the elaboration of the method, therefore, it was necessary for us to determine what percentage of ammonia might be used without creating either of the two errors indicated above. For this purpose we have carried out several tests of which the following is a type: Into five small test-tubes were placed, respectively, 2 c.c. water; 2 c¢.c. 0.5 per cent ammonia; 2 c.c. 1 per cent ammonia; 2 clemas per cent ammonia; and 2 c.c. § per cent ammonia. In each of 1 Vide DRECHSEL, HOFMEISTER, NOLF: Loc. cit. 2 Throughout the paper these figures represent per cent of agua ammonia. Specific gravity, 0.9. The Quantitative Estimation of Carbamates. 453 the solutions 0.001 gr. calcium carbamate! was dissolved, and the tubes corked and left standing at room-temperature. In about one minute the watery solution had become densely opaque from the separation of calcium carbonate. In a few minutes calcium carbonate had also commenced to separate from the 0.5 per cent solution, and a little later from the 1 per cent solution; but the 2.5 per cent solution was perfectly clear, even after standing one hour. A tension of 2.5 per cent ammonia, therefore, prevents the decom- position of 1 mg. calcium carbamate. With regard to the second precaution, interference with the accuracy of the gas-determination in the Barcroft-Haldane apparatus, a 5 per cent solution of ammonia gives a dense precipitate with tartaric acid, and this interferes with the reading. With a 2.5 per cent solution, however, only a few crystals separate out and the reading is accurate. The more carbamate present, the higher must be the tension of ammonia to prevent decomposition, and the choice of 1 mg. for the above tests may seem somewhat arbitrary; but 1 c.c. of blood serum could seldom be expected to contain more carbamate than I mg., for this amount of calcium carbamate itself gives 0.262 c.c. of carbon dioxide gas, and 1 c.c. of blood under normal conditions from 0.320-0.400 c.c. Even should the blood serum contain considerably more carbamate than this, it is improbable that any decomposition into carbonate would occur within the hour which is required for precipitation and centrifugalization. The presence of such considerable amounts of ammonia introduces other difficulties besides the formation of a precipitate when the fluid is transferred to the gas bottle, viz., that it takes a large amount of the tartaric acid solution to neutralize all the ammonia, and that when the solution is shaken in the bottle, a minus reading, due to suction pressure, may result. This suction pressure persists after the fluid in the blood-gas bottle has been rendered distinctly acid, and its cause we cannot explain. Both these difficulties can be removed by nearly neutralizing the fluid before connecting up the blood-gas bottle with the manometer, and completing the acidifying after the manometer has been connected by adding 0.25 c.c. of tartaric acid previously placed in the spoon attached to the stopper of the apparatus. To insure complete precipitation of carbonates, we also add, besides. the above reagents, a small amount of barium chloride solution. 1 As nearly as could be weighed. 454 J. J. R. Macleod and H. D. Haskins. The method for blood and similar fluids 1s briefly as follows : The blood is collected directly from the blood-vessel in a clean, dry cen- trifuge tube of 12-16 c.c. capacity, and containing a little mercury. The tube is tightly corked and then vigorously shaken. By this process the blood is defibrinated. The tube is then placed on the centrifuge. While the blood is being centrifuged, a mixture of 7 c.c. of clear baryta solution, 2 c.c. water previously boiled to remove all gas, and 0.5 c.c. of a I per cent solution of barium chloride is placed in two weighing bottles (A and B) of 25 c.c. capacity. One c.c. of the blood serum is then delivered under each of the solutions in the weighing bottles, and to A is then added 3.5 c.c. of 10 per cent ammonia solution. Both bottles are firmly stoppered. The bottle A is placed on a mechanical shaking apparatus. The bottle Z is placed on a water bath, and the temperature kept at 60° C. for about fifteen minutes. By thus warming, any carbamate which the solution may contain is converted into carbonate The bottle B is then cooled, and 3.5 c.c. 10 per cent ammonia added; it is then placed on the shaker. After shaking for half an hour, the contents of the weigh- ing bottles are transferred to centrifuge tubes which are tightly corked, and placed on a rapid centrifuge for about fifteen to twenty minutes. Seven c.c. of the supernatant fluid (corresponding to 0.5 c.c. serum) from each tube is then transferred to the gas bottle of the Barcroft-Haldane apparatus. An amount of a saturated watery solution of tartaric acid, almost sufficient to produce neutral- ization — determined by previous titration — is then delivered under- neath the fluid in the bottles, and 0.25 c.c. of tartaric acid solution is placed on the spoons. The bottles are then attached to the manometers, — which, for convenience, are marked A and 8#, corre- sponding to the bottles, —and the temperature of the water bath registered on a thermometer reading one-tenth of a degree, brought to a fixed point near that of the temperature of the air of the room.” The temperature in the bottles having become constant, — as deter- mined by the manometers, — the acid in the spoon is spilled into the 1 Where the presence of a considerable amount of carbamate is expected —as after injecting carbamates into the circulation, or where carbamates have been directly added to the blood, etc. —the serum used in the control B should be some of the normal serum of the same animal, removed before the injection or addition of carbamates is made. 2 To stir the water in the water bath we have used a mechanical stirrer con- nected with a water motor. The Quantitative Estimation of Carbamates. 455 fluid in the bottles, and these are vigorously shaken, When all the gas has been expelled, the bottles are replaced in the water bath and cooled to the previous temperature. The meniscus of fluid in the manometers is then adjusted and read, and the calculation made as follows: The reading in the manometer B is deducted from that of A, and the product multiplied by the capacity of the gas bottle and tubing. This latter value, divided by 10,000,? gives the number of cubic centimetres of carbon dioxide gas derived from carbamates. To calculate how much calcium carbonate this represents, the above result should be multiplied by 3.8. 1 c.c. carbon dioxide equals 0.0038 gr. calcium carbamate (approx.). A small positive reading is frequently obtained from the control B,. This may be due to the presence of carbonates formed from the air during transference of the fluid to the gas bottle, or to the forma- tion of carbamates by the action of the ammonia on the carbonates, or to the incompletely precipitated carbonates of the serum. The most probable of these causes is the formation of carbonates from contact with air. A mixture of the reagents alone, treated exactly as above, but with less ammonia, gave on three different occasions readings of 7 mm., 6 mm., and 6 mm., and a saturated baryta solution alone, 8 mm. and 6 mm. With the same per- centage of ammonia as above, the readings were —19 mm. and —18.5 mm. Whether this be the correct explanation or not is of little impor- tance for our purpose, for in any case the same reaction will take place in both bottles. That none of this gas is derived from unpre- cipitated carbonates of the blood serum is proved by the fact that the reagents alone give a similar reading. That the difference be- tween the readings of A and B really corresponds to carbamate, can be further shown by examining, by the above method, blood serum which has stood for some time, and in which, therefore, all carbamate will have become converted into carbonate. In such a case these readings were found to be in one experiment: —19 mm. for A, and 1 The bottles A and BZ are practically of the same capacity, so that this deduc- tion is allowable. 2 The fluid in the manometers is a solution of ehioniic acid of 10.30 specific gravity. If water were used in the manometers, the factor for division would be the barometric pressure in millimetres of water, viz., 10,300; but by having the fluid in the manometer of the above specific gravity, adjustment of the decimal is all that is required. The use of the apparatus in this way was suggested to us by Dr. Barcroft. 456 J. J. R. Macleod and H. D. Haskins. —15.5 mm. for &. It will be noted that in these cases a marked negative reading was obtained. To demonstrate the value of this method we offer the following results: To 1 c.c. of blood serum diluted with ammonia water to make 7 c.c. of fluid containing 5 per cent ammonia was added I-mg. calcium carbamate.! When the carbamate had all dissolved, 7 c.c. of TABLE IV. It II. IUOE. ION As Vi: VI. VET: Awntaxe Carbon Pr race Gas of carbamate Difference I ae dioxide ob- Peas oe re (difference between between Wee ® tained from |) sae anew ee IV andV). amount No. | amount of half of II super- super- founda carbamate ( hacia natant hatant, |" >= aiane added.t 5 a) fluid of A.| fluid of 2. Pres- cubic cen- dded ee ae sure. timetres. a _————= | gram c.c. mm. mm. mm. c.c. 1 0.0011 0.1207 55.5 17 38.5 0.131 0.0103 2 0.0012 0.1310 omitted in notes 39.5 0.134 0.0030 3 0.0012 0.1310 2 os 18 37.5 0.129 0.0020 43 0.0011 0.1207 2 40.0 7 33.0 0.122 0.0013 5 small amt. 0.1510 49.5 6 43.5 0.148 0.0030 6 es 0.1235 24.0 —8.5 32.5 0.112 0.0110 1 Calcium carbamate was used throughout. 2 Those values were taken on the basis of estimation in Nos. ] and 2. 3 In Nos. 1, 2, and 3 no barium chloride was used. a clear saturated solution of baryta was added, and the weighing bottle corked and placed on the shaking apparatus. The rest of the process was carried out as described above. Into another weighing bottle were placed the above reagents and serum, but no carbamates. 1 The weighing of such a small amount is naturally subject to considerable inaccuracy. THE INFLUENCE OF FEVER ON THE REDUCING ACTION OF THE\ANIMAL ORGANISM. BY Gy A. BPERTER: N a previous paper ! I described the effects of a depression of the body-temperature on the reduction of methylene blue to leuco- methylene blue, and laid special emphasis on the impaired reducing action which is demonstrable in the muscles and gray substance of the central nervous system. Since making these observations I have carried out other experiments of a similar character upon animals in which the temperature had undergone an elevation. The outcome of these experiments appears of sufficient interest to justify me in placing on record some of the details. The main result of the experiments which form the subject of the present paper is the demonstration of a greatly increased reducing action during fever by means of suitable color indicators (methylene blue or indophenol). The result is, as was anticipated, the opposite of that obtained by the experiments concerned with subnormal tem- peratures. It was found, however, that the acceleration of the reduc- ing action which is occasioned by a rise of 3° or 4° C. gives indications of being as pronounced as the diminution in reducing action that is demonstrable by a depression of temperature through 8° or 10° C. Hence the study of cell-reductions in fever is a peculiarly satisfactory field for the application of intravital colorimetry. The contrast between the fever animals and the controls can easily - be made striking. The best results have been obtained under con- ditions slightly different from those recommended for the study of the effects of cold. In most of the experiments upon the action of fever, the rabbits have been infused intravenously with a 25 per cent solution? of methy- 1 HerTER: This journal, 1904, xii, p. 1'28. 2 At first, the percentage of salt in the solution was 0.85 per cent, but as this concentration salts out some of the dye at low room-temperatures the sodium chloride was reduced to 0.4 per cent. 457 458 C. A. Herter. lene blue at the rate of I c.c. per minute, and from 50 to 75 c.c. have usually been introduced. Instead of waiting five or ten minutes after the close of the infusion before examining the organs, the animals have been killed (by ventricular incision), without delay, at the end of the infusion. These changes have been found advisable owing to the rapidity of reduction at fever heat. Two different methods of elevating the temperature have been employed. In one set of cases the animal was enveloped in cotton batting, and an incandescent electric light was passed over the sur- face until the temperature of the body had been gradually elevated to the desired point, where it was without difficulty maintained during the infusion. In another set of experiments, the temperature was raised by means of hog-cholera infection. The infection was induced by means of cultures of hog-cholera of known virulence furnished me by Prof. Theobald Smith. One of the cultures sent me by Professor Smith gives rise to a temperature of 42° or 43° C. on the fourth or fifth day after inoculation with ss}oq c.c. of a twenty- four hour bouillon culture. Higher temperatures are observed later in the course of this fatal infection, but there appears to be no advan- tage in employing them in connection with the present investigation. In order to illustrate the effect of elevation of temperature upon reduction in the organism, I shall give two typical protocols, one from an experiment in which the rise was caused by the external applica- tion of heat, and another in which the fever was of infectious origin. An examination of the results recorded in Experiment 1 plainly shows the accelerating effect of elevation of temperature on the reduction of methylene blue by various types of cells. The differ- ence in color between the corresponding parts in the two animals is the expression of the difference in reducing activity. The first evi- dences of this difference appeared during the infusion of the dye in the readily visible pectoral muscles, which were bared for purposes of observation. After the close of the infusion the animals were killed promptly enough to enable one to obtain an idea of the state of reduction at the end of life. The differences in color were especially striking in the brain, skeletal muscles, heart, spleen, pancreas, and liver. In all these situations reduction was more intense in the superheated animal than in the control, and, excepting the kidney, it may be stated here that this was the case not only in this experiment, but also in all the experiments of the sort that were conducted. It should be noted also that the warmed rabbit secreted less blue by Reducing Action of the Animal Organism. 459 the urine and by the stomach than did his fellow, and further, that the blood in the warmed animal contained less blue than the control. These are features of interest in relation to the actual distribution of EXPERIMENT 1. NoRMAL Rasgit (CONTROL). Temp. 389-399 C. Weight, 1565 gms. Received intravenously 42 c.c. of 0.25 per cent methylene blue solution in 42 minutes. Killed about 3 minutes after infusion’s close. Muscles: pectorals bluish during life, quick reduction after death, the muscles of back of neck, however, remaining blued for some time. Application of H,O, shows considerable leuco-blue in muscles. Heart found unblued (ventricles). Brain: blue (turquoise). Spleen: blue. Pancreas: moderately blue. Suprarenal: colorless, does not blue on addition of H,Ox,. Kidneys: blue. Urine: scanty, deep blue. Liver: purple, contains unreduced blue. Liver pulp contains considerable paired leuco-methylene blue. Bile: moderate amount ; contains consider- able methylene blue; paired leuco- methylene blue, moderate amount. Blood: considerable methylene blue; con- siderable leuco-methylene blue ; paired leuco-methylene blue, scanty. Stomach: much blue on surface of contents. SUPERHEATED RABBIT. Temp. 42°-43° C. Weight, 1510 gms. Received intravenously 42 c.c. of 0.25 per cent methylene blue solution in 42 minutes. Killed about 3 minutes after infusion’s close. Muscles: pectorals quite colorless during life. Muscles of back of neck com- pletely reduced. Application of H,O, shows considerable leuco-blue in muscles, more than in control muscles. Heart found unblued (ventricles); contains considerable leuco-blue, more than control. Brain: unblued. Oxidation with H,O, shows that the brain contains more leuco-blue than the control. Spleen: unblued; on oxidation with H,O, blues to about the color of control. Pancreas: pale blue. Suprarenal: colorless; blues, cortex and medulla, on addition of H,Ox. Kidneys: uncolored. Urine: scanty, moderately blue (greenish) Liver: red; complete reduction of blue. Liver pulp contains considerable paired leuco-methylene blue. Bile: scanty; contains little methylene blue. Moderate amount of paired leuco-methylene blue (slightly more than control). Blood: little or no methylene blue; con- siderable leuco-methylene blue; paired leuco-methylene blue, scanty. Stomach: little or no blue on surface of contents. dye in the body. The criticism might be made that the differences in color in the two animals were possibly referable to variations in the distribution of the dye rather than to unequally energetic reduc- tion. In answer to such a criticism it might be said that if one 460 C. A. Herter. organ or a group of organs in the warmed animal received less blue than the corresponding organ or group of organs in the control, there would necessarily (since the infusions are of the same magnitude) be a compensating difference in the opposite direction, in some other part of the body, whereas in reality the color contrasts between the organs of the two animals were so widespread that they may be designated universal. Further and more convincing evidence is, however, not wanting to show that the color differences are not dependent on unevenness of distribution. Experiments made with eosene (which undergoes no reduction or other demonstrable chemi- cal change in the body) showed that the distribution of this dye is essentially the same in the organs of the normal rabbits as in the corresponding organs of animals in which the temperature has been raised, — the conditions of infusion having, of course, been kept the same in these experiments. Finally, the ease with which methylene blue undergoes oxidation gives us a direct method of determining the actual distribution of the dye in the organism, at least in the case of the central nervous system and the muscles. If we have any doubt whether the brain has taken up as much dye as its fellow, we have only to pour upon a freshly cut surface a solution of hydrogen perox- ide, when the leuco-blue or reduced blue is rapidly oxidized to the dye itself. By means of this method, it was easy to show that the brain and muscles of the warmed animal held not less but more dye than the corresponding parts of the normal rabbit. When we come to the other organs, especially the liver and kid- neys, the conditions are not so simple, for they are complicated by the occurrence of a synthesis or pairing of the methylene blue with some unknown constituent of the cells. The substance thus formed I have called paired leuco-methylene blue, or more briefly “ paired substance.” The important characteristic of this substance in the present connection is that it fails to be oxidized (in neutral or alka- line medium) to methylene blue, and hence escapes recognition in the simple process of oxidation. The dye can, however, be unmasked by boiling the organ pulp with a dilute acid — acetic being perhaps the best for this purpose. Until I recognized the pairing process in the liver and elsewhere, it was impossible to account for the disappearance of the infused methylene blue, since it was evident that the sum of the simple leuco-blue and the unreduced blue was far from being equal to the amount of dye infused. It would be incorrect to give the impression that the total amount Reducing Action of the Animal Organism. 461 of methylene blue in the febrile organs was always or even usually in excess of the amount in the corresponding organs of the controls. Even in animals of the same temperature, infused under the same conditions, the brain and muscles (structures holding little or no EXPERIMENT 2. NorRMAL RABBIT (CONTROL). Temp. 39° C. Weight, 1450 gms. Infused intravenously 31 c.c. 0.25 per cent methylene blue solution, 1 c.c. per minute. Killed at close of infusion. Muscles: blue during life. At death, blue was rapidly reduced. With H,O, mod- erately blue. Heart: Ventricles blued. Brain: almost completely reduced; blues deeply on exposure to air; with H,O, very deep blue. Spleen: blue. Pancreas: not noted. Suprarenals: not noted. Kidneys: unblued; considerable leuco-blue in cortex. Urine: none. . Liver: bluish; leuco-blue, considerable; pulp yields considerable paired leuco- methylene blue. Bile: quantity moderate; some methylene blue; some paired methylene blue. Blood: serum blue; very little leuco-blue ; considerable paired leuco- methylene blue. Stomach: contains considerable blue. INFECTED RABBIT (HOG CHOLERA). Temp. 40.5° C. (4 days after inoculation).1 Weight, 1435 gms. Infused intravenously 31 c.c. 0.25 per cent methylene blue solution, 1 c.c. per minute. Death at close of infusion. Muscles: no bluing during life; with H,O, less blue than control. Heart: ventricles unblued. Brain: completely reduced, but blues only slightly on exposure to air; with H,O, blues somewhat less deeply than control. Spleen: red. Kidneys: blue; papillae and medulla blue on exposure to air. Urine: none. Liver: gray; contains little or no leuco- blue; nearly all blue exists as paired leuco-methylene blue; quantity about same as in control. Bile: scanty; no methylene blue (as such) ; very little paired leuco-methylene blue (z. e. no secretion). Blood: serum uncolored; little or no leuco- blue. All dye is present as paired leuco- methylene blue. Stomach: contains very little blue. 1 The temperature of this animal had been higher, but it fell when the animal was placed on the holder, as is often the case. paired substance) often show, on oxidation with hydrogen peroxide, moderate inequalities in the content of dye. In the case of experi- ments made at an elevated temperature, the greater amount of dye is found sometimes in the muscles and brain of the control, sometimes in the structures of the superheated mate. The evidence of increased reduction is, of course, especially pronounced in those cases where the 462 C. A. Herter. brain and muscles not merely are colorless, but contain more total dye-stuff than their fellow-organs. In essential respects the conditions in Experiment 2 after infusion resembled those found in Experiment 1; that is to say, the various parts, including brain, muscles, heart, spleen, and liver, showed the presence of less blue in the febrile rabbit than in his fellow. The differences were, however, less pronounced than in Experiment 1, and this is hardly surprising, inasmuch as the disparity in temperature was only 2.5° C.; whereas, in Experiment 1 the inequality amounted to probably not less than 4° C. In other observations in animals in- fected with hog cholera, higher temperatures were obtained, and in such instances the results were indistinguishable from those seen in animals whose temperature had been raised to an equally high level by external application of heat. There is usually a distinct difference between the color of the blood of the febrile animal and that of the control, the former containing less methylene blue as such and more leuco-methylene blue. The bile in the gall-bladder of the febrile animal is almost always scanty, and contains less methylene blue than the bile from the control. The stomach also shows a diminished secretory activity when the tem- perature is elevated, for little or no blue finds its way into the interior of this organ, whereas in the control animals the food is found cov- ered by a layer of mucus mixed with unreduced blue. Once, how-. ever, I found these conditions reversed. As regards the quantity and character of the urine secreted during the infusion the results are extremely variable. I have generally found that the febrile urine contains rather less methylene blue and more leuco-methylene blue than the urine from the normal control, but there may be more blue in the febrile urine. It seems probable that the wide variability noted in wholly normal animals with the same temperature affords a sufficient explanation of these irregularities. Occasionally, as in Experiment 2, there is no secretion during the infusion. In general, it can be stated that the quantity of dye recovered in the urine is too small to exert a material influence on the quantity or distribution of that which remains in the organism. The observation that methylene blue is capable of serving as an admirable indicator of the acceleration of reduction resulting from an elevation in temperature naturally suggested that the various kinds of cells in which the reducing powers had been watched during life might be advantageously studied in vitro in respect to this activ- Reducing Action of the Animal Organism. 463 ity. The prospect of being able to study individually the properties of different tissues, unhampered by the uncontrollable and unmeasur- able interactions characteristic of the living organism in its entirety, made it appear worth while to seek a method of conducting such experiments under conditions permitting a measurement of the pro- cesses in question. ) After some unsuccessful trials, the following method of measuring approximately the reaction velocity of reduction was adopted. The tissues to be studied (liver, kidney, muscles, etc.) were taken from a dog or rabbit (which had been bled) and subdivided in an ordinary meat machine. A finer degree of subdivision was secured by tritura- tion with fine sand. Definite weights of tissue thus prepared and mixed with sand were placed in thin walled capacious test-tubes of hard glass, one and one-quarter inches in diameter, to which a fixed volume of distilled or tap water was added. For example, to 2 or 4 ems. of triturated liver was added usually 25 gms. of water. To this mixture was added at the proper time I c.c. of a weak methylene blue solution (0.025 per cent in distilled water) at the proper tem- perature. In order to secure better contact of the dye with the par- ticles of tissue, a constant stream of washed and neutralized nitrous oxide gas was passed through the mixture.! As our gas liberated a very slight quantity of oxygen, its action, aside from a mechani- cal one, must have been to slightly retard reduction. Carbon dioxide was abandoned because of the disturbing effect exerted by its acid properties. The temperature within the tubes was easily regulated, and there was no difficulty in keeping their contents in a practically anaérobic state. In order to insure the rapid and thorough mingling of the dye with the remaining contents of the tube, it has been found helpful to in- ject the methylene blue directly into the rubber delivery tube through which the gas enters the lowest part of the tube. The reduction tube is prepared for action by allowing the nitrous oxide to bubble actively through it for five minutes before the dye is introduced. As regards the endpoint of the reaction, it must be said that this is not always as sharp as could be desired, and that it seldom happens 1 A drawing of the apparatus will be given in a subsequent paper. Each tube is fitted with a rubber cork with three openings, one for the tube delivering the gas, one for a thermometer, and a third for the exit of the gas. Two tubes are generally operated simultaneously. The desired temperature within the tubes is secured by their immersion in a beaker of water. 464 C. A. Herter. that the original color of the tube is regained, even through prolonged action of the tissues on the blue. For this reason, it has been found best to take the reading when the last trace of the green-blue dis- appears from the mixture in the tube. Where control observations are being made, it is usually not difficult to fix upon an arbitrary endpoint which is the same for both tubes. By means of the method thus indicated, it has been possible to study the influence of many conditions and substances upon the velocity of reduction, including the action of acids and alkalies, the ions of neutral salts, the effects of colloidal solutions,! etc. At present reference will be made only to a few typical observations on the in- fluence of temperature. The following are a few examples of the influence of temperature and mass of tissue (rabbit’s liver) : 1 gm. liver + 1 c.c. blue solution + 25 c.c. H,O at 38° C. reduced in 243 minutes. 1 gm. liver + 1 c.c. blue solution + 25 c.c. H,O at 43° C. reduced in 18 minutes. 2 gms. liver + 1 c.c. blue solution + 25 c.c H,O at 38° C. reduced in 3} minutes. 2 gms. liver + 1 cc. blue solution + 25 c.c. H,O at 43° C. reduced in 2 minutes. Material from the liver of another rabbit gave the following results: 1 gm. liver + 1 c.c. blue solution + 25 c.c. H,O at 38° C. reduced in 8 minutes. 1 gm. liver + 1 c.c. blue solution + 24 c.c. H,O + 1 c.c. 0.85 per cent NaCl solution at 38° C. reduced in 8 minutes. 1 gm. liver + 1 c.c. blue solution + 25 c.c. HO at 43° C. reduced in 4 minutes. 1 gm. liver + 1 cc. blue solution + 24 c.c. HJO + 1 c.c. 0.85 per cent NaCl solution at 43° C. reduced in 5 minutes. 2 gms. liver + 1 c.c. blue solution + 25 c.c. H,O at 38° C. reduced in 1} minutes. 2 gms. liver + 1 c.c. blue solution + 25 c.c. H2O at 43° C. reduced in 45 seconds. With larger quantities of triturated tissue, say 4 or 5 grams of liver, reduction may occur so rapidly that it is difficult to measure the difference in velocity at 38° and 43° C. At one time it seemed to me that our experiments indicated a greater reaction velocity than is characteristic of simple chemical 1 An interesting field for observation is a comparison of the reducing action of different structures in the same organism, and of the corresponding structures in different groups of animals. In the present experiments the great activity of liver and kidney as compared with brain and muscle is an obtrusive feature. The superiority of the dog’s liver over the liver of the rabbit in regard to the power of reduction can be easily shown. The action of poisons has not yet been taken up with this method, except to a limited extent. Reducing Action of the Animal Organism. 465 reactions; but at present it is doubtful if that view can be main- tained. In order to gain information as to the influence of tempera- ture on the velocity of reduction in the absence of organized material, the tubes were filled with a solution of 10 c.c. of (approximately) 3.71 percent glucose solution, 1 c.c. methylene blue solution, 2 c.c. »% sodium hydrate solution, and 12 c.c. water. At 38° C., reduction occurred in nine and one-quarter minutes ; at 43° C. in six minutes ; at 48° C.in four minutes. The results thus derived are apparently of the same order as those obtained with living tissues. It is clear that many factors influence the velocity of reduction by living animal tissues. In some of the above experiments the effect of a small amount of sodium chloride is seen,! and it can be shown that there is a high degree of sensitiveness to the presence of hydrogen and hydroxyl ions. The differing activity of the livers of different indi- viduals in vitro is a striking thing, and reminds one of the individual differences noted in the course of intravital infusions. A factor which may become disturbing in test-tube experiments is the postmortal decline in activity of the cells. In the case of muscle, especially rabbit muscle, this is rapid. The liver, however, usually retains a high grade of reducing activity for several hours after death. It is believed by the writer that the demonstration of the action of fever on the organism by the intravital method described in these pages will not only prove serviceable in the study of the pathology of fever, but will afford a highly instructive object lesson in class work. Probably the method of studying reduction in vitro will prove of considerable use in the analysis of the many factors that affect the process of reduction. I hope before long to report on the behavior of different kinds of living tissue under physiological and pathological conditions, I wish to acknowledge the valuable assistance I have received from my laboratory assistant, Mr. Edward O’Brien, in the conduct of these experiments. 1 At 38° C. sodium chloride, in larger amount than in the experiment given above, was found to accelerate reduction, but at higher temperatures it was fre- quently observed to exert the opposite influence. THE PRODUCTION OF FAT FROM PROTEID BY eis BACIELYS: PYOCYANEUS: By S. P. BEEBE AND BB. H. BUXGEORE [From the Department of Experimental Medicine in the Cornell University Medical College, New York City.] HE production of fat from proteid has been among the most interesting of physiological and pathological problems. In this paper are given the conditions under which such a production due to the growth of a microorganism was found to take place. That certain microorganisms form fat is admitted, notably the bacillus tubercu- losis, the intracellular fat of which has been the subject of recent studies by Bulloch and Macleod,! and Levene.? In the present in- stance it is the belief of the writers that the conditions under which the fat is formed are somewhat more definite than in the previous instances. A stock culture of the bacillus pyocyaneus, of unknown origin, was observed to form a thick wrinkled pellicle on the surface of the agar or broth in which it was growing. On making a hanging drop prepa- ration from the pellicle, it was seen to consist of masses of bacilli among which bundles of long needle-shaped crystals were frequent. The only paper in the literature which could be found referring to the formation of crystals by the bacillus pyocyaneus was a note by Dorset in the Centralblatt fiir Bakteriologie (Vol. 20, 1896, p. 217). Agar cultures of Dorset’s bacillus, which was isolated from a guinea- pig killed by inoculation of oleomargarine, formed small needle-shaped crystals which penetrated beneath the surface of the agar, —tests proving conclusively that these crystals were not due to mere drying- out of the medium. Chemically, the crystals were shown to be phosphates, principally of calcium with. traces of magnesium, and were apparently ‘due to a separation of the phosphates normally present in the medium, caused 1 BULLOCH and MACLEOD: Journal of hygiene, 1904, iv, pp. I-Io. 2 LEVENE: Journal of medical research, 1904, xii, pp. 251-258. 466 The Production of fat from Proteid. 467 by the specific action of certain products of the bacillus.” The dis- position of the crystals in this case was very different from that in our cultures. Instead of crystals being formed in the agar, we find them only in the pellicle, in intimate connection with the bacilli. How- ever, in view of Dorset’s statements, we subjected our crystals to a preliminary test for phosphates with negative results. Further examination showed that: 1. The crystals are soluble in alcohol, ether, chloroform, and to some extent in petroleum ether. 2. They give staining reactions for fats with osmic acid and sudan III. 3. The alcoholic solution on saponification affords a soap from a solution of which fatty acids can be precipitated by a mineral acid, and then extracted by shaking out with ether. The photograph shows the similarity of the crystals with those of certain of the higher fatty acids, and the conclusion was soon reached that they must be of a fatty nature. The next step was to prepare the fat in sufficiently large quan- tities for chemical examination; and, after trial of several methods, the following was adopted as being the simplest and most effective. Two litres of meat-extract broth with 3 per cent peptone are pre- pared at a time and distributed into eight 1000 c.c. Erlenmeyer flasks, 250 c.c. in each, a large surface for the formation of the pel- licle being thus secured. After sterilization, the flasks are inoculated from an agar culture and allowed to stand, without being disturbed for two or three weeks, in the incubator. At the end of that time a thick wrinkled pellicle has formed on the surface, and with a little care the underlying fluid can be decanted off, leaving ‘the pellicle lying intact at the bottom of the flask. After drying for two or three days, each flask is filled with chloro- for 1 until the pellicle is entirely covered, and allowed to stand for twenty-four hours with occasional shaking. Usually two flasks at a time were so treated, the chloroform being then poured into the next 468 S. P. Beebe and B. H. Buxton. two. After all the flasks have been thus treated, the chloroform, which has assumed a yellowish tinge, is filtered to free it from the fragments of the pellicle remaining, consisting of masses of bacteria. The chloroform is then mostly recovered by distillation, and the dark-brown concentrated residuum evaporated to dryness at 100° C. in a current of hydrogen gas, the flask cooled over sulphuric acid and weighed. The dried residue, of a yellow-brown color, is dissolved in ether, and the solution kept in a stoppered bottle for future use. In this way we collected about 10 grams of fatty substances from 30 litres of broth culture, —an amount considered sufficient for pre- liminary examination. The quantity of fat obtained from each litre was about 0.3 to 0.4 gram after we had once hit upon the best method of procedure. Chemical examination.— The chemical examination of the fat has been rendered difficult because of the small quantity at our disposal. The following fat constants were determined: (1) Melting point, 70°. (2) Acid number, 47. (3) Saponification number, 94. (4) Iodine number, 70. (1) Melting point.— It must be noted that the fat does not have a sharp melting point, nor does it, even at a much higher temperature than 70°, become fluid as the ordinary fats do, but softens to a thick, syrupy condition. (2) Acid number. — This number gives the milligrams of potassium hydroxide required to neutralize r gram of the fat. Forty-seven, as the acid number, indicates that a considerable portion of the substance is free fatty acid. The fat from the tubercle bacillus has an acid value of 23. The ordinary fats vary from 0.3 to 9 in acid value. (3) Saponification value.— This gives the number of milligrams of potassium hydroxide required to saponify 1 gram of the fat. The value in this instance is low, indicating the presence of a considerable amount of unsaponifiable substance. The tubercle bacillus fat had a saponification value of 60.7, and was also found to contain a large amount of unsaponifiable substance. The saponification value of a few common fats may serve for comparison: cocoa butter, 192; mutton tallow, 195; horse fat, 195; goose fat, 195. (4) Iodine number.— The iodine number gives the percentage of iodine which a fat may take up. The unsaturated fats under certain conditions readily take up definite amounts of iodine so that this number serves as a convenient means of identifying them. The number, 70, is somewhat higher The Production of Fat from Protedd. 469 than that of most solid fats. In animal fats, the iodine is absorbed by the oleate radical, but there are other unsaturated fatty acids which absorb even more than oleic acid. A considerable portion of the fat was saponified by one of the common methods, with alcoholic potash, and the resulting solution of soap was evaporated to dryness after adding 5 grams of sodium chloride. The dried mass was powdered and extracted with ether. After evaporation of the ether a waxy substance was left which was found to weigh 78 per cent of the fat used “for saponification. In order to determine if a portion of this wax could be saponified by more drastic methods, it was dissolved in absolute alcohol, and the solution slowly saturated with metallic sodium. ‘The alcoholic solu- tion was boiled for four hours, evaporated to dryness, the residue taken up in water and shaken out with ether. None of the wax had been saponified. The wax does not dissolve readily in ether, and scarcely at all in petroleum ether; alcohol and chloroform are its best solvents. It gives none of the color reactions for cholesterin. It stains with osmic acid and absorbs iodine; its iodine number was found to be 60.6. When the wax is boiled with acetic anhydride, it dissolves readily, but crystallizes out to some extent on cooling. Its behavior in this respect is like that of an alcohol. The soap from the saponification of the original fat yielded a fatty acid having a low melting point, 41°. (The fatty acid was probably not pure enough to have much reliance placed on this figure.) It absorbed iodine and stained with osmic acid. The question now arises as to the origin of the fats. Sugar-free meat broth and sugar-free extract broth with varying quantities (1 to IO per cent) of Witte’s peptone were prepared, and in these the crystals were readily formed. Free sugar, therefore, could not be the source of the’fat. On sugar-free solutions of nutrose and casein there was no development of the pellicle, so it might be supposed that the fat is derived from the carbohydrate group contained in the albumoses of the peptone (so-called), but the amount formed is so great (about 0.35 gram in 1000 c.c. of 3 per cent peptone broth, or in other words over 10 per cent of the peptone used) that the carbohydrate nucleus of the synalbumose could hardly account for all of it. That the crystals could not be formed from fat pre-existing in the broth was made evident by extracting 1000 c.c. of the latter with ether. The merest trace of ether-soluble substances could be found. 470 S. P. Beebe and B. H. Buxton. Again 50 grams of peptone extracted with chloroform afforded negative results. During the growth of the bacillus the medium becomes very alka- line, the alkalinity being due almost entirely to the presence of free ammonia. Of the total nitrogen found in one case, 33.7 per cent was split off by boiling with magnesium oxide. Obviously during the growth of the bacillus there is much ammonia split off from the proteids, and the resulting products may become oxidized to fats. It seems probable, therefore, that the fat is formed, at any rate partly, by oxidation of fragments of the albumoses and peptones apart from any carbohydrate nucleus they may contain. That it is an oxidation process appears more than likely from the fact that the crystals are formed solely in the surface pellicle. Again under anaérobic conditions the bacilli grow sparingly, and no trace of pellicle or fat crystals is formed. In addition to the fats a considerable amount of a mucinous sub- stance is formed during growth. Although this substance gives cer- tain reactions for mucin, such as precipitation by acetic acid and solubility in alkaline solution, yet, strange to say, we have been unable to detect the presence in it of a reducing substance except in mere traces, and that only on complete destruction of the albumin molecule after prolonged boiling in weak acid (2.5 per cent H,SOs,). Of those who have studied the mucin of the bacillus pyocyaneus, Rettger! found it yielded a compound which reduces Fehling’s solu- tion, while Charrin and Desgrez? obtained a reducing substance in two cases, a third yielding negative results. Since the mucinous substance, therefore, does not appear to call for any carbohydrate, we may relegate all of the carbohydrate nuclei of the albumoses to the formation of the fat; but even so, as already remarked, there could not be sufficient to account for all of it. We expect to study more closely the ammoniacal, mucinous, and wax-like substances of our cultures and make them the subject of a future paper. 1 RETTGER: Journal of medical research, 1903, p. 102. 2 CHARRIN and DESGREZ: Comptes rendus de l’académie des sciences, 1898, CXXviii, p. 596. FURTHER EVIDENCE OF THE NERVOUS ORIGIN OF THE HEART-BEAT IN LIMULUS. By “A. J. CARLSON. [From the Hull Physiological Laboratory, University of Chicago.] N a previous paper ! I have shown that the heart of Limulus contra- dicts the myogenic theory, bothas regards the origin of the beat and as regards the tissues concerned in co-ordination or conduction in the heart. The experiments reported in that paper were per- formed at the Marine Biological Laboratory, Woods Hole, during the early part of June, but the work was continued till September, with the result of not only completely verifying the earlier observations, but also of obtaining data which place related questions of heart- physiology in a new light. This refers especially to the nature of cardiac inhibition and the nature of the action of drugs on the heart. As a preliminary to taking up those questions, the following observa- tions may be recorded as further evidence of the nervous origin of the heart-beat: I, DEGREE OF AUTOMATISM OF THE DIFFERENT REGIONS OF THE HEART. It was stated in the previous paper that any segment of the heart will continue to beat rhythmically for some time after being isolated from the adjoining segments by transverse lesions, provided the nerve-cord is left intact in the segment. It was noted, however, that the three anterior segments may not beat spontaneously after such lesions. In the experiments during July and August spontaneous contraction in either of the first three segments, when isolated, were observed only in four out of sixty hearts tested. This refers to hearts from adult specimens. The hearts from young specimens worked on shortly after being brought into the laboratory usually exhibited a greater degree of automatism of the anterior or aortic end, just as did 1 CARLSON: This journal, 1904, xii, p. 67. 471 472 AF Carkson. the hearts of adults worked on in the early part of June. This slight difference between the specimens worked on in June and those used later in the season was probably due to a poorer state of nutrition and general condition of the animals later in the summer. Theadult specimens are obtained in abundance only during the breeding sea- son in May, when they are collected and kept during the summer in aquaria anchored inthe bay. The animals keep well in this confine- ment; but as they are not fed, or do not take food, it is evident that the condition of nutrition grows steadily poorer as the season advances. But even in young specimens, or in adult specimens in good condi- tion, the greatest automatism is exhibited by the posterior two-thirds, or more correctly by the middle third of the heart. The anterior end of the heart is the aortic end, the posterior third or two-thirds of the heart correspond to the venous end. We can thus perform the Stan- nius experiment on the Limulus heart. The heart is prepared by carefully freeing the lateral nerves and the nerve-cord from the heart in the third or fourth segments, so that they may be severed trans- versely without injury to the heart-muscle. The lesion of the nerves and the nerve-cord in this region does not affect the rhythm of the posterior or venous end of the heart. Anterior to the lesion the rhythm ceases temporarily in every case, and it may or may not be resumed, usually the latter. Conduction takes place in the nervous elements, hence any rhythm anterior to the lesion of the nerves and the nerve-cord is not due to transmission of the contraction from the posterior end of the heart, but must be generated in the heart anterior to the lesion. Thus we find in the heart of Limulus a condition similar to that in the vertebrate heart, the venous end of the heart exhibiting the great- est automatism, the aortic end the least or no automatism. On the myogenic theory, this condition in the heart of Limulus would be ex- plained as in the vertebrate heart by the subsidiary hypothesis that the muscle at the venous end retains more of its embryonic character, and is therefore more automatic; but when this hypothesis is tried ‘in the case of the Limulus heart, it is found wanting, for after extir- pation of the nerve-cord or ganglion on the dorsal side of the heart, the heart-muscle exhibits the same degree of automatism in all regions of the heart, that is, no automatism at all. In the distribution of the ganglion cells, we find still another sim- ilarity between the heart of Limulus and that of vertebrates, namely, Nervous Origin of the Heart-Beat in Limulus. 473 that the regions of the heart exhibiting the greatest automatism have the greatest number of ganglion cells. In the Limulus heart, the ganglion cells are collected in the nerve-cord on the dorsal side of the heart. The nerve-cord is thickest in the middle third of the heart, that is, in the fifth, sixth, and seventh segments, which means a greater number of ganglion cells in this region. In the first three segments, the nerve-cord is very slender and made up for the most part of nerve-fibres. There are some individual variations in this regard. Thus in two specimens I found an unusual ganglionic en- largement on the nerve-cord at the level of the first pair of ostia. But the few ganglionic cells in the nerve-cord of the anterior seg- ments can in most cases not maintain rhythmic activity when sep- arated from the region of the cord having the greatest automatism, just as the tortoise ventricle, when severed from the auricles, does not beat except when stimulated artificially! In all probability the ganglion cells (or at least some of them) situated in the aortic end of the heart are similar in function to the ganglion cells of the respiratory centres in the spinal cord of vertebrates in not being active under normal conditions except when being stimu- lated by the impulses from the centre of greater automatism. There is still another similarity in the distribution of the ganglion cells in the heart of Limulus and in the vertebrate heart, namely, in their position with reference to the myocord. In the vertebrate heart, they are situated in the main on the surface of the myocard, and this is also their position in the Limulus heart. The absence of automatism in the isolated anterior segments of the Limulus heart may thus be due, in part, to the absence of impulses from the cells of the cord having the greatest automatism, but it is also due to the severing of the nerve fibres which pass from the nerve- cord in the middle of the heart directly to the muscle of the anterior segments. These fibres reach the muscle along the lateral nerves, as well as in the nerve-cord itself, as can be shown by the following ex- periments. If the nerve-cord is cut transversely in the third segment, and the portion of the cord anterior to the lesion extirpated, this end of the heart continues to beat in perfect synchrony with the rest of the heart, but the strength of the contractions is reduced. This rhythm is due solely to the impulses from the cord posterior to the lesion reaching the muscle in the lateral nerves, for lesion of either nerve diminishes still further the strength of the beats, and lesion of 1 MARTIN: This journal, 1904, xi, p. 103. 474 A. F. Carlson. both nerves stops the rhythm completely. When both lateral nerves are cut transversely in the third segment, leaving the nerve-cord in- tact, the amplitude of the beats anterior to the lesion is similarly re- duced ; but in this case it is difficult to determine whether the rhythm now maintained is due to impulses from the nerve-cord behind the lesion passing to the muscle directly, or to the activity of the ganglion- cells situated in the’ nerve-cord of the anterior segments. Both factors are probably present. By a similar series of lesions, it can be shown that nerve-fibres from the ganglion in the fifth and sixth seg- ments pass in the lateral nerves and the nerve-cord to the muscle of the posterior segments of the heart (seventh, eighth, and ninth seg- ments). We must call in the aid of the histologists to determine the mechanism effecting the co-ordination between the impulses passing directly to the muscles in all parts of the heart from the middle region of the cord, and impulses reaching the same or adjoining muscle cells from the ganglion cells situated in the same segment. The co-ordi- nation of the Limulus heart is to be compared to the co-ordination of the auricle or the ventricle, rather than with the co-ordination of the whole heart in the vertebrates. Assuming that the physiology of the ganglion cells in the vertebrate heart is the same as in the heart of Limulus, the mechanism of co-ordination of the latter is in all prob- ability similar to that effecting co-ordinated activity of the ganglion cells scattered all over the myocord of the vertebrate auricle. There is some individual variation in the course of the fibres from the nerve- cord of the middle region to the muscle of either end of the heart, as determined by the effect of the lesions on the strength of the contrac- tion. More fibres may pass in the left than in the right lateral nerve, or vice versa. It is rare that the majority of the fibres pass in the nerve-cord. ; In all experiments involving the determinations of the rate and the strength of the beats, the ordinary graphic method was employed. For this purpose it is preferable to work on the excised heart. A recording lever may be adjusted to the heart zw sztw, but removing the heart from the body does not affect the rhythm, and in either case the heart is empty, the mere opening of the pericardial cavity pre- venting the entrance of blood into the heart. The contraction of the heart-muscle diminishes the diameter of the heart without chang- ing its length, as the muscle cells are arranged circularly. For graphic registration, one needs only to secure one side, or rather lat- eral angle, in any region of the heart to a fixed support and connect Nervous Origin of the Hleart-Beat in Limulus. 475 the opposite side with a recording lever. In case any of the first four segments are used for recording the contractions, the attachment of the heart to the support and the lever is most conveniently made by glass hooks in the lateral arteries in the manner shown in Fig. 10. When it is desired to make use of any of the middle or posterior seg- ments, the hooks are fastened to the muscle directly. A record thus obtained represents the contractions of one, or part of one, seg- ment only. The heart of large specimens is about 12 cms. long. The parts of the heart not connected with the recording apparatus were supported independently so that their contractions did not influence the lever. When the first or second segments were used for contraction, the posterior part of the heart was usually placed in a shallow dish filled with plasma, and thus supported on a level with the recording segment. MAQIMOWONIOAIDAMJANDNAARNOOMUNNGSN sSNA ANA NNNNsINUAnns sn NN NNR x x’ FicurrE 1]. — Tracing of contractions of first segment. -X, beginning of severing the con- nectives between the nerve-cord and the lateral nerves in the last segment, the lesion being continued forwards till at X’ the nerve-cord is isolated up to the third segment. Diminution of contractions of the anterior segments following the lesion. Continued activity of the ganglion cells, although the nerve-cord is isolated from the greater part of the heart. Tracings illustrative of the effect of the lesions of the nervous complex described above are reproduced in Figs. 1 to 3. The record in Fig. 1 is from the contraction of two anterior segments, the hooks being fastened in the first pair of lateral arteries in the manner shown in Fig. 10. At X the beginning of the lesion of the connectives between the nerve-cord and the lateral nerves was made in the seventh segment and proceeded anteriorly until at X! the nerve-cord was completely isolated from the lateral nerves up to the third seg- ment. In this particular experiment that part of the cord was at the same time almost completely isolated from the heart-muscle, as the lesion of the nerves was effected by cutting through both the nerve plexus and the dorsal wall of the heart at either side of the cord and at a distance of 1 mm. from it. It is not possible to sever all the connectives in any other way. The nerve-cord must not be touched by the instruments, as that causes acceleration and inco-ordination of the rhythm. Ifthe operation is made by a pair of sharp scissors, and the nerve-cord not touched, the whole ganglion may thus be isolated 476 A. F. Carlson. from the heart-muscle, save in the first three segments, without greater change in the activity of the ganglion cells than that exhibited by the tracing in Fig. 1. It will be observed in this tracing that as the severing of the connectives between the nerve-cord and the lat- eral nerves proceeds forward, the strength of the contractions dimin- ishes by degrees, until at X' the amplitude of the beats is only about half that of the original. This diminution in the strength of the beat does not imply any change in the activity of the ganglion cells. It is simply due to the severing of the nerve-fibres, and thus cutting off the impulses that pass in the lateral nerves from the ganglion in the middle region of the heart to the heart-muscle of the first two seg- ments. The only change in the activity of the ganglion cells is the slight but steady increase in the rate of the beats. This is, without exception, the first result of the lesion, and in the majority of the preparations the acceleration was much greater than that shown in Fig. 1. In these preparations the acceleration was accompanied by inco-ordination, the beats soon became diminutive and the rhythm ceased. When the operation succeeds, as in Fig. I, a strong and fairly regular rhythm of the first two segments is maintained for from ten to thirty minutes, but inco-ordination sets in before the final ces- sation of the rhythm. Now, this is the condition. The rhythm of the first two segments after severance of the lateral connectives up to the middle of the third segment is still caused by the activity of the ganglion cells situated behind this region of the heart, asis shown by the fact that the cross-section of the cord in the third segment stops this rhythm instantaneously. But behind the second segment the nerve-cord is completely isolated from the heart-muscle, save that portion of the dorsal wall immediately beneath the cord. Yet the activity of the ganglion cells goes on for some time, and, in some preparations, without change in rate or intensity. So far, then, the activity of the ganglion cells does not appear to be primarily depend- ent upon afferent impulses from the heart, although such afferent impulses are probably present under normal conditions, as will be shown later. It may be argued that the afferent impulses from that part of the heart-muscle still in connection with the ganglion are sufficient to maintain the activity of the ganglion cells, or that the severance of the nerves produces sufficient injury to cause a series of impulses in the nerve-fibres, which thus take the place of the afferent impulses present under normal conditions. The acceleration of the activity of the ganglion is probably due to stimulation from the Neroous Origin of the Fleart-Beat in Limulus. 477 unavoidable handling, as well as to the chemical changes taking place at the cut end of the nerve-fibres. The dependence of the rhythm of the anterior end of the heart on the nerve-cord in the middle region of the heart, and the function of the nerve-cord and the lateral nerves in affording paths for these nervous impulses, are clearly shown in the tracing reproduced in Fig. 2. This record is also from the contraction of the first two seg- ments. The posterior part of the heart was supported in a dish filled with plasma. Prior to suspending the heart for graphic registration, the nerve-cord and the lateral nerves were isolated from the heart for a distance of 1 cm. in the third segment, so as to be readily severed with a pair of scissors without injury to the _heart- muscle. At a@ the nerve-cord is cut transversely in the third seg- ment. This is followed by acceleration of the rate and diminution of the strength of the beats. The acceleration is only temporary, the a b c FiGuRE 2.— Tracing of contractions of the first segment. «a, cross-section of the nerve- cord in the third segment; 4, cross-section of the right lateral nerve in the third segment; c, cross-section of the left lateral nerve in the third segment. Nerve-cord and lateral nerves isolated from the heart in the region of the lesions prior to the experiment, so that the lesions did not injure the heart-muscle. diminution in the amplitude of the contractions is permanent. The acceleration is evidently due to the mechanical stimulation in cutting the nerve-cord, as the mere touching the cord produces the same results (Fig. 18). The diminution of the beats is due to elimination of a part of the impulses reaching the anterior end of the heart from the ganglion cells situated behind the lesion. At & the lateral nerve on the right side is cut in the third segment. This lesion does not affect the rate, but reduces the amplitude of the beats still further by eliminating all the motor impulses reaching the heart-muscle of the reacting segments through that nerve. At ¢ the left lateral nerve is severed in the same region. In these lesions of the nerve- cord and the lateral nerves, the heart-muscle has not been touched, but all the nervous connections between the first two segments and the ganglion cells situated beyond the third segment have been severed, and in consequence of this the rhythm of the anterior seg- ments ceases, in most cases permanently. The portion of the heart posterior to the lesion continues its rhythm unchanged, in the same 478 A. F¥. Carlson. manner as do the sinus and the auricles of the frog after severance of the ventricle, except that the rhythm is temporarily accelerated on the cross-section of the nerve-cord, an acceleration clearly due to the mechanical stimulation. The tracing in Fig. 2 shows a rather exceptionally equal distribu- tion of the nerve-fibres in the nerve-cord and the lateral nerves. Frequently the section of one of the lateral nerves produced a much greater diminution of the beat than the section of the opposite nerve. In some specimens the greater part of the fibres appear to pass in the nerve-cord itself, as shown in Fig. 3. The individual variations NU FicureE 3.— Tracing from the first segment. X, nerve-cord cut in the third segment leaving the lateral nerves intact. Unusual diminution of the contractions on cross- section of the cord in that region. can also be made out anatomically, the size and number of the con- nectives between the nerve-cord and the lateral nerves being either relatively few or greater on the one side than on the other. II. THE PHYSIOLOGY OF THE CENTRIFUGAL NERVE-FIBRES FROM THE NERVE-CORD TO THE HEART-MUSCLE. The preparation of the heart for the study of the function of the nerve-fibres passing from the cells in the nerve-cord to the heart- muscle is represented in Fig. 4. The lateral nerves are isolated up to the second segment, the median nerve-cord is extirpated, and the heart severed transversely in the middle of the second segment. By means of hooks in the first pair of lateral arteries, the isolated anterior end of the heart is connected with the lever for graphic registration. The electrodes may now be applied to the lateral nerves 5 to 6 cms. from the muscle. If it is desired to study the influence of the nerve- cord on the quiescent muscle, the nerve-cord may be left in connec- tion with the first two segments, and severed in the fourth segment. After previous dissection of the lateral nerves, lesion of the nerve- cord in the fourth segment is usually followed by complete cessation of the rhythm of the first two segments. In the few preparations in which a feeble rhythm persists, it ceases after a few periods of stimulation of the nerve-cord with the interrupted current. When the nerve-cord is extirpated in the manner shown in Fig. 4, the Nervous Origin of the Heart-Beat in Limulus. 479 musculature of the first two segments remains perfectly quiescent. This preparation thus offers an unequalled condition for the study of the influence of the intrinsic heart-nerves on the quiescent heart- muscle. It was stated in my previous paper that ¢he nerves passing from the nerve-cord to the heart-muscle are of the ordinary motor type. This has now been further confirmed by aid of the graphic method. The heart-muscle responds toa single induced shock applied to either one or both of the lateral nerves, but the nerves possess a relatively low excitability to the induced current. The induced shocks must there- fore be relatively strong to produce a contraction when applied to the FicuRE 4.— Preparation of the lateral nerves and the first two heart-segments for studying the influence of the nerves on the resting muscle. Nerve-cord extirpated. e, electrodes; %, hooks for suspending the preparation for graphic registration; /, lateral nerves. lateral nerves. There is, however, in these experiments, no chance of escape of the current directly to the muscle, as the electrodes may be placed 5 to 6 cms. from the reacting segments. Stimulation of both lateral nerves produces a greater contraction than when either nerve is stimulated singly. This is readily under- stood from the fact that each nerve is confined in its action almost wholly to its own side of the heart, the two nerves thus supplying separate portions of the musculature. It is frequently not possible to augment the strength of the con- traction by increasing the strength of the induced shock applied to the nerves above that which first proves effective, but such an aug- mentation in the amplitude of the contractions is always produced by two or more shocks following one another in rapid succession. Both these reactions are illustrated by the tracings in Figs. 5 and 6. In Fig. 5 the lateral nerve on the left side is being stimulated by break shocks of gradually increasing intensity. The numbers indicate the 480 A. F. Carlson. position of the primary coil with reference to the secondary in centi- metres. It is true that at 6, 5,and 4, the strength of the contractions is somewhat greater than at 7, but even with the secondary pushed clear over the primary (0), the contraction is not much greater than that produced by the minimal stimulus to the nerve. There is nothing that may be interpreted as an “ all-or-nothing ” response in these re- Se ee eee en Sain ac i ae a a a es 8 Zi 6 5 4 3 2 1 0 FIGURE 5.— Tracing of the contraction of the first segment on stimulation of the lateral nerves with single induced shocks. The strength of the shock is indicated by the figures which give the distance of the secondary from the primary coil in centimetres. actions, for when small gradations in the strength of the induced shock are employed, corresponding slight gradations in the amplitude of the contractions may be observed; but the maximal response is quickly reached, beyond which no increase in the amplitude of the beat is produced by increasing the strength of the induced shock. The accumulative effects of two or more induced shocks following one another at intervals of from one-eighth to one-fifth second is strikingly shown in Fig. 6. In this particular experiment both the 1 2 3 4 5 6 ae Sa ed oe FIGURE 6. — Tracing from contraction of second and third segments. The lateral nerves stimulated with series of induced shocks. The figures give the number of shocks in the group, showing summation. lateral nerves were placed on the electrodes. The figures give the number of shocks sent through the nerves. When the shocks are applied to the nerves at the rate of six to eight per second, an almost complete tetanus is produced. It is not clear whether this summa- tion is to be referred to the nerve-fibres, so that a series of shocks in rapid succession produce a single strong nervous impulse, or whether the summation is in the muscle cells. The latter appears the more probable. That would imply that the heart-muscle of Limulus does Nervous Origin of the Heart-Beat in Limulus. 481 not conform to the “ all-or-nothing ” law, as usually interpreted for the vertebrate heart-muscle, and we shall see later that this is actually the case. The normal automatic beat of the Limulus heart is not maximal, as a stimulus applied at the beginning of systole may increase the strength of the contraction. Stimulation of the lateral nerves with the interrupted current pro- duces a continuous supermaximal contraction or tetanus of the heart- ee eS ee ——_li i a FIGURE 7.— Tracings of contraction of first segment. Lateral nerves stimulated with the interrupted current. a, from a fresh and vigorous preparation; 4, from a fatigued preparation. Tetanus. ; muscle. It is evident from the record in Fig. 6 that application of induced shocks to the nerves at the rate of eight to ten per second suffices to produce an incomplete tetanus, but in these experiments the automatic interrupter of the inductorium was usually employed. Fig. 7a gives a typical curve of complete tetanus of the muscle of the first two segments on application of the interrupted current to the lateral nerves in a fresh and vigorous preparation. The particular character of the curve is the rapid attainment of the maximal degree Se eee < < a FicureE 8.— Tracing from first segment on stimulation of lateral nerves with the inter- rupted current of increasing strength. Showing more complete tetanus with the stronger current. of contraction, and the slow but steady relaxation that follows imme- diately upon the muscle’s reaching this greatest degree of shorten- ing. In this regard the form of the tetanus curve differs from that of the vertebrate skeletal muscle. When a preparation is fatigued, or in a poor condition, this immediate fall of the lever becomes more pronounced, as will be seen from Fig. 74. The relaxation also sets in sooner and proceeds more rapidly the weaker the strength of the interrupted current (Fig. 8). It is not clear whether the rapid fall of the curve is due to failure or fatigue of the muscle or of the nerve- fibres. It will probably be found that both factors are concerned. A82 A. F. Carlson. When the nerves are being stimulated with a relatively weak in- terrupted current for some time, the tetanus curve usually exhibits the irregularities shown in Fig. 9. As the muscle relaxes, inco-ordi- nate or fibrillar contractions take the place of continuous contraction. This condition, not unlike the delirium of the vertebrate heart, is evidently only an incomplete tetanus, and may be due to fatigue of the muscle. I have shown in a former paper! that stimulation of the cardio- accelerator nerves in molluscs produces contractions in the quiescent heart. To the various species of molluscs cited in that paper, in which this reaction may be observed, may now be added the large marine gasteropod Sycotypus. Stimulation of the nerve passing from the right visceral ganglion to a ganglion on. the aortic end of the ventricle, or stimulation of the latter ganglion itself with the weak interrupted current, produces a series of beats or incomplete tetanus _[ rr FIGuRE 9.— Tracing from the first segment. Lateral nerves stimulated with a weak interrupted current. Tetanus curve passing into incomplete tetanus on long continued stimulation. in the quiescent ventricle. But in reviewing the present status of the question whether the quiescent vertebrate heart can be made to beat by stimulation of the sympathetic nerves, I overlooked Stewart’s observations? to the effect that the stimulation of the sympathetic produces contractions in the frog’s heart brought to a standstill by raising the temperature of the heart. Stewart gives evidence to the effect that the heat standstill of the heart is not a vagus standstill. That the heart, including the sinus and the great veins, is really quiescent in the so-called heat standstill can readily be determined by aid of a lens. Actual motor effects, that is, the causing of a con- traction in the quiescent heart-muscle on stimulation of any extrinsic or intrinsic cardiac nerves cannot, according to Engelmann, be ad- mitted on the myogenic theory of the heart-beat,® as that would imply 1 CARLSON: This journal, 1904, xii, p. 55. 2? STEWART: Journal of physiology, 1892, xiii, pp. 92, 93. 8 ENGELMANN: Das Herz und seine Thatigkeit im Lichte neurer Forschung, Leipzig, 1903. Nervous Origin of the Heart-Beat tn Limulus. 483 so great a modification of the theory as to virtually amount to aban- doning it. Such true motor effects are, however, produced in the molluscan and the arthropod heart. Ill. Tue NERVE-CorD ON THE DorRSAL SIDE OF THE HEART A REFLEX CENTRE. Are the ganglion cells in the nerve-cord “automatic,” or is their activity dependent on or influenced by impulses from intrinsic sensory nerves? The heart of Limulus lends itself admirably to the solution of this question. For studying the influence of possible afferent nerves on the ganglion cells, the heart is prepared in the manner shown in Fig. 10. One of the lateral nerves is carefully isolated from e &«* FicurE 10.— Preparation of the heart for the study of the local cardio-reflex. Dorsal view. zc, nerve-cord; /z, lateral nerves; #, hooks for connecting a region of the heart with the lever; ¢, electrodes. the muscle in the first three segments, and the isolated nerve stimu- lated. Any region or level of the heart may be connected with the recording lever, as this dissection of the nerve does not affect the rhythm or the co-ordination. The only effect is a diminution in the strength of contraction of the first three segments. If it is desired to study the effects of stimulation of both lateral nerves, it is best to connect the fifth or the sixth segment with the lever, as the dissec- tion greatly diminishes the strength of the contraction in the anterior segments. Now, when one or both lateral nerves prepared in this manner are stimulated with the weak interrupted current, the strength of the beats zs increased, the rate of the beats may or may not be augmented, and zn case the rhythm has ceased from exhaustion, a series of beats are produced during the stimulation. The augmentation of the amplitude of the beats is shown in the tracing reproduced in Fig. 11. In this particular experiment the tracing records the contraction of the 484 A. F. Carlson. seventh segment ; but the augmentation of the strength of the beats is evidenced in every part of the heart that remains in physiological connection with the nerve-cord. When the heart is beating vigor- Ln niiiiirnnnnnnninn ANN annnnone FIGURE 11. — Tracing from the first segment, showing reflex augmentation of the rhythm on stimulation of the lateral nerve in the manner shown in Fig. 10. ously, the augmentation is usually less than shown in Fig. 11. The augmentation in the strength of the beat is usually accompanied by a quickened rate. In case the heart is beating very slowly, the aug- mentation of the rate may be very marked (Fig. 12). wnt ARK FIGURE 12.— Tracing from sixth segment. Reflex augmentation of the rhythm on stimulation of the lateral nerves in the second segment. Heart beating slowly. These effects of stimulating the lateral nerves in the anteropos- terior direction are duplicated by stimulating any of the large i —— FicureE 13.— Tracing from seventh segment. Lateral nerves stimulated in the second segment. Heart quiescent from exhaustion. Showing reflex contractions of quiescent heart. branches leading from the nerve-cord to the lateral nerves in the middle region of the heart (Fig. 14 a). This was to be expected, as in this way we are stimulating the same neurones as in the case of the lateral nerves, only a little nearer the nerve-cord. When any one of these main connectives between the nerve-cord and the lateral Nervous Origin of the Heart-Beat in Limulus. 485 nerves is stimulated with a strong interrupted current, the very oppo- site effect, or inhibition of the rhythm is produced (Fig. 144). I have reasons for thinking that this is not a true reflex inhibition, but due to the escape of the current directly to the ganglion cells in the nerve-cord. When these connectives are stimulated, the electrodes are at the most only 8 to 10 mm. from the cord, and as this inhibi- tion is obtained only with very strong currents, escape of the current SC — 0, ———=$=$_ .. 2 b FicurE 14.—Tracing from second segment. Stimulation of one of the connectives between the cord and the lateral nerves in the sixth segment. a, weak current; 4, very strong current. Showing opposite effects. to the cord is very probable. A certain strength of the interrupted current applied directly to the nerve-cord produces inhibition. If there were any local inhibitory reflex mechanism in the heart, it ought to be revealed on stimulation of the lateral nerves 4 to 5 cms. dis- tant from the cord; but when the position of the electrodes is removed that distance from the cord, even the greatest strength of the inter- rupted current produces only acceleration. Still, the possibility of the presence of a local inhibitory reflex mechanism must be admitted. ANAM ttey ih HAA NM \\ \ \ Mt AEANE KN Ny | Figure 15.— Tracing from the first segment. Stimulation of one of the connectives between the cord and the right lateral nerve in the fifth segment. Strong interrupted current. Showing rhythmical variations in the strength of the beats. WY The discussion of the nature of the inhibition produced by stimulating the cord directly, will be deferred to a later paper. The inhibitory effects produced by stimulating the connectives with a strong inter- rupted current persist much longer than the stimulation, and occa- sionally there appears a peculiar rhythmical variation in the strength of the beats before the original rhythm is resumed. A tracing show- ing these periodic variations is reproduced in Fig. 15. The base line remains horizontal, hence the variations in the excursions of the lever are not “tonus” contractions. They are actual variations in the 486 A. F. Carlson. strength of the beat. Hence if one could have measured the blood- pressure in the aorta, one would have obtained typical “ Traube- Hering-waves.” The ‘“ Traube-Hering-waves” in the blood-pressure of mammals are ascribed to rhythmical variations in the size of the arterioles.1_ A periodic variation in the strength of the ventricular beat would produce the same effects on the arterial blood-pressure. These periodic variations in the strength of the beat in the heart of Limulus are due, not to any variations in the condition of the heart- muscle, but to variations in the activity of the nervous mechanism, — possibly the ganglion cells, as I have obtained tracings exhibiting these periodic variations after subjecting the nerve-cord to the action of certain solutions under conditions which absolutely prevented the solutions from coming in contact with the muscle of the region of the heart from which the records were taken. The reflex contractions produced in the quiescent heart by stimu- lating the centripetal nerve-fibres in the lateral nerves are shown in the record reproduced in Fig. 13. In all the experiments I satisfied myself, by examining the heart with a lens, that it was actually quies- cent in every segment. Only a few contractions, or at most a short series of contractions, follows the stimulation, even when the strength of the current is considerable. This is evidently due to the exhaus- tion of the nervous mechanism. But the contractions are sufficient in number to be recognized as a part of a true motor reflex. These effects of stimulating the isolated lateral nerves of the first segment in the anteroposterior direction are also produced by stimu- lating the isolated lateral nerves of the last two segments in the pos- teroanterior direction. But as the lateral nerves are tiny and not so readily isolated in the eighth and ninth segments, nearly all the experiments were carried out on the lateral nerves of the anterior end of the heart. Sufficient work was, however, done on the nerves of the posterior end to prove that afferent nerve-fibres pass to the nerve- cora from every region of the heart. What reasons are there for attributing these effects on the activity of the ganglion cells to stimulation of centripetal or sensory nerves? May they not be simply ‘axon reflexes,’ in Langley’s sense? The following facts speak against ‘axon reflexes.” The change in the rhythm involves every part of the heart-muscle not severed from its connection with the nerve-cord. This is the case whether one or 1 WoLF and PLuMIER: Journal de physiologie et pathologie générale, 1904, vi, p. 213. Nervous Origin of the Heart-Leat tn Limulus. 487 both lateral nerves are stimulated, as well as when the electrodes are applied to any one of the main connecting branches between the nerve-cord and the lateral nerves. On the “ axon reflex”’ hypothesis, every motor neurone in the nerve-cord would thus have to send axis cylinder processes to practically every muscle cell in the heart, which is highly improbable, if not impossible. The principle of reciprocal conduction in nerve-fibres does not suffice as an explanation of the augmented rhythm. It appears to me fairly well established in the case of the vertebrates, that the nervous impulses that pass cen- tripetally in the motor nerve-fibres do not produce any effect on the nerve-centres. Sensation or motor responses cannot be elicited in that way. There is no reason for supposing that the physiological polarity of the conducting elements in the nerve-centres is less marked in Limulus than in the vertebrates. We must, therefore, fall back on the most natural explanation, viz., that the change in the rhythm. on stimulation of the lateral nerves is a true reflex. The nerve-cord is the reflex centre. Sensory nerves pass from the walls of the heart to the nerve-cord. The existence of this local cardiac reflex mechanism appears to me sufficient evidence of its being func- tional in the normal heart activity, but whether the rhythmical activ- ity of the ganglion cells is actually caused by or merely influenced by these centripetal nervous impulses cannot be answered without further study. | That the rhythm of the vertebrate heart can be influenced reflexly by stimulation of sensory nerve-endings or nerves in the heart has been shown by Muskens.! But according to this observer, the centre for these reflexes is in the central nervous system and not in the ganglia in the heart itself. That the ganglia in the vertebrate heart function as reflex centres has, to my knowledge, not yet been demon- strated. It is of interest to note that the reflex effects observed by Muskens on the frog’s heart were inhibitory, while the local reflexes in the Limulus heart appear to be solely motor or accelerator. I expect that the local reflexes in the vertebrate heart will also prove to be motor or accelerator. IV. THe EFFECT OF TENSION ON THE HEART WALLS. It is well known that in vertebrates tension on the heart walls up to a certain limit is favorable to the rhythm, and may thus be said to act 1 MuSKENS: Archiv fiir die gesammte Physiologie, 1897, Ixvi, p. 328. 488 A. F. Carlson. as a stimulus. The beats become stronger, and the rate may be augmented. Tension may also cause a beat or a series of beats in the quiescent heart. Thus it is said that by raising the pressure of the blood in the ventricle the apex of the frog’s ventricle rendered quiescent by Bernstein’s crushing is made to beat rhythmically. Gaskell believes that mechanical tension may cause a rhythm in iso- lated strips from the apex of the tortoise’s ventricle. I have at times obtained a series of beats of the intact but quiescent ventricle of the dog simply by stretching the ventricular wall. This stimulating effect of tension on the heart walls is very manifest in the inverte- brates, as shown by the works of Foster and Dew-Smith, Biedermann, Ransom, Schoenlein, and Straub.!. These observers worked on the hearts of various molluscs, all obtaining substantially the same results. These results I have confirmed on the hearts of several molluscs. Up toa certain limit, the augmentation of the rhythm is directly proportional to the pressure of the fluid in the cavity of the heart. If the pressure is rapidly increased, the heart may go into a condition of tonus or fibrillary contractions. Simple mechanical ten- sion exerted on the heart wall is on the whole less efficient as a stim- ulus to the rhythm than the pressure exerted by a liquid in the heart cavity. Yet in some molluscs (for example, the giant California slug Ariolimax), there is little difference between the results of these two ways of producing tension in the heart walls. Theexcised and there- fore empty and collapsed ventricle of the slug (Ariolimax) beats with perfect rhythm for hours. If the ventricle is suspended and loaded with a light recording lever, the rhythm becomes more rapid and vigorous. The stimulating action of the tension from the load is shown still more clearly when the empty heart has been beating for six to eight hours, and is becoming exhausted. With the ventricle in an exhausted condition, releasing the tension stops the rhythm, and re- applying the tension is followed by the reappearance of the rhythm. Simple mechanical tension on the ventricular walls is thus able to pro- duce a rhythm in the guiescent ventricle. But this does not prove that the rhythm is produced by the tension acting on the muscle. The heart of arthropods is not so well adapted to studying the 1 FosTER and DEw-SMITH: Proceedings of the Royal Society, 1875, xxiii, p- 318 ; BIEDERMANN: Sitzungsberichte der Wiener Akademie, 1884, lxxxix, 3, p- 191; RANSOM: Journal of physiology, 1884, v, p. 261; SCHOENLEIN : Zeitschrift fur Biologie, 1894, xxx, p. 187; STRAUB: Archiv fiir die gesammte Physiologie, 1901, 1xxxvi, p. 504. Nervous Origin of the Hleart-Beat in Limulus. 489 effects on the rhythm of varying the pressure of the liquid in the heart cavity, because of the several ostia opening into and the several arte- ries leading from the heart. In the Limulus heart the pressure in the cavity may be raised by ligaturing the lateral arteries and admitting plasma or sea-water by a cannula at the aorticend. ‘The pressure of the liquid may be varied by varying the inclination of the cannula. The Limulus heart responds to variations in pressure in every way like the heart of vertebrates and molluscs. Increasing the pressure up to acertain limit, augments the rhythm, a further sudden increase causes inco-ordination, fibrillary contractions, and incomplete tetanus or tonus, just as Biedermann observed in the ventricle of the snail. If the Limulus heart is quiescent from exhaustion, the mere filling of the heart with plasma or sea-water under slight pressure may start a series of beats. That the tension or pressure of the liquid on the heart walls, and not any chemical action, is the stimulating factor seems to be shown by the fact that immersion of the whole heart in the plasma or sea-water does not produce similar results. Augmen- tation of the rhythm can also be produced by simple mechanical ten- sion on the heart. If the suspensory ligaments attached to either lateral angle of the heart in the fifth or sixth segments are gently stretched, and the heart thus put under tension, the rhythm of the whole heart is augmented, despite the fact that the tension affects at the most only two segments. The same results may be obtained by inserting hooks at opposite angles of the heart in the middle region, securing the one toa fixed support, and to the other one attaching weights of different values. When mechanical tension is thus applied in the middle region, I have even succeeded in starting a series of beats in perfectly quiescent hearts, that is, in hearts quiescent from exhaustion. The mechanical tension is most efficient in the middle region of the heart. When the first two or the last two segments are similarly stretched, the augmentation, if produced, may be confined to the region of the tension, thus resulting in inco-ordination of the rhythm. Lhe pressure of the liquid in the heart cavity and the simple stretch- mg of the heart walls produce these effects only in hearts with the ganglion or nerve-cord intact. After extirpation of the nerve-cord I have never succeeded in producing a beat or a series of beats by these means. When the nerve-cord is extirpated, the heart becomes quiescent. Now if hydrostatic pressure or mechanical tension acted as stimuli only on the beating heart, we would not get far in our 490 A. F. Carlson, analysis of the nature of these stimuli, but inasmuch as they produce a series of beats in the quiescent heart as long as the ganglion is intact, but not after extirpation of the ganglion, it is evident at once that their action is in some way dependent upon the nerve-cord. In the first place, tension does not produce a rhythm by direct action on the heart-muscle or on the motor nerves and nerve-endings. We know, however, that the stretching of a nerve or nerve-fibre up toa certain limit increases its excitability. It is therefore not impossible that the tension on the heart walls increases the excitability of the motor fibres so that the nervous impulses reaching the heart-muscle from the ganglion cells become of greater strength. That may suf- fice to explain the augmentation of the amplitude of the beats, but it does not account for the increased rate. That must be due either to direct action of the tension on the ganglion cells, or to stimulation of sensory nerve-endings and fibres in the walls of the heart, making it a true reflex action. We have shown that such a local reflex mechanism exists in the heart of Limulus, and it is not improbable that these sensory nerves are readily stimulated by tension. The tension or pressure may furthermore act directly on the ganglion cells. Theoretically it ought to be possible to produce a rhythm in the heart of Limulus by simple tension of the heart walls, even after extirpation of the nerve-cord. By stretching a motor nerve up toa certain limit, we can produce a contraction or a series of contractions in the muscle supplied by the nerve. Now if stretching the heart walls put sufficient tension on the motor nerve-fibres to start a ner- vous impulse, the contraction of the muscle would immediately release the tension on the nerves, while the subsequent relaxation of the muscle would again allow the full strength of the tension to act on the nerves, resulting in a second stimulus of the nerve-fibres and a second contraction of the heart-muscle, etc. But the motor nerve- fibres in the heart of Limulus are evidently sufficiently elastic not to be stimulated by a moderate amount of stretching of the heart walls. There may, however, be hearts having sufficiently inelastic and excitable motor nerves for tension to produce rhythmical contractions in this manner in the absence of ganglion cells or a reflex nerve- centre. The stimulating action of tension in the vertebrate heart is usually interpreted as being on the heart-muscle directly. In view of these results on the Limulus heart, it would seem that the ques- tion should be re-examined in the vertebrates. Nervous Origin of the Heart-Beat in Limulus. 491 V. THE INFLUENCE OF SODIUM CHLORIDE ON THE HEART, WITH AND WITHOUT THE NERVE-CORD. When a Limulus heart is removed from the plasma and placed in an isotonic (372) sodium chloride solution, the rhythm is at once augmented. The rate of the beats is particularly increased. A heart that is quiescent from exhaustion starts to beat within one to three minutes after being placed in an isotonic sodium chloride solution. It does not concern us for the present how long this rhythm is main- tained. It suffices to note that an isotonic sodium chloride solution augments the rhythm of the beating heart at once or after a latent period of one to two seconds, and starts a series of beats in the quiescent heart after a latent period of not exceeding three or four minutes. Now when the heart is placed in the sodium chloride solution after extirpation of the ganglion, it remains perfectly quies- cent for thirty to forty-five minutes. The same is true whether the _whole heart is immersed in the solution or only a portion of the heart wall is used. There is, then, this difference in the action of the sodium chloride on intact and on the ganglion-free heart ; viz., in the first case, the solution stimulates at once, in the latter case, it also stimulates or causes a rhythm of the heart-muscle, but only after a latent period of more than thirty minutes. The action of the solution on the intact heart is therefore an action on the ganglion, or a local reflex dependent on the integrity of the ganglion. The belated action on the heart deprived of the nerve-cord must be an action on the motor. nerve-fibres or on the muscle cells directly. It is only in a pure sodium chloride solution that this rhythm is developed in a heart from which the nerve-cord has been removed. The rhythm does not appear when the heart is left in plasma, or sea-water, or in an artificial solution containing the chlorides of sodium, calcium, and potassium in approximately the same proportions as the sea-water, while the heart with the nerve-cord intact beats rhythmically in either of these solutions, as well as in the air, when protected from evaporation. The ganglion-free heart of Limulus thus reacts to a solution of pure sodium chloride in the same manner as do vertebrate skeletal muscle or the apex of the frog’s and tortoise’s ventricle. For notwithstanding the observations of Gaskell, it has recently been asserted! that the ventricular tissue of the tortoise’s heart is not spontaneously rhythmical under the conditions of normal life. 1 MARTIN: This journal, 1904, xi, p. 103. Ag2 A. F. Carlson. The nature of the action of sodium chloride on muscle in produc- ing rhythmical contractions is not yet understood, but this action of the salt is frequently cited as a proof of the myogenic theory of the heart-beat. That inference is entirely unwarranted, for the combina- tion of salts in the blood does not stimulate muscle or nervous ele- ments in this way, and no tissue is bathed in a pure sodium chloride solution under normal conditions. VI. DrrREcT STIMULATION OF THE HEART. It has been claimed that the heart of cephalopods and tunicates exhibits a refractory condition during systole similar to that of the heart of the higher vertebrates! This is not true if we understand by refractory period a condition of inexcitability. The excitability of the arthropod, the molluscan, and the tunicate heart is lowest at the beginning of systole, but a strength of the induced shock can be found which affects the heart in any phase of contraction.? The heart of Limulus makes no exception to this rule. An induced shock of sufficient strength sent through the heart at the beginning of systole produces a supermaximal beat. The so-called “all-or- nothing” law is, furthermore, not applicable to the invertebrate heart. I have recently shown that the ‘“all-or-nothing” principle and the principle of absolute inexcitability during systole does not apply to the heart of the lowest vertebrates.? It is true that the heart, especially that of cephalopods and crustaceans, when in good condition, tends to respond with a beat of uniform strength to stimuli of varying intensity within a wide range, but increase in the strength of stimulus above this range produces supermaximal contractions. The heart of pulmonates and the heart of Limulus exhibits perhaps the least of this tendency to a response of uniform strength. A Limulus heart with the nerve-cord tntact exhibits a much greater excitability and a greater tendency to uniform contractions in response to stimuli of varying strengths than does the same heart after the nerve-cora has been extirpated. There is the further difference that a single induced shock, however strong, sent through a ganglion-free heart produces only a single contraction, while the same shock ap- plied to a quiescent but intact heart may produce a series of beats. 1 Ransom: Journal of physiology, 1884, v, p. 261; SCHULTZE: Jenaische Zeitschrift fiir Naturwissenschaft, 1901, xxxv, p. 221. 2 CARLSON: Science, 1903, xvii, p. 548. 8 CARLSON: Zeitschrift fiir allgemeine Physiologie, 1904, iv, p. 259. Nervous Origin of the Fleart-Beat in Limulus. 493 The single induced shock is evidently not able to produce a series of beats when acting on the motor nerve-fibres and the muscle alone, while a single stimulation of the ganglion cells may cause a rhythmical activity of the latter. The greater tendency of the heart to uniform response when the ganglion is left intact, is illustrated by the tracings ht dt tt pt at yt tt A i FIGURE 16.— Contractions in response to direct stimulation of the heart with single- induced shocks of increasing strength. 4, from the third segment, with the nerve- cord intact; B, from the sixth segment, with the nerve-cord extirpated. Showing tendency to greater uniformity in the contractions when the cord is intact. in Fig. 16. The upper record is from the contraction of the third segment in an intact heart, the latter from the contraction of the sixth segment in a ganglion-free heart. In both cases the induced shocks are sent through the heart from side to side in the segment connected with the recording lever. It is needless to add that the heart of Limulus can be tetanized by direct stimulation with a strong interrupted current, as that follows WU FIGURE 17.— Tracing from first segment. At X the nerve-cord was touched with the forceps in the fifth segment. Showing acceleration of the rhythm on mechanical stimulation of the ganglion. x x necessarily from the fact that the muscle can be tetanized by stimu- lation of the motor nerves. The difference in the response of the heart to direct stimulation, according as the ganglion is intact or extirpated, goes to show that at least a part of the peculiarity of the properties of cardiac tissue in response to direct stimulation is due to the presence of nerve-centres or ganglion cells, and a relatively low excitability of the motor nerve- fibres and the muscle to the induced current. It is scarcely neces- A494 A. F. Carlson. sary to point out that we do not yet know the properties of the heart- muscle apart from the intrinsic heart-nerves, as we know of no drug that will paralyze the motor nerve- endings i in the heart without injury to the heart-muscle.! Since the recent (1902) comprehensive review of the literature on the vertebrate heart with special reference to the neurogenic or the myogenic nature of the heart-beat, and the co-ordination in the heart by Langendorff,? no contributions to vertebrate heart physiology of decisive bearing on these questions have been made. But it appears to me not out of place in this connection to call attention to some erroneous statements in current literature regarding the presence of nervous elements in the heart of invertebrates. Thus Langendorff says in his otherwise able review: “It has been determined by numerous researches that in the case of the heart of many inverte- brates (molluscs, arthropods, tunicates, crustaceans), the presence of ganglion cells and nerve-fibres cannot be demonstrated (pp. 334, 339)- In his recent (1903) address on the physiology of the heart, Engel- mann gives expression to the same view: “In the heart of higher invertebrates (tunicates, molluscs, arthropods) Foster and his stu- dents were unable to find ganglion cells, and the painstaking re- searches of later investigators Caaahaayy ° Knoll, Straub, Schultze, Heine) have led to the same results” (p. 10). Practically the same statement is found in many textbooks of physiology. Porter puts it in this way: ‘The hearts of many invertebrates in which ganglion cells are apparently absent beat rhythmically.” ® Turning our attention first to the molluscs, we find that Foster,* Foster and Dew-Smith,® Darwin,® and Biedermann? claim that there are neither nerve-cells nor nerve-fibres in the heart of the snail. Darwin cautions us against placing too much value on his negative results. Biedermann had recourse only to the method of macerating the heart in caustic potash and teasing. Results similar to these 1 HERING, H. E.: Archiv fiir die gesammte Physiologie, 1903, xcix, p. 253; CARLSON: Science, 1904, xx, p- 684. 2 LANGENDORFF: Ergebnisse der Physiologie, 1902, i, pp. 263, 345. 3 PORTER: American textbook of physiology, 1900, i, p. I5I. 4 Foster: Archiv fiir die gesammte Physiologie, 1875, v, p. 191. 5 FosTER and Dew-SmirH: Proceedings of the Royal Society, 1875, xxiii, p. 318; Archiv fiir mikroskopische Anatomie, 1877, xiv, p. 317. 6 DARWIN: Journal of anatomy and physiology, 1876, x, p. 506. 7 BIEDERMANN: Sitzungsberichte der Wiener Akademie, 1884, lxxxix, 3, pera, Nervous Origin of the Heart-Beat in Limulus. 495 were obtained by Knoll? on the heart of some heteropods. This investigator was equally unable, making use of the gold chloride method, to find nerves or ganglion cells in the heart. These obser- vations are certainly erroneous as regards the presence of merve-fibres in the heart, as shown by the works of Ransom,? Young,? Dogiel,* Bottazzi and Enriques,® and Budington.6 Ransom has shown both by histological and physiological methods that the snail heart is pro- vided with nerves. Young and Budington have proved that the heart of several lamellibranchs is supplied with nerves from the visceral ganglion. The same has been shown histologically by Dogiel. Ransom and Bottazzi and Enriques have further demon- strated that the heart of Aplysia is similarly provided with nerves. That the heart of the cephalopods is provided with nerves has been known since the observations of Paul Bert on Sepia, more than thirty years ago. To these observations showing the presence of nerves in the heart of various molluscs may be added my own (the details of which have not yet been published), made on several representatives from every group in the phylum, and they are to the effect that cardio-regulative nerves, accelerator or inhibitory, or both, are uni- formly present. This makes it highly probable that the molluscan heart ts without exception provided with cardio-regulative nerves. Regarding the presence of werve-cells in the molluscan heart, it is true that Foster, Foster and Dew-Smith, Darwin, Biedermann, Ransom, and Knoll failed to find any, but these observers (excepting Ransom) also failed to find nerve-fibres in the heart, although the nerve-fibres are certainly there. Their methods were therefore at fault, and it is probably just as erroneous to conclude from their observations that no ganglion cells are present in the heart, as it would be to conclude that cardio-regulative nerves are absent. We have, moreover, positive evidence of the presence of ganglion cells in the molluscan heart. Dogiel (/oc. cit.) described and figured nerve cells in the auricles and the ventricles of Helzx, Picten, Ana- donta, and Aplysia. Haller’ describes'and figures both multipolar 1 KNOLL: Sitzungsberichte der Wiener Akademie, 1893, cli, 3, p. 387. 2 Ransom: Journal of physiology, 1884, v, p. 261. 8 YounG: Archives de zoologie expérimentale, 1881, ix, p. 429. 4 DoGIEL: Archiv fiir mikroskopische Anatomie, 1877, xiv, p. 59. 5 BoTTAZzI and ENRIQUES: Archives italiennes de biologie, 1901, xxxv, p. III. 6 BuDINGTON: Biological bulletin, 1904, vi, p. 311. 7 HALLER: Zoologisches Jahrbuch, 1883, ix, p. I. 496 A. F. Carlson. and bipolar ganglion cells in the heart of Fzsurell/a and the Muricide, and he traces the processes from the cells to nerve plexuses about the muscle-fibres. The same author describes under the ectocardium of the heart of Chitons a plexus of ganglion cells and nerve-fibres very similar to that in the sinus and the auricles of the vertebrates.1_ More than fifty years ago Hancock and Embleton ? described a ganglion on the aortic end of the heart of Doris. This I have confirmed on a related species of the Pacific coast. Inthe marine gasteropod Sycoty- pus, | founda similar ganglion at the ventriculoaortic junction. Iam not familiar with the work of Heine, referred to by Engelmann in the statement cited, but Engelmann, citing Straub’s work as showing the absence of nervous elements in the heart, is certainly erroneous. For I take it that Engelmann refers to Straub’s work on the heart of Aplysia.® In that paper Straub assames the absence of nervous elements in the heart, but he brings no proofs, either histological or physiological, in support of the assumption, and he furthermore appears to have been ignorant of the observations of Dogiel and Ransom proving the presence of cardio-accelerator nerves to the Aplysia heart. In the case of the Arthropod heart there can be no question of the presence of cardio-regulative nerves from the thoracic ganglion, at least in the crustaceans,’ Knoll’s negative results notwithstanding. I have confirmed the earlier observers both on Macrura and Brachy- ura. The papers of Milne-Edwards,® and Patten and Redenbaugh ® leave no doubt as to the presence of werves in the heart of Arach- nids. The presence of ganglion cells in the heart of the Arachnids (Limulus) is also conclusively proven by Patten and Redenbaugh, and they are demonstrated with such comparative ease that any one in doubt as to the accuracy of the observations of Patten and Reden- baugh may readily convince himself of the presence of ganglion cells in the nerve-cord and of their true nervous nature. The evidence for 1 HALLER: Arbeiten aus dem Zoologischen Institut der Universitat in Wien, 1882, iv; 1883, v. ° 2 Hancock and EMBLETON: Philosophical transactions, 1852, Part I, p. 207. 3 STRAUB: Archiv fiir die gesammte Physiologie, 1901, Ixxxvi, p. 504. 4 PLATEAU: Bulletin de l’académie royal de belgique, 1878, xiv; Archives de biologie, 1880, i, p. 595; YouNnG: Archives des zoologie expérimentale, 1878, vii, p. 401; JOLYET and VIALLANES: Annales des sciences naturelles, zoologie, 1892, xiv, p. 387; CONANT and CLARKE: Journal of experimental medicine, 1896, i, p- 341: Borrazzi: Centralblatt fiir Physiologie, 1901, civ, p. 663. 5 MILNE-EDWARDS: Annales des sciences naturelles, 1873, xvii, ser. 5. 6 PATTEN and REDENBAUGH: Journal of morphology, 1899, xvi, p. 91. Nervous Origin of the Heart-Beat in Limulus. 497 the presence of ganglion cells in the crustacean heart is no less con- vincing. It is true that some of the early observers failed to find nerve-cells in the heart of crabs and crayfishes, but the positive results of Berger,! Dogiel,?, Deszé,3 and Pogoschewa* must take precedence over the earlier negative findings. It is inconceivable that the elements figured and described by Dogiel in one of his late papers on the crustacean heart (1894) can be anything but nerve- cells. Dogiel has, furthermore, described nerve-cells in the heart of the Corethra larva. The tunicate heart was the last to give up its secrets in that nervous elements in it were sought for in vain up till very recently. The reader will find the references to the earlier observations in the recent paper by Schultze. Schultze was equally unable to find any nerves or nerve-cells in the heart of Ciova or Salpa. But all these failures, it appears to me, do not count against the clear demonstra- tion of nerve-cells and nerve-fibres in the heart of Molgula by Hunter.’ It is significant that in the tunicate heart the ganglion cells are massed particularly at either end, that is, in the two regions ‘where the reversible rhythm originates. Nerves and nerve-cells have thus been shown to be present in the heart of a great number of molluscs and arthropods, and at least in one tunicate. There can be little doubt but that with greater accuracy in histological and physiological methods and observations, nervous elements will be found to be present in the heart of all invertebrates. But even on the basis of the data so far at hand, it is obvious that statements like those cited on page 504 are wholly erroneous. SUMMARY. The ganglion cells of the venous end (fifth, sixth, and seventh segments) of the heart of Limulus are more numerous and exhibit greater automatism than the ganglion cells of the aortic or anterior end. From the ganglion cells in the fifth, sixth, and seventh seg- 1 BERGER: Sitzungsberichte der Wiener Akademie, 1876, Ixxiv, Abt. i, p. 422. 2 DoGIEL: Archives de physiologie, 1877; Archiv fiir mikroskopische Anatomie, 1894, xliii, p. 223. 8 DeEsz6: Zoologischer Anzeiger, 1878, i, p. 126. * PoGOSCHEWA: Bote fiir Naturwissenschaften, 1890, No. 5. 5 DOGIEL: Mémoires de l’académie St. Petersburg, 1877, xxiv, No. Io. 6 SCHULTZE: Jenaische Zeitschrift fiir Naturwissenschaften, 1901, xxxv, p. 22I. 1 HUNTER: Anatomischer Anzeiger, 1902, xxi, p. 241; Science, 1903, xvii, No. 424; This journal, 1903, x, p. I. 498 A. F. Carlson. ments nerve-fibres pass in the nerve-cord and the lateral nerves directly to the heart-muscle of every segment. The nerve-fibres passing from the ganglion cells in the nerve-cord to the heart-muscle are ordinary motor fibres. They exhibit a very low excitability to the induced current ; nevertheless the heart-muscle responds to single induced shocks applied to the nerves.- Stimula- tion of the nerves with the interrupted current produces tetanus of the heart-muscle. The ganglion or nerve-cord on the dorsal side of the heart is a reflex centre. Stimulation of the sensory or centripetal nerve-fibres passing from the walls of the heart to the nerve-cord augments the rhythm, and may start a rhythm in a heart that is quiescent from ex- haustion. There is no evidence of the presence of a local inhibitory reflex mechanism. On direct stimulation of the heart, the amplitude of the contraction varies, within limits, directly with the strength of the stimulus; but there is a greater tendency to uniform contractions in response to stimuli of varying intensity in hearts with the nerve-cord intact, than in hearts from which the nerve-cord has been removed. A certain degree of mechanical tension on the heart walls, or pres- sure of an indifferent liquid in the cavity of the heart, augments the rhythm, and may start a rhythm in hearts that are quiescent from exhaustion. These effects are produced only in case the nerve-cord is intact. After extirpation of the ganglion, neither hydrostatic pressure in the heart cavity, nor mechanical tension on the heart walls, produces contractions. The action of inorganic salts, particularly sodium chloride, in solution on the ganglion appears to be of the same nature as that on the muscle, except that they act more rapidly on the ganglion cells. An isotonic sodium chloride solution stimulates the ganglion practi- cally instantaneously, while the heart from which the nerve-cord has been removed develops a more or less rhythmical series of contrac- tions only after a prolonged (thirty to forty-five minutes) immersion in the solution. Calcium chloride counteracts the stimulating effects of the sodium chloride both on the ganglion and on the muscle. The com- bination of salts in the blood or in the sea-water does not produce contractions of the heart after the nerve-cord has been removed. I am indebted to Prof. G. N. Stewart for valuable criticism of the manuscript. —_— = Fg en ne IN DES TOv.V.OL:. Xt. oa ATION of blood-corpus- cles, Alcohol, ae on excretion of uric acid, 13. Allantoin, production, 85. Autolysis of animal organs, 276. EEBE,.S. P. The effect of alcohol and alcoholic fluids upon the excre- tion of uric acid in man, 13. BEEBE, S. P. The chemistry of malignant growths. II.— The inorganic constitu- ents of tumors, 167. BEEBE, S. P., and B. H. Buxton. The production of fat from proteid by the bacilius pyocyaneus, 466. Biood, laking, 184, 363. Bowen, W. P. See HIGLEY and BowEN, Bits Brain affected by starvation, 116. Brown, O. H. See MATTHEWS and Brown, 173. Brown, O.H. See NEILSON and Brown, 374- Buiot, G. On the swelling of organic tissues. — Researches on the cornea, 297. Buxton, B. H. See BEEBE and BuXToN, 466. ON: W. B. The passage of dif- ferent food-stuffs from the stomach and through the small intestine, 387. Carbamates, estimation of, 444. CARLSON, A. J. The rhythm produced in the resting heart of molluscs by the stimu- lation of the cardio-accelerator nerves, 55. Carson, A. J. ‘The nervous origin of the heart-beat in Limulus, and the nervous nature of co-ordination or conduction in the heart, 67. CARLSON, A. J. Further evidence of the nervous origin of the heart-beat in Limu- lus, 471. Cholin, from lecithin and brain-tissue, 353. Cook. F. C. The chemical composition of some Gorgonian corals, 95. Corals, composition of, 95. CorrAT, I. H. The production of cholin from lecithin and brain-tissue, 353. | Dae electrical potential during devel- opment, 241. Electrical potential in developing eggs, 241- Enzymes, proteolytic, I. Exercise, effect on excretion of carbon. dioxide, 311. | clean produced from proteid by bacillus pyocyaneus, 466. Fever, influence on reducing action of organism, 457. Food, passage from stomach and through intestines, 387. UTHRIE, C.C. The effect of the in- travenous injection of formaldehyde and calcium chloride on the hemolytic power of serum, 139. ime 139, 184, 363. Haskins, H. D. The identity of so-called urcine (Moor), 162. Haskins, H. D. See MACLEOD and Has- KINS, 444. Harat, S. The effect of partial starvation on the brain of the white rat, 116. HarcHer, R. A. The fate of strychnine in the intestiné of the rabbit, 237. Heart, accelerator nerves in molluscs, 55. Heart-beat, nervous origin, 67, 471. Heart, conduction, 67. , co-ordination, 67. HeERTER, C. A. On the reducing action of the animal organism under the influence of cold, 128. 499 500 HERTER, C. A. The influence of fever on the reducing action of the animal organ- ism, 457. HERTER, C. A.,and A. N. RicHarps. The influence of chloroform on intravital stain- ing with methylene-blue, 207. HIGLey, G. O.,and W. P. BowEN. Changes in the excretion of carbon-dioxide result- ing from bicycling, 311. Hypbeg, Ipa H. Differences in electrical potential in developing eggs, 241. Hydrolysis of glands, 276. [BRICAN, 176. Ions in physiological processes, 374. Isotony, 99. K IDNEY, decapsulation of, 304. i of blood by ether, 184. LEVENE, P. A. Hydrolysis of spleen- nucleic acid by dilute mineral acid, 213. LEVENE, P. A. The autolysis of animal organs. II.— Hydrolysis of fresh and self digested glands, 276. LEVENE, P. A., and L. B. SrookKrEy. On the combined action of proteolytic en- zymes, I. LEvIN, LI. ney, 304. Lyon, E. P. On rheotropism. tropism in fishes, 149. On decapsulation of the kid- I. — Rheo- hp Screen, Jie Jo Ika ciel Wet 1D) NSE KINS. The quantitative estimation of carbamates, 444. MatruHews, A. P. The toxic and anti-toxic action of salts, 419. MatrHews, S. A., and O. H. Brown. Inhibition of the action of physostigmin by calcium chloride, 173. MENDEL, L. B., and E. W. Rockwoobp. On the absorption and utilization of pro- teids without intervention of the alimen- tary digestive processes, 336. MENDEL, L. B., and B. WHITE. On the intermediary metabolism of the purin- bodies: the production of allantoin in the animal body, 85. Metabolism of purin bodies, 85. EILSON, C. H., and O. H. Brown. Further proof of ion action in physi- ologic processes, 374. Nucleic acid, hydrolysis of, 213. ~~ 665 L[ndex. VA, changes in anisotonic solutions and in saponin, 99. ARAMCECIUM, reactions of, 220. PESKIND, S. Ether-laking: a contri- bution to the study of laking agents that dissolve lecithin and cholesterin, 184. Physostigmin antagonized by calcium chlo- ride, 173. Proteid absorption without digestion, 336. Proteid, production of fat by bacillus pyo- cyaneus, 466. Purin-bodies, 85. RR eDUCne action of animal organism, 1200-41577 Rheotropism in fishes, 149. RicHArbs, A. N. See HERTER and RICH- ARDS, 207. Rockwoop, E. W. The elimination of endogenous uric acid, 38. Rockwoop, E. W. See MENDEL and ROCKWOOD, 336. OLLMANN, T. Structural changes of ova in anisotonic solutions and sapo- nin, 99. Staining, intravital, 207. Starvation, effect on brain, 116. STEWAR?, G. N. Further experiments on the hzmolysinogenic and agglutinino- genic action of laked corpuscles, 363. Stomach, emptying of, 387. STrookeEy, L. B. See LEVENE and Sroo- KEY, I. Srorey, T. A. Tonus rhythms in normal human muscle and in the gastrocnemius of the cat, 75. Strychnine, fate in the intestine, 237. Swelling of organic tissues, 297. ONUS in mammalian muscle, 75. TowLs, ELIZABETH W. A study of the effects of certain stimuli, single and combined, upon Paramcecium, 220. Toxic and anti-toxic action of salts, 419. Tumors, inorganic constitutents, 167. NDERHILL, F. P. On the origin and precursors of urinary indican, 176. Ureine, 162. Uric acid elimination, 38. Uric acid excretion affected by alcohol, 13. W HITE, B. See MENDEL and WHITE 85. + Ag Wy ; ese. tans beat ae Bes fe ee BINDING SECT. MAR 4 1966 nl American Journal of Physiology A5 Vel2 J cop,.2 As a (q9 Biological & Medica] Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY \GE : . ; - Rae, ono 8 hg 0d 9 2 ne eee w "he pees - 2 5 - arg 2 et ~ + renee een fe 7 . : Sp penne S pit Ta ¥ x ey 7 - ? - . Moe « * <0 DRA et hoe rele i ci de aia 2 a ; a | Pe IF nae ay MAR oe 2 ne, Sa See Se at oO aa Saeed ihn Sie eae ee ae Pad or a Toe: Cpaoe smi eae, SE eRe es ume ee wae — fortes ne Se it b or ee ee ST vans, ai a ep ft ky Ow we Ort fey apanp yee ee tee™ ahetae